Let’s, as the president would say, review the bidding on this. I favor a constitutional amendment that would block both the federal and the state courts from instituting same-sex marriage, civil unions, or from leaving marriage eligibility alone but extending some of its benefits to the unmarried. Such an amendment, I believe, would deserve the support not only of opponents of same-sex marriage but even of some supporters of it: supporters who believe that same-sex marriage would have greater acceptance in the long run if it were achieved legislatively rather than judicially, or who believe that it is wrong in principle for the courts to impose same-sex marriage however good an idea it might be, or who believe both things.
The great defect of this amendment is that, as far as I know, nobody but me has proposed it. What has been proposed is a constitutional amendment reading as follows: “Marriage in the United States shall consist only of the union of a man and a woman. Neither this Constitution or the constitution of any State, nor state or federal law, shall be construed to require that marital status or the legal incidents thereof be conferred upon unmarried couples or groups.” NR has editorialized in favor of this amendment (while endorsing my proposed amendment as a kind of fallback or compromise). I myself have not endorsed it, preferring to critique what I regard as bad arguments for it or, more often, against it.
But most recently, what I’ve been debating is what the amendment means–and, even more narrowly, what the second sentence of it means. This is not to say that there is no controversy about the meaning of the first sentence. Most supporters of the amendment believe that the first sentence reserves the word “marriage” for opposite-sex couples. A few supporters, however, believe that this sentence would also block certain kinds of civil-union laws. Professors Robert P. George of Princeton University and Gerard Bradley of Notre Dame Law School fall into this camp. In their view, the only way to rule out the interpretation that prohibits some forms of civil union would be to recast the first sentence as follows: “In the laws of the United States and the states, the word ‘marriage’ shall refer to the union of a man and a woman.”
But Bradley and George agree with every other supporter of the amendment I know in thinking that the second sentence of the amendment, taken by itself, allows legislatures to create civil unions and other benefits packages for same-sex couples. It blocks only the judicial creation of these civil unions and benefits. So, to sum up what we’ve got so far: Almost all supporters of the amendment believe it allows the legislative creation of civil unions; all supporters believe that the second sentence of the amendment does (or would if the first sentence were rewritten).
This is precisely the assertion that opponents of the amendment have chosen to contest. In recent weeks, bloggers Andrew Sullivan, Eugene Volokh, and Jacob Levy have all argued, against me, that the “legal incidents” sentence blocks even legislatures from creating civil unions. Evan Wolfson, a leading activist for same-sex marriage, has adopted this argument. So has law professor Andrew Koppelman.
There are some differences among the arguments the bloggers make. Volokh merely argues that my interpretation of the second sentence is wrong: As written, it blocks legislated civil unions. Levy argues that the precise wording does not matter: It is impossible in principle to devise language for a constitutional amendment that would allow legislative, but not judicial, creation of civil unions. Sullivan buys both men’s arguments, saying the former has “demolished” my own and the latter’s are “devastating.” He adds that social conservatives generally know that Volokh’s interpretation is true and are lying about the amendment’s meaning.
Sullivan generously exempts me from the charge of deliberate deception. But he does not provide much by way of evidence for its accuracy with regard to others. As near as I can tell, his main argument for the theory is that there would be no reason for social conservatives to want an amendment that allowed legislatures to create civil unions. If they really thought it meant what they say it means, in other words, they wouldn’t be interested. I’m afraid I don’t follow this at all. Most social conservatives, especially the activists, are much more alarmed at the threat of judicially imposed civil unions than that of legislatively enacted ones. They figure they can win fights in the legislatures more easily than the courts, and plenty of them will take an amendment that blocks the judicial threat even if it does not also restrict legislatures. (Some of them, on the other hand, refuse to support the amendment because it gives legislatures too much latitude.)
Unlike Sullivan, I take the disagreement on the meaning of the second sentence to be in good faith. I assume, for example, that Sullivan, Volokh, and Levy are not simply adopting their interpretation in a duplicitous effort to make the amendment look extreme, and that if it were enacted in its present form they would continue to claim that it bars legislatively enacted civil unions. If the current language passes, no doubt they will, in the future, urge Wolfson to file legal briefs to that effect.
Here’s why Volokh thinks that the amendment’s legal-incidents language means no civil unions, period. He envisions a scenario in which a legislature provides same-sex couples with some benefit or set of benefits that used to be reserved for married couples. But the executive branch refuses to provide the benefit, because it cannot, under the amendment, provide a legal incident of marriage to unmarried couples. And the courts can’t help, either: They can’t “construe” any statute to confer an incident of marriage on an unmarried couple.
In response, I have argued that the “legal incidents of marriage” are, essentually, whatever state legislatures say they are. They are not some set of historical benefits and duties that have tended to come with marriage. If the Indiana legislature, today, does not treat a particular benefit as marital, that benefit is not an incident of marriage in Indiana. That is true even if most states have historically offered that benefit as a marital incident and even if books compiling typical incidents of marriage list it as one.
In my view, the Indiana legislature could take some benefit formerly tied to marriage–say, joint adoption rights–and provide it both to married couples and gay couples in civil unions. In that circumstance, the legislature would have simultaneously taken two conceptually distinct actions: abolishing an incident of marriage and extending a benefit.
Volokh inclines more toward the view that an incident of marriage is any benefit that is traditionally associated with marriage. He offers examples of judicial decisions that have spoken of marital “incidents” in that way. None of these decisions, however, seems to be a judicial attempt to figure out the meaning of a statutory reference to “the legal incidents of marriage”–still less a constitutional reference to them.
A reasonable court trying to interpret the marriage amendment would, in my view, pay attention to the word “require.” What is that doing in the amendment? The text could simply declare that no statute or constitutional provision shall be “construed to confer” a benefit to same-sex couples. What’s the point of saying that a law can’t be “construed to require” that a benefit be conferred? It might be thought to suggest that a benefit can be conferred on same-sex couples but does not have to be. It is an option; a legislative choice.
Volokh goes on to say that if my interpretation is correct, a court would be able to create civil unions anyway. It could simply abolish incidents of marriage and then extend them, just as I say legislatures could do under the amendment. I think Volokh is right about this, to a limited extent. If a court is determined to read the amendment in complete abstraction from its clear intent, then it can indeed work its will.
But that scenario at least has some plausibility. It seems highly unlikely that, once the marriage amendment passes, the federal courts will be straining to interpret it in the way most hostile to same-sex couples.
Levy accepts Volokh’s points about the meaning of the amendment, but goes on to suggest that no possible amendment language could allow legislatures, but not courts, to create civil unions. He writes,
If you try to prevent the judiciary from deeming that something like the equal protection clause creates a civil union–which is what supporters of the FMA [Federal Marriage Amendment] hope to do–then you also prevent the judiciary from deeming that the “Civil Unions Act of 2004″ creates a civil union. The more moderate proponents of the FMA suggest that they’re open to the latter but not the former. But there’s simply no way to allow a judge to find legal justification for civil unions in an act of the legislature while forbidding him to find legal justification for them in the text of the constitution. Doing so would mean prohibiting any judicial interpretation of the more sweeping constitutional clauses, like equal protection, which would be the height of the kind of judicial activism conservatives frequently complain about. The judicial authority to construe one source of law can’t be parceled off so neatly from the authority to construe others.
Try as I might, I can make no sense of this. You would not have to prohibit “any judicial interpretation” of the equal protection clause; you would be prohibiting an interpretation of it that produced civil unions. It might be a bad thing to pass an amendment that abolished judicial review altogether, but that’s not the same thing as saying it’s impossible. Passing such an amendment would not be judicial activism; nor would it be activism for a judge to abide by it.
Levy goes on to suggest that there is something particularly unsettling about the federal imposition of a rule of construction on state judiciaries. Again, I don’t see his point. It may be true that nowhere else in the Constitution are state courts told how to interpret state laws. But they are effectively prohibited from giving effect to state laws that allow, say, slavery. I don’t see what the force of the distinction is supposed to be.
Or let’s look at Bush v. Gore. There you had a Florida supreme court that had “interpreted” state election laws. The Supreme Court found, rightly, that the court had trespassed on a power (the establishment of procedures by which electors are chosen) that the Constitution reserves to state legislatures. The state court’s interpretation of state laws implicated the federal Constitution. That’s what would happen, in some cases, if the marriage amendment becomes part of the Constitution.
As for the suggestion that no possible language will work, let me offer a suggestion: Strike the words “state or” from the second sentence of the amendment. That is to say, make it possible for a state law to be construed to require the conferral of benefits on same-sex couples. But continue to block the construal of a state constitution that way. That deletion should make Volokh’s scenario go away. And amendment supporters would not be giving up very much. What they are most concerned about is the idea that a court will take one of the “majestic generalities” typically found in constitutions and give it a specific meaning they do not believe it can bear. Statutes, by their nature, are less susceptible to this kind of thing. And if a state court took interpretive liberties with a statute, it would be much easier for the state legislature to correct the problem–it would simply need to pass a new statute.
I end with a question for Volokh. Does he agree with Levy that it is impossible to devise amendment language that would block the federal and state courts from creating civil unions while allowing state legislatures to create them? Never mind whether he thinks that such an amendment is a good idea. Can he come up with language that, to his mind, effects the goals that most amendment supporters say they have for it and none of the ones they have disavowed? If the language can be improved, it could make for a more edifying debate.