The Volokh Conspiracy

In a recent New York Times column, prominent Supreme Court commentator Linda Greenhouse grossly misrepresents the federalism argument against the Defense of Marriage Act put forward in an amicus brief co-signed by several federalism scholars, including co-bloggers Randy Barnett, Jonathan Adler, Dale Carpenter, and myself. She claims that it is a “Trojan horse” for an effort to block same-sex marriage, and that it is somehow inconsistent with various Supreme Court decisions striking down state laws that violate individual constitutional rights protected by the Fourteenth Amendment:

Beware of conservatives bearing gifts.

Ever since last week’s frustrating Supreme Court argument in the Defense of Marriage Act case, I’ve been wondering whether the attack on DOMA will turn out to be a constitutional Trojan horse. It may bring victory: the demise of a spiteful federal statute, enacted by an opportunistic Congress and signed into law 17 years ago by a cowardly Bill Clinton. But at what price?..

[S]triking down DOMA on federalism grounds is a truly bad idea, and the campaign for marriage equality would be worse off for it. To explain the argument is to reveal its dangers. A ruling that left the states to their own devices when it comes to marriage would take the equal protection guarantee out of the picture...

It would, of course, provide the full benefits of marriage to those living in states that chose to recognize same-sex marriage. But it would snatch away the promise for those living elsewhere, particularly if the decision was based not only on the asserted absence of federal authority but on exaggerated notions of state sovereignty anchored in the Tea Party’s favorite constitutional amendment, the 10th....

Just such a vigorous constitutionally based states’-rights argument was put forward in a brief filed in the case last month by a half-dozen law professors under the title, “Brief of Federalism Scholars.” One of the group, Prof. Randy Barnett of the Georgetown University Law Center, is justly regarded as the intellectual father of the Commerce Clause attack on the Affordable Care Act. “DOMA falls outside Congress’s powers” because “marriage is not commercial activity,” the brief declares. It argues further that by limiting federal recognition to marriages between men and women, the law “undermines the states’ sovereign authority to define, regulate and support family relationships.”

There is much that’s questionable about this assertion of implicitly boundless state authority over family affairs. A famous pair of Supreme Court decisions from the 1920s armed parents with rights under the Due Process Clause to educate their children as they see fit, in resistance to state laws. Pierce v. Society of Sisters gave parents the right to choose private or religious schools despite an Oregon law that required public school education for all. Meyer v. Nebraska struck down a state law that barred the teaching of modern foreign language (the law’s post-World War I target was German.)...

And of course the most famous federal intervention of all was Loving v. Virginia, the 1967 decision (shockingly recent) that overturned the laws of Virginia and 15 other states prohibiting marriage between people of different races.

Greenhouse’s argument is based on a fundamental error: she conflates structural limits on Congress’ authority outlined in Article I of the Constitution with individual rights constraints on the states. In reality, it is perfectly possible for a particular law to be both beyond the scope of Congress’ authority if enacted by the federal government and a violation of constitutional individual rights if adopted by a state. Our federalism brief merely claims that Congress lacks the constitutional authority to enact DOMA because the law exceeds the scope of Congress’ enumerated powers. That conclusion is perfectly consistent with the view that state laws banning gay marriage violate individual rights protected by the Fourteenth Amendment. Indeed, I myself have argued that the Court should strike down California’s Proposition 8 because it is unconstitutional sex discrimination. Similarly, we would argue that a federal law banning interracial marriage is also outside the scope of congressional power under Article I, while simultaneously endorsing the holding of Loving v. Virginia that state laws barring such marriages violate the Equal Protection Clause of the Fourteenth Amendment. The states’ sovereign authority over marriage – like all state sovereign authority – is constrained by constitutional individual rights. Nothing in our brief is inconsistent with that basic principle.

Greenhouse’s piece also contains several other errors. For example, our argument in the federalism scholars brief does not rely on the Tenth Amendment, and most of us are not “conservatives.” At least three of the six signers (Randy, Jonathan, and myself) are libertarians. Perhaps most important, the majority of us are actually gay marriage supporters, and therefore it is unlikely that we are somehow trying to “prop up” the anti-gay marriage “status quo,” as Greenhouse suggests. Dale Carpenter is one of the leading academic advocates of gay marriage, and has been involved in that cause and other gay rights issues for many years. Few people are less likely defenders of the “status quo” on these matters than Dale.

Unfortunately, this is not the first time that Greenhouse has misrepresented the views of her opponents on constitutional federalism issues, including myself. There is plenty of room for legitimate disagreement about both our federalism argument and other important issues at stake in the gay marriage cases. But the debate is not advanced by crude misrepresentation of our adversaries’ views.

UPDATE: I have fixed what was previously a broken link to Greenhouse’s column. My thanks to readers for pointing out this problem.

In his last post on the subject, Nick Rosenkranz concludes that “the mere fact of a federal definition, for purposes of federal law, does not violate principles of federalism.” On this we are entirely agreed. And if all Congress sought to do with Section 3 of DOMA was to define the semantic meaning of a word, there would be no problem. But that’s not all Congress sought to do. Further this is not a case in which we (or anyone else) is asking the Court to pour through legislative history to divine Congressional intent, as the true purpose of DOMA has never been contested.

The ultimate question in the DOMA litigation is whether Congress has a legitimate federal interest in having a particular definition of marriage that supports the traditional form, and whether this interest is sufficient to justify the differential treatment (and disregard of marriages recognized under state law) that DOMA produces. One might have a definition so that we can know what the words mean in federal statutes, and yet still not be able to defend them with the same interests that a state might assert (as I explained here). In order to strike down DOMA, the Court need not conclude that federal law must forever and always accept state law definitions. All it needs to do is recognize that defining marriage is a traditional function of the states, not the federal government, and, as a consequence, there is no federal interest sufficient to justify DOMA.

Section 3 of the Defense of Marriage Act (DOMA) defines the word “marriage” for purposes of federal law. In a recent post, Randy argues that this provision must be unconstitutional, because otherwise one could be “married” under state law but not under federal law, or vice versa, and “that is crazy.”

With all due respect, this is not crazy. It is, in fact, an utterly commonplace feature of our federalism that the exact same word may mean different things for purposes of state law and federal law. Ed Whelan demonstrates that this is true of the word “marriage” itself, even without DOMA. Orin Kerr points out that it is actually true of Randy’s favorite example, the word “property,” even though property law is generally core state law. And I have pointed out that it is true of the entire Dictionary Act, including the word “county,” see 1 U.S.C. 2, even though one might have thought that defining subdivisions like counties was the single most fundamental state function.

I would just add that this is also true of literally hundreds of other words throughout the U.S. Code. Many, perhaps most, of the words in the U.S. Code also appear in some state statute. It is utterly unsurprising to find that many of these words mean different things in different statute books.

To be sure, Congress can, if it wishes, piggyback on state definitions. But the Court has never suggested that Congress is required to do so. In fact, the presumption traditionally runs the other way. See Jerome v. United States, 318 U.S. 101, 104 (1943) (“in the absence of a plain indication to the contrary, … Congress when it enacts a statute is not making the application of the federal act dependent on state law.”); cf. McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 424 (1819) (“To impose on [Congress] the necessity of resorting to means which it cannot control, which another government may furnish or withhold, would render its course precarious, the result of its measures uncertain, and create a dependence on other governments, which might disappoint its most important designs, and is incompatible with the language of the constitution.”).

In short, no legislature has a monopoly on any particular word in the English language, and no legislature may impose its definition on any other. Congress has the power to define the terms that it uses in its own statutes. A definition is merely a cut-and-paste function. Where you see X, you should read Y. After performing the particular cut and paste, one must analyze the result to see if it runs afoul of any substantive constitutional provision. But the mere fact of a federal definition, for purposes of federal law, does not violate principles of federalism. See Federal Rules of Statutory Interpretation.

Ilya Shapiro is probably right to predict that the “federalism argument” will be rejected by at least eight Justices. But I would go further. Justice Kennedy has strong and sure federalism instincts. When he considers the radical and illogical implications of this particular “federalism argument,” which actually turns federalism on its head, I believe he will not stand alone on this ground.

AMENDMENT: Dale and Ilya Shapiro have pointed out that I overstated Ilya’s prediction. He does not necessarily predict that eight Justices will affirmatively “reject” the federalism argument; after all, some of those eight may not need to reach this argument, and so may say nothing about it. Rather, to be precise, Ilya predicts that eight Justices will not endorse the federalism argument. I predict that the number is nine.

Whatever the merits of the federalism concerns I and others have raised about the Defense of Marriage Act (DOMA), the issue is garnering a fair amount of attention.  Here’s a brief round-up of some recent commentary:

• Mike Dorf, “The Federalism Argument that Should Have Been Made Against DOMA” - arguing that DOMA’s opponents missed an opportunity at oral argument to appeal to Justice Kennedy with an embrace of federalism concerns;
• Jeffrey Rosen, “Flip-Flopping Federalists” - suggesting a tension between the usual conservative impulse to limit federal power and support for DOMA;
• John Eastman, “The Faux Federalism Argument Against DOMA” – as suggested by the title, an argument that DOMA presents no real federalism concerns (which is interesting coming from someone who has argued the federal government has no power to spend money beyond the scope of its enumerated powers);
• Geoff Stone, “The Same-Sex Marriage Cases: Snatching Victory . . .” – the federalism argument “makes no sense”
• Erwin Chemerinsky, “What Might Happen?” – the federalism argument, if adopted, would produce a “radical change” in constitutional law;
• Peter Lawler, “Same-Sex Marriage, Federalism, and Judicial Restraint” – noting the DOMA litigation rasises issues of both “judicial activism” and “congressional activism.”
• Ilya Shapiro, “If Supreme Court Has Jurisdiction, DOMA Likely Falls, With a Controlling Federalism Opinion from Kennedy Alone” – a rosy prediction.

In a post below, Orin notes that many federal statutes define the meaning of the term “property,” even though property is generally defined under state law.  If this is so, Orin understandably wonders, how could there be a problem with Congress defining marriage for purposes of federal law in DOMA?

The short answer to Orin is that all of the examples he cites are fairly straightforward examples of Congress adopting definitions that do no more than help facilitate the implementation of a given federal program and all lie within the scope of federal authority. What matters is not whether Congress invokes a particular word — there is no list of “special words” immune from the feds — but what it is that Congress is actually doing. So,  the claim is not that any effort to define “property” for the purposes of a given federal law or program necessarily “undermin[es] the institution of property,” but that where Congress actually acts to “undermine the institution of property,” it cannot defend the constitutionality of such action by claiming that all it has done is adopt a simple definition.

None of us have disputed that Congress has the power to define terms where doing so is necessary and proper to carry into execution the federal government’s enumerated powers.  So, for example, our brief notes that federal immigration law contains an antifraud marriage provision at 8 U.S.C. § 1186a(b)(1)(A)(i).  Though this provision addresses marriage, it is not particularly problematic.  As we explain in our brief:

this provision limits resident-alien status to members of a “qualifying marriage,” which excludes marriages that were “entered into for the purpose of procuring an alien’s admission as an immigrant.” 8 U.S.C. § 1186a(b)(1)(A)(i). It is at least conceivable that, in particular situations, the national government could demonstrate a need (apart from desiring to encourage a particular definition of marriage) to exclude same-sex couples. But, although the Necessary and Proper Clause might support a targeted limitation of state-conferred marital status for federal purposes, DOMA is a sawed-off shotgun. A federal definition of marriage that indiscriminately applies to more than 1100 federal statutes and programs can be “plainly adapted” to none of them.

The same is true for property. None of the examples Orin cites seem particularly problematic. Yet I think it quite clear that were Congress to enact an across-the-board definitional statute that, say, excluded automobiles, corporate stock, or other forms of property traditionally recognized under state law as “property” for all federal purposes, such a statute would be unconstitutional. Such a statute would do more than provide a handy definition for the administration of one or more federal schemes. It would represent an assault on the traditional state function of defining property, and could not be defended as “plainly adapted” to the implementation of federal law. (It might also violate other constitutional guarantees as well.)

The question, again, is not whether Congress adopted a definition of some special term, but the actual effect and intent of the legislative act in question. And while I understand the reluctance to launch open-ended inquiries into legislative pretext, in the case of DOMA, no such inquiry is necessary.  Congress was quite explicit about what it sought to do: “Defend” a traditional definition of marriage against changes adopted under state law. That Congress sought to do this through the adoption of a legal definition is of no import. Congress has no power to pursue such a goal, and there is no distinctly federal interest to invoke in Section 3′s defense.

For those interested, we’ve had some sparring over at NRO as well. I commented here. Ed Whelan has rejoinders here, here, and here. I hope to respond more directly to Ed’s points on NRO tomorrow.

In a recent op ed, Harvard Law Professor Noah Feldman argues that striking down the Defense of Marriage Act on federalism grounds – as advocated in an amicus brief I signed along with several other federalism scholars, including co-bloggers Jonathan Adler, Randy Barnett, and Dale Carpenter – would lead to “litigation chaos”:

[T]he problem with this gradual strategy envisioned by court observers and attributed to [Justice] Kennedy is that it would create anomalies leading to a nightmarish barrage of new litigation....

To understand the mess that would result if the court struck down DOMA without finding a general right to same-sex marriage, consider what would happen if the federal government recognized marriages performed in states that allow gay couples to marry while continuing to deny marital status to couples in other states.

In the first, most optimistic scenario, one or several marriage-friendly states might allow anyone from any state to get married there, creating a Las Vegas-style business in same- sex marriage. Gay couples would return to their home states with a piece of paper that should, in principle, entitle them to federal marital tax status, immigration benefits and more. But their home states would probably decline to recognize those out- of-state marriages, and deny them state-level marriage benefits.

If the Supreme Court’s decision to strike down DOMA depended on finding that states have an inherent right to define marriage in which the federal government cannot infringe, then the home states’ policy would probably be upheld. The result would be couples who are both married and unmarried for purposes of the same tax returns, mortgages and hospital visits. Each of these conflicts would be brought to the courts. State and federal courts would probably render divergent conclusions — across all 50 states and 13 federal circuits. If this isn’t legal chaos, nothing is.

If no state wanted to attract business by becoming the same-sex-marriage hub for out-of-state residents, then the anomaly would arise when legally married gay couples moved to states that didn’t recognize their unions. Presumably they would nevertheless bring their federal benefits with them — giving rise to the same legal issues just described.

To be honest, I simply don’t see the problem here. If same-sex couples get married in State A, receive federal marriage benefits, and then move to State B, which denies them state-level marriage benefits, there is no need for litigation of any kind. The couple in question will have federal marriage benefits, but no state ones (at least none from State B). Obviously, the couple might want to get state benefits from B and could potentially file lawsuits claiming that State B is engaging in unconstitutional discrimination against same-sex couples. But they could file the exact same lawsuits even if DOMA remains in place. We already have married same-sex couples that move from states that recognize same-sex marriage to those that don’t and seek to obtain benefits as a result. State and federal courts will have to address those cases regardless of whether or not DOMA gets invalidated on federalism grounds. The fact that if DOMA is struck down these couples will get federal marriage benefits does not require states to grant them state benefits.

To be sure, many states base tax status on residents’ federal tax returns. If same-sex couples could file as “married” on their federal tax returns, states that don’t recognize same-sex marriage might end up implicitly giving such people “married” tax filing status. However, if the state government doesn’t like that result, they could simply alter state tax law to forbid same-sex couples married in other states from filing as “married” on their state tax returns. This would be easy for states to do and need not give rise to any complex litigation. Many states already have tax deductions that are different from those on federal returns.

Feldman also scares readers with the following creative scenario:

Some scenarios are downright funny. Suppose I married someone of the same sex in New York and that marriage wasn’t recognized in Pennsylvania. If I then decided to marry someone of the opposite sex in Pennsylvania, the state would presumably recognize that marriage while New York recognized my previous one. And both marriages would be recognized by the federal government, which would treat me as a lawful bigamist. That would be good news for 19th century Mormons, who were denied a federal constitutional right to plural marriage — but most people today would find the conclusion truly bizarre.

As a legal scholar myself, I enjoy clever law professor hypotheticals as much as the next guy, probably even more. But this one has little real-world bite. As a practical matter, it is unlikely that any significant number of people will want to enter both a same-sex marriage and an opposite-sex marriage at the same time. If someone did try to do it, states that don’t recognize same-sex marriage could protect themselves against the possibility simply by enacting a law stating that they will not recognize a marriage contracted by a person who is still a participant in a same-sex marriage or civil union contracted in another state. Doing that would not require the state to recognize same-sex marriage generally. As for the federal government, refusing to recognize a second marriage entered into by a person who has never divorced their first spouse need not involve any federal encroachment on state governments’ powers to define marriage. After all, no state allows people to enter into multiple marriages at the same time either.

Some degree of conflict and legal uncertainty is inevitable in a federal system where different states have divergent marriage laws. But invalidating DOMA on federalism grounds would not significantly exacerbate these problems, much less lead to legal “chaos.”

UPDATE: Congress could also try to address these possible issues by basing eligibility for federal benefits on the marriage law of either the state of residency or the state where the marriage in question was initially contracted. Either approach would avoid the federalism problems created by DOMA, since neither would be a pretext for an effort to try to pressure states to define marriage in a particular way.

UPDATE #2: Ed Whelan responds to this post here:

Those crediting the federalist argument against DOMA have generally been claiming (wrongly) that the federal government lacks authority under the Constitution to create and enforce a definition of marriage different from that of the state in which a couple resides....

Somin’s claim is radically different. Under his account, the federal government would not be deferring to and applying the definition of marriage in the state in which a couple resides. Instead, for same-sex couples that lawfully marry in any state that allows same-sex marriage, the federal government would provide federal marriage benefits for those same-sex couples if they live, or if they later move to, a state that doesn’t recognize same-sex marriage. In short, under the false flag of federalism, Somin would have the federal government nationalize federal marriage benefits for same-sex marriages.

Somin’s position also gives the lie to the absurd claim that he and his co-signatories make that DOMA’s definition of marriage for purposes of federal benefits somehow intrudes on state authority over marriage. Whether or not Somin’s position is correct, Somin, by taking it, necessarily acknowledges that it is permissible for federal marriage benefits to apply to those living in a state without tracking that state’s definition of marriage. So much for the federalist argument against DOMA...

If [Somin] is right, he and the other signatories to the amicus brief of “federalism scholars” have submitted a highly deceptive brief that, in the mistaken name of federalism, would nationalize federal marriage benefits for same-sex spouses. If he is wrong, he has signed his name to an amicus brief without having giving serious attention to what his position actually means.

Whelan is wrong to claim that there is some deep contradiction in my position, much less that the brief I signed was “deceptive.” The position defended in the brief is that the federal government cannot adopt a general national definition of marriage and must instead respect state autonomy in that area. Deferring to the definition of marriage in force in the state where the marriage was contracted is deferring to the states. No one can enter a same-sex marriage unless they do so in a state where it is permitted. If a same-sex couple migrates, the receiving state can still maintain its refusal to recognize SSM. And as a practical matter, we are only talking about a small number of cases here. Gays and lesbians are only about 3% of the population and those who enter SSMs in states that permit it and then move to states that don’t are likely to be a tiny fraction of that.

Whelan’s claim that allowing couples that entered same-sex marriages to continue receiving federal marriage benefits when they move to a different state interferes with state definitions of marriage just like DOMA does is also weak. There is no meaningful comparison between an act that affects over 1100 different benefits and impacts all SSM participants everywhere and is deliberately intended to promote “traditional marriage” with the framework that I describe, which involves no such pretextual intent on the part of Congress, and in practice would have only very minor effects.

Finally, I would add that Congress can choose to condition federal marriage benefits on the law of the domicile state rather than on the law of the state where the marriage was contracted. Either approach is consistent both with my original post and with the position we took in our brief, because both avoid establishing a federal definition of marriage and both ultimately defer to state law. In my discussion above, I merely assumed the framework outlined in Noah Feldman’s op ed, and noted that it would not result in the “legal chaos” he predicted. But it’s not the only possible framework for determining eligibility for federal marriage benefits. Randy Barnett makes some related points here.

In his most recent post, Randy responds to my critique (here and here) of the DOMA “federalism brief.”

He begins by conceding: “In most every conceivable case, there is no objection to any particular definition adopted by Congress for purposes of federal law. Most definitions can easily be shown to be both necessary and proper to an enumerated power.” This is absolutely correct.

But if that’s so, I asked: “what is it that makes DOMA different? The fact that it applies to 1100 statutes…?” And Randy answers: “Yes exactly. It was indiscriminately adopted to apply to all statutes regardless of whether the definition was a necessary or proper means of executing any one of them….”

But this is true of the entire Dictionary Act, all of which was “indiscriminately adopted to apply to” hundreds of federal statutes, enacted under all different heads of federal power. Do the proponents of the “federalism brief” believe (contra, e.g. Bill Eskridge, p.92) that the entire Dictionary Act is unconstitutional? If not, why not?

The remainder of Randy’s post argues that the title of DOMA proves that it was not motivated by the execution of an enumerated power, and that it is thus a “pretext” in the McCulloch sense. See McCulloch v. Maryland (“should Congress, under the pretext of executing its powers, pass laws for the accomplishment of objects not intrusted to the [national] Government, it would become the painful duty of this tribunal, should a case requiring such a decision come before it, to say that such an act was not the law of the land.”)

This seems a very slender reed on which to hang the federalism argument. Is the title of the Act really the linchpin of the objection? Is this really the crucial distinction between DOMA and the rest of the Dictionary Act? Does anyone believe that if 1 U.S.C. 3, which defines the word “vessel,” had been entitled “The Defense of Vessels Act,” it would therefore be unconstitutional?

And even if the title of the Act were some sort of smoking gun, the fact is that the Court has more or less repudiated McCulloch’s “pretext” inquiry, in cases like United States v. Darby. See, e.g., Randy E. Barnett, Commandeering the People: Why the Individual Health Insurance Mandate Is Unconstitutional, 5 NYU J.L. & Liberty 581, 591-92 (2010) (acknowledging that, in Darby, the Court “abandoned” the “pretext” inquiry and “discard[ed] this aspect of McCulloch”). I’m not sure that the Court has struck down even a single federal statute on McCulloch “pretext” grounds for at least 75 years. If this is really the heart of the argument, it faces an uphill battle indeed.

Again, I take no position on the ultimate merits of the case; there may be substantive constitutional problems with the specific definition adopted by DOMA. But the mere fact of a federal definition — which differs from some state definitions but applies only to federal law (like the rest of the Dictionary Act) — does not violate principles of federalism.

Richard Epstein has two recent pieces discussing the Hollingsworth and Windsor cases.  One for Hoover’s Defining Ideas, the other for Ricochet.  In these pieces he notes some of his doubts about the libertarian case against DOMA and Proposition 8, but also suggests that Justice Kennedy — if he is to be consistent with his prior opinions — should not have such reservations.

I am still uncertain of how I would come down in these two cases . . . . But my equivocation on the case should not slow down Justice Anthony Kennedy. If he wants to maintain his own definition of liberty consistently, the author of the Lawrence opinion has to go the whole nine yards and come down in favor of gay marriage. . . .

The Wall Street Journal editorial page — always a steadfast friend of federalism — is nevertheless unconvinced by the “federalism brief” in the Defense of Marriage Act (DOMA) case. The editors write:

[L]iberals and some libertarians argue [that DOMA] is an offense to federalism. We disagree …. In the Constitution’s system of dual federal-state sovereignty, each coequal sovereign has the power to define marriage for its own sphere.

As I have explained, here and here, this is exactly right.

The question presented in United States v. Windsor is whether Section 3 of the Defense of Marriage Act (DOMA) violates the Fifth Amendment’s guarantee of equal protection. Why, then, are some of us talking about federalism? It’s a fair question.

In an equal protection case, courts are called upon to weigh the weight of the government’s asserted interest and evaluate whether the allegedly discriminatory policy is sufficiently related to that interest. So, for instance, under the rational basis test, a court considers whether the challenged policy is rationally related to a legitimate government interest; under intermediate scrutiny a court considers whether the challenged policy is substantially related to an important government interest; and so on. Under each test, courts must consider the nature and weight of the asserted governmental interest — and not just any interest will do. Animus or the naked desire to harm or oppress another group won’t do, nor will any other interest that is not properly pursued by the government — and that is where federalism comes in.

Because the federal government is a government of limited and enumerated powers, the range of interests it may assert in defense to an equal protection challenge is necessarily limited. It cannot assert just any governmental interest to sustain a policy, such as Section 3 of DOMA, because not all governmental interests are federal interests. Whether or not one believes the range of federal interests is limited to the objects of enumeration in the Constitution, there should be no doubt that the federal government, lacking a plenary police power, does not have recourse to the same range of potential interests as do state governments. And those interests that remain may only justify so much. Efficient administration of a federal program may suffice (at least under rational basis), but it’s hard to argue that Section 3 of DOMA is about the efficient administration of anything, particularly since in some applications it will actually make program administration more difficult and it costs the federal treasury millions. (See, e.g., here.)

Incorporating federalism concerns into Equal Protection analysis means that there are certain policy choices states may make, but that the federal government many not. As a consequence, the Supreme Court could well conclude that California’s Proposition 8 survives Equal Protection challenge, but Section 3 of DOMA does not. Should the Court decide to apply some form of heightened scrutiny, Section 3 would be in particular trouble (for reasons identified by the Second Circuit). But even under heightened scrutiny, Proposition 8 and other state laws refusing to recognize same-sex marriage might survive.

In his latest post on the federalism argument against DOMA, my co-blogger Nick Rosenkranz suggests that there can be no federalism-based constitutional objection to a statute that “defines [a word] only for purposes of federal law.” According to Nick. such a definition can only be unconstitutional if it offends some substantive constraint on government power, such as the Equal Protection Clause, and that federalism concerns have “nothing” to do with it. Randy makes several points in response to Nick’s argument, as did I in a prior post. Here I just want to focus on the claim that it is impossible for Congress to define a word, for purposes of federal law, that could exceed the scope of federal power.

For starters, we agree that Congress may define terms in order to carry into execution the federal government’s enumerated powers. Where we disagree, apparently, is the extent to which this imposes a meaningful constraint on federal lawmaking and whether it is possible for Congress to adopt statutory definition that exceeds the scope of federal power. Yet not only is this possible, the Supreme Court has said as much in scrutinizing the definitions Congress and federal agencies have adopted for statutory terms.

Let’s take one of Nick’s examples: “It shall not be lawful to tie up or anchor vessels or other craft in navigable channels in such a manner as to prevent or obstruct the passage of other vessels or craft.” 33 U.S.C. 409. As Nick notes, in 1 U.S.C. 3 Congress elsewhere defined “vessel” for the purposes of this and other federal laws. There’s no problem here, but that’s because insofar as Congress has the authority to regulate vessels in navigable waters (under the Commerce Clause) and military vessels (under those clauses conferring authority over the military), it can define what constitutes a vessel for such purposes. Such an action is necessary and proper for carrying into execution enumerated federal powers and neither threatens state authority nor aggrandizes federal power. So far so good.

Yet suppose instead of defining “vessel” Congress sought to define “channel” as used in this and other statutes. And suppose Congress defined “channel” to include, not just those waterways we may all recognize as “channels,” but also all navigable-in-fact waterways, as well as all interstate and intrastate waters including (but not limited to) intrastate lakes, rivers, streams (including intermittent streams), mudflats, sandflats, wetlands, sloughs, prairie potholes, wet meadows, playa lakes, natural ponds, drainage ditches, and ephemeral streams. Such a definition does more than provide semantic meaning to a statutory term. It would also operate to extend the scope of federal power. And insofar as federal regulatory authority does not extend to all such waters – as the Supreme Court has suggested twice in the past dozen years – it would be no defense to claim that all Congress sought to do was provide a handy definition for the purposes of federal law. This is because the practical effects of this definition would be expand the scope of federal power so as to supplant state authority in an area of traditional state concern. [Note also that in the cases concerning waters, Congress at least had a long tradition of asserting federal authority over some U.S. waters, and a clear textual hook for the assertion of federal power under the Commerce Clause. With DOMA, however, Congress has neither.]

Once we have established that Congress lacks some general, all-purposes “definitions” power that is immune from judicial review, we can focus on the real question: Whether the enactment of a particular definition is necessary and proper to carry into execution the federal government’s powers. Where Congress adopts definitions to facilitate operation of constitutionally authorized federal programs or to attain enumerated purposes, there may be no problem. Where, however, Congress enacts a statutory definition so as to leverage preexisting authority so as to aggrandize federal power and intrude on an area of traditional state concern – such as family law – greater care is required.

There has been a lot of commentary about the federalism argument against the Defense of Marriage Act (DOMA). Several of my co-conspirators signed on to the “federalism brief” in the DOMA case. I expressed serious doubts about the argument a few weeks ago (as did Sasha, and Ed Whelan over at National Review Online), and Jonathan Adler and Ernie Young responded. In deference to my co-conspirators, I thought I would leave the matter there. But now that George Will and Michael McConnell have taken to the editorial pages in support of the federalism argument (as Dale and Jonathan note), I feel obliged to explain in greater detail why I think this argument is unsound.

Once again, Section 3 of DOMA provides: “In determining the meaning of any Act of Congress, or of any ruling, regulation, or interpretation of the various administrative bureaus and agencies of the United States, the word ‘marriage’ means only a legal union between one man and one woman as husband and wife, and the word ‘spouse’ refers only to a person of the opposite sex who is a husband or a wife.” The key point here is that this provision defines the word “marriage” only for purposes of federal law.

Indeed, Ernie Young, primary author of the federalism brief, concedes that “Congress can, of course, define terms where this is ‘necessary and proper’ for ‘carrying into execution’ its enumerated powers.” But he insists that DOMA’s definition of “marriage” is not necessary and proper. I confess that I do not understand this argument.

Imagine Congress has enacted only two statutes total. The first is an exercise of the Commerce Clause power. It provides: “It shall not be lawful to tie up or anchor vessels or other craft in navigable channels in such a manner as to prevent or obstruct the passage of other vessels or craft.” 33 U.S.C. 409. The second is an exercise of the power to “make rules for the government and regulation of the land and naval forces.” It provides: “The Secretary of the Navy shall designate boards of naval officers to examine naval vessels.” 10 U.S.C. 7304(a).

I don’t think anyone doubts that Congress could add a definition of the word “vessel” to statute one, or statute two, or both. Surely, therefore, Congress can instead enact a general definitional provision, defining the word “vessel” once for purposes of both statutes. In fact, Congress has done exactly that, at 1 U.S.C. 3 (just a few sections before DOMA), and the Court has never suggested that it is unconstitutional. I imagine that all the signatories of the “federalism brief” are with me so far. They have not, thus far, expressly argued that the entire Dictionary Act is unconstitutional.

So what is it that makes DOMA different? The fact that it applies to 1100 statutes, rather than just two? The Dictionary Act, which defines “vessel” at 1 U.S.C. 3, also applies to many, many federal statutes. The fact that the many federal statutes that use the word “marriage” were originally enacted under assorted, different heads of congressional power? That’s just as true of “vessel.” (Compare 33 U.S.C. 409 with 10 U.S.C. 7304(a), two paragraphs above.) The fact that states have also chosen to use the word “marriage” in many of their statutes? “Vessel” appears in lots of state statutes too. The fact that some states have chosen to define “marriage” differently, for purposes of state law, than Congress has for purposes of federal law? Again: this is also true of “vessel.” (And even if it weren’t true of “vessel” today, could it possibly matter if California tomorrow defined “vessel,” for purposes of state law, to include sports cars? Would California thus somehow render the federal Dictionary Act unconstitutional?) The fact that state changes in the traditional definition of “marriage” are recent, and post-date DOMA? Surely, if anything, that fact cuts the other way. Cf. U.S. v. Raich, 549 U.S. at 29 n.38 (“California’s decision (made 34 years after the Controlled Substances Act] was enacted) to [legalize medical marijuana] cannot retroactively divest Congress of its authority under the Commerce Clause.”); The Subjects of the Constitution, 62 Stan. L. Rev. 1209, 1279, 1283-84 (2010).

The fact that marriage is special, and of particular state concern? Well, ok. But that is only to say that the word “marriage,” as used in state law, connotes a special relationship. It hardly follows that Congress is either required or forbidden to use the same word, or to have it denote the same idea. Congress could forbid selling a kilo of “marriage” in interstate commerce and then define “marriage,” for purposes of this statute, to mean cocaine; this would be an odd use of language, but it would hardly exceed Congress’s Commerce Clause power or violate the Tenth Amendment.

Indeed, to this point, consider another provision of the Dictionary Act, 1 U.S.C. 2, defining the word “county” for purposes of federal law. One might have thought that defining its own subdivisions, like counties, was perhaps the single most fundamental reserved power of a state – more fundamental even than domestic relations. Cf. Coyle v. Smith (state has power to choose its capital city). Yet no one has ever suggested that Congress cannot define the word “county” for purposes of federal law.

In short, no legislature has a monopoly on any particular word in the English language, and no legislature may impose its definition on any other. If DOMA is unconstitutional, it is because this particular federal definition of “marriage” offends some substantive constitutional provision. But that has nothing to do with federalism.

One final note: Jonathan Adler reminds us that “federalism should be respected even when it cuts against conservative preferences,” and Randy makes the same point. About this, I emphatically agree. But the implication, it seems, is that Sasha and Ed Whelan and I are, perhaps, fair-weather federalists, abandoning our principles because they do not suit our preferred policy result in the DOMA case. For the record, I yield to no one in my steadfast defense of federalism. Moreover, I have no particular fondness for DOMA, and I have taken no position on the ultimate merits of the case. I have critiqued “the federalism brief” only because, with all due respect, I’m afraid that it is unsound.

For a more thorough and scholarly treatment of this topic, see Federal Rules of Statutory Interpretation, 115 Harv. L. Rev. 2085, 2102-25 (2002).

On Thursday, Stanford law professor and former federal appellate judge Michael McConnell suggested the Supreme Court should avoid ruling directly on the constitutionality of same-sex marriage by denying Proposition 8′s supporters standing to defend the initiative’s constitutionality and holding that the Defense of Marriage Act exceeds the scope of federal power. NRO’s Ed Whelan, who has previously criticized the federalism arguments against DOMA, responded to McConnell’s arguments here and here. McConnell sent Whelan a reply, which Whelan has posted with his response here.

In Friday’s WSJ, Stanford law professor Michael McConnell, a former judge on the U.S. Court of Appeals for the Tenth Circuit, argues that the Supreme Court should hold that Proposition 8 supporters lack standing to defend the ballot initiative in federal court and that the federal government lacks the power to define marriage.

the court need not base its decision in Windsor on the merits of the same-sex marriage question. The leading argument against DOMA all along has been that the federal government lacks authority under the Constitution to create and enforce a definition of marriage different from that of the state in which a couple resides. It is hard to think of an issue more clearly reserved to state law under constitutional tradition than the definition of marriage.

The court has held that “regulation of domestic relations” has “long been regarded as a virtually exclusive province of the States” (Sosna v. Iowa, 1975). In the past, the court has recognized a “domestic relations exception” to federal judicial power. Although the legal question is close, the court could take the same path in Windsor—holding that DOMA improperly intrudes on the reserved powers of the states.

If the court dismisses the Proposition 8 case on standing grounds and strikes DOMA down on federalism grounds, the combined effect would be to reaffirm America’s democratic, decentralized decision-making process without imposing an answer—one way or the other—to the same-sex marriage question.

By taking such a path, the court would be spared from imposing a single nationwide definition of marriage as a matter of constitutional law, and from having to rule, for all time, that there is or is not a constitutional right to same-sex marriage—a momentous step that some justices might be reluctant to take. It would leave the issue to the states, at least for the time being. This course might appeal to centrist justices like Anthony Kennedy, John Roberts and Stephen Breyer—and perhaps could even command a unanimous court, which would have a welcome calming influence on the nation’s culture wars.

I’m skeptical either judgment could command a unanimous court, and am not yet convinced Prop. 8′s supporters lack standing, but certainly agree this would be a welcome approach for the Court to take (and certainly agree that the “leading argument” against DOMA is that it exceeds the scope of federal power).

I’m also one of those who are skeptical of the federalism brief against DOMA. But I have a few questions, based on perusing the recent posts here and elsewhere by Dale, Randy, Jonathan, Nick, and my former boss Ed. (I read the brief itself a while back, but declined to sign; forgive me if the answers to my questions are right there in the brief.)

1. Ed distinguishes three types of statutory schemes: (a) a statute attaching consequences to marriage, with a definitional section defining “marriage”, (b) a statute where the word “marriage” is simply replaced with its definition, and (c) a statute without a definition of marriage, but where “marriage” is defined in a separate statute. Would the federalism theory be fine with (b) but not with (a) and (c)? Or, based on the idea that Congress has no power to care about what’s a good marriage, would the theory would also invalidate (b)? What’s the dealio?

2. If the problem is that Congress enacted DOMA for the purpose of furthering traditional marriage, does that mean I have to buy into purposivism or intentionalism to buy this theory?

3. What about all the statutes granting benefits to a spouse derivatively of the eligibility of a primary person, like Social Security (for purposes of survivorship), the immigration laws, etc.? The statute could have been written to apply to the actual eligible person: no Social Security benefits for widows, no special immigration treatment for the wife of an eligible person, etc. And yet they did write the statute to give special treatment to spouses that isn’t available for boyfriends/girlfriends, business partners, best friends, chess partners, etc. What was Congress’s power to do so, thus privileging the marital relationship over other kinds of relationships (even long-standing non-marital romantic relationships)? Does Congress have a legitimate interest there?