Is Same-Sex Marriage Already Constitutional?
(Author’s note: This essay was written some ten years ago and may not reflect my current views in every detail although it does still represent the general tenor of my thought.)
This essay is an argument that same-sex marriage is already Constitutional under Romer v. Evans, 517 U.S. 620 (1996), and Lawrence v. Texas, 000 U.S. 02-102 (2003). The current controversy over same-sex marriage is at bottom a conflict between two views of marriage. Opponents of same-sex marriage tend to think that legal recognition of a marriage is somehow a way of bestowing society’s blessing or approval on a marital union. Proponents of same-sex marriage tend to think that getting married is simply the exercise of a personal right that doesn’t require anyone else’s approval.
The first of these views appears to be based on a confusion between religious marriages and civil marriages, or at least on a sense that the two should be somehow the same. And it is surely true that most religious marriage rites do involve some sort of endorsement or blessing of a marital union, a communal or institutional ‘stamp of approval’ that requires at least the tacit permission of a community or its authorities.
But civil marriage involves no such thing. There is a long line of Supreme Court cases holding that there is a right to marry (and we’ll mention a few of them in what follows). This right to marry, according to the Court, is a fundamental liberty right, and its exercise — though, like the exercise of all liberty rights, subject to some regulation — neither requires nor involves societal approval of one’s choice of spouse. (And of course the sort of marriage at issue here is civil marriage, i.e,, marriage under the jurisdiction of the Constitution or of the states that are parties to it. Private individuals’ and organizations’ rights of free association and religious expression and so forth are unaffected.)
Indeed, it’s the very nature of a right that no one’s ‘permission’ is required for its exercise. It isn’t very plausible to read same-sex civil marriage as somehow conferring societal ‘approval’ when opposite-sex civil marriage doesn’t. Insisting on the recognition of a fundamental right and the liberty to exercise it is hardly a demand for social approval. (Of course this argument cuts both ways; even with Constitutional approval of same-sex marriages, people are entirely free to disapprove of them and to refuse to recognize them for religious and other strictly private — here meaning ‘non-government’ — purposes.)
So far, though, we don’t have explicit word from the Supreme Court that same-sex marriage definitely does fall within the scope of the fundamental right to marry. Despite some language in Romer and Lawrence suggesting a favorable attitude toward same-sex marriage, the Court has never ruled on the subject directly. However, I think those two cases provide the foundation for a strong argument that same-sex marriage is already Constitutionally protected.
The argument is in two parts, and understanding each part requires a little familiarity with current Fourteenth Amendment jurisprudence. I’ll try to provide the requisite background as we go along.
The reason there are two parts is that there are really two distinguishable aspects (perhaps even different meanings) of the ‘right to marry’, and they’re generally not clearly distinguished in arguments on this subject. The first is the liberty right to enter into a long-term, committed, intimate relationship with another consenting adult (and to express or solidify that commitment through a cermony or ritual); the second is the positive right to legal recognition of such a relationship. (Lawrencedistinguishes these two rights carefully, although without explicitly naming them: Justice Anthony Kennedy’s majority opinion deliberately refrains from addressing the question whether same-sex relationships are entitled to any kind or degree of legal recognition, while clearly protecting the right to enter into such relationships.)
You’ll understand why I distinguish these two rights so carefully when you see the two-part argument. Here it is:
First: Under Lawrence, the Fourteenth Amendment’s Due Process Clause protects the liberty right of same-sex couples to enter into such relationships; any state laws criminalizing or otherwise prohibiting such relationships (if there are any) are unconstitutional. Moreover, under Lawrence, the Fourteenth Amendment also prohibits any state laws that burden the exercise of this right — so any state laws (or state constitutional amendments) forbidding the recognition of such relationships are also unconstitutional. It’s also a Due Process violation to deny same-sex couples certain other legal rights (e.g. the right not to testify against one’s spouse and the right to prevent one’s spouse from testifying about private communications). So under the Due Process Clause alone, same-sex couples are entitled to some degree of legal recognition.
It’s possible, though, that all the Due Process problems could be handled by recognizing ‘domestic partnerships’ or ‘civil unions’ that were not fully equivalent to civil marriages in every way. That’s why we can’t stop with a pure Due Process analysis; the liberty right to enter into a same-sex committed relationship without legal impediment doesn’t necessarily entail the positive right to ‘equal’ legal recognition of that relationship.
Second: Under Romer, it’s not enough to grant same-sex unions legal recognition that falls short of equivalence to civil marriage. The Equal Protection Clause forbids such a distinction; every jurisdiction is Constitutionally bound to give same-sex civil marriages exactly the same kind and degree of recognition it gives opposite-sex civil marriages.
As you can see, to each right there corresponds a somewhat different meaning of ‘constitutionality’. By the ‘constitutionality’ of the right to enter into committed same-sex relationships, we mean that the Constitution forbids the enactment of laws that prohibit people from entering into and maintaining such relationships or that burden their exercise of their right to do so. By the ‘constitutionality’ of the right to legal recognition of such relationships, we mean that the Constitution requires such recognition.
(By the way, this argument doesn’t show that the ‘best’ way to recognize this right is to have the Supreme Court actually rule on it. In fact, my strong preference would be that the voting public in each state become persuaded that same-sex unions deserve civil equality with opposite-sex marriages, and vote accordingly. My second choice would be that state courts rule on the question. A SCOTUS decision would be, primarily for political reasons, my distant third choice.)
Now let’s set out each prong of the argument more fully.
The Due Process argument
Essentially, Lawrence holds that under the Fourteenth Amendment’s Due Process Clause, in order for a state law restricting private conduct to have a ‘rational basis’, it has to further some interest beyond the mere promotion or protection of public morality. The Court used this principle to overturn Bowers v. Hardwick, 478 U.S. 186 (1986), and strike down state anti-‘sodomy’ statutes as unconstitutional.
Lawrence also has a less commonly appreciated effect on Fourteenth Amendment jurisprudence. By way of background, here’s a somewhat simplified summary of pre-Lawrence Fourteenth Amendment Due Process analysis.
Suppose a state law affects the exercise of a libery right. Under what had been the standard Fourteenth Amendment Due Process analysis, a court would first look at the affected right and determine whether it was ‘fundamental to scheme of ordered liberty’. If it was, then a law curtailing the exercise of that right would be subjected to ‘strict scrutiny’ in which the government would be required to show that the law was narrowly tailored to serve a compelling government interest. Otherwise, the law would need only to have a ‘rational basis’ in order to pass Constitutional muster — and the burden was on the challenger of the law to show that it lacked such a basis. This latter task is all but impossible; pretty much any law would survive such a test.
Now, though, for any law touching on the exercise of rights associated with intimate relationships, Lawrence puts teeth into the ‘rational basis’ test: such a law has to have some purpose beyond the promotion of morality. Moreover, Lawrence apparently places the initial burden of proof on the government to show that its law doeshave such a purpose. That means that, on a Lawrence-based analysis, a law touching on certain sorts of private conduct may be struck down without ever reaching the question whether the right in question is ‘fundamental’. The usual Fourteenth Amendment analysis doesn’t even start unless the government can meet its initial burden, because there’s no need for it; obviously, if a law can’t pass a rational-basis test, it wouldn’t survive strict scrutiny either.
The logic of Lawrence also indicates that laws burdening the exercise of privacy-liberty rights are subjected to the same test. Lawrence did not, as some ‘conservatives’ like to pretend, strike down anti-sodomy laws on the basis of a ‘right to sodomy’; it struck them down on the basis of a right to enter into certain sorts of personal relationship. Indeed, the majority could hardly have been clearer that the right ultimately at issue was not (as Bowers had suggested) merely the right to engage in certain sorts of sex. The Court’s overarching point is that anti-sodomy laws are unconstitutional because they burden the exercise of the broader right to enter into and maintain intimate relationships with persons of one’s own choice — a right that includes, but of course is not limited to, same-sex relationships.
This understanding of Lawrence receives support from the famous trio of ‘contraception’ cases: Griswold v. Connecticut, 381 U.S. 479 (1965), Eisenstadt v. Baird, 405 U.S. 438 (1972), and Carey v. Population Services International, 431 U.S. 678. In these cases the issue of distribution of contraceptives was considered in light of a ‘right to privacy’ that protected their use. It is possible to argue that at least part of the logic of these cases is that since the right to use contraceptives is protected, limitations on their distribution may unconstitutionally burden the exercise of that right.
The argument is made much more explicitly in U.S. v. Extreme Associates, Inc. et al. [that link is to a PDF document], in which federal judge Gary Lancaster declared certain federal obscenity-trafficking statutes unconstitutional because burdensome to the Constitutionally protected right to view ‘obscene’ materials in the privacy of one’s own home. Whether or not Extreme Associates survives appeal, it is still relevant here for the light it sheds on the underlying logic of Griswold and Lawrence. (And even if it is overturned on other grounds, I still think its analysis of Lawrence is correct in the respect discussed here.)
What this means for same-sex marriage is that any laws burdening the right to enter into committed relationships (including same-sex relationships) are unconstitutional from scratch unless they survive Lawrence‘s rational-basis test. That is, Lawrence stands not only for the proposition that people have the right to enter freely into (and remain in) such relationships, but also for the proposition that the state has no legitimate interest in making it more difficult for people to enter into them (or remain in them).
This conclusion alone means that state laws and constitutional amendments prohibiting the recognition of such relationships are unconstitutional — except, that is, in the unlikely event that they have some purpose beyond promoting a single societal ‘moral’ view of marriage.
(And even then, such laws would still be subject to the ‘old’ Fourteenth Amendment analysis, in which case it would still be possible for the SCOTUS to rule that the rights at issue here are ‘fundamental’. There is a long line of cases holding that the right to marry is a ‘fundamental right’, and although these cases don’t reach the specific issue of same-sex marriage, it wouldn’t be a long leap to extend them to cover such marriage. One of the most important of these cases is Loving v. Virginia, 388 U.S. 1 (1967), in which the Court struck down Virginia statutes forbidding interracial marriage on both Due Process grounds and Equal Protection grounds. We’ll revisit this case shortly, in the Equal Protection portion of our analysis.)
It also means that same-sex couples deserve to have available to them, as a matter of due process, at least some of the legal rights and other protections available to opposite-sex married couples. For example, the spousal testimonial privileges have as their purpose the protection of marital privacy. Under Lawrence, same-sex couples enjoy a similar right to privacy, and the unavailability of such protections surely places a burden on their ability to maintain privacy in their relationships. On the logic of Lawrence, then, such couples are being deprived of liberty without due process. (And the two spousal privileges are just one of a long list of legal rights; seethis page for quite a few more. Scroll down to the part headed ‘Other Reasons why Same-sex Marriage Rights are Important’.)
So under the Due Process Clause alone, same-sex relationships are entitled to some legal recognition parallel to that of opposite-sex marriage. (As we’ll see, this point is crucial to our Equal Protection analysis.)
Lawrence takes us quite a bit of the way, then, to a Constitutional mandate for the recognition of same-sex committed relationships — both the liberty right to enter into such relationships, and positive rights to a degree of legal recognition in at least some of the same respects as opposite-sex marriages. But to get all the way to mandatory recognition of same-sex civil marriage on equal footing with opposite-sex marriage, we need to rely on the Equal Protection Clause.
The Equal Protection argument
Loving, which was decided in part under the Equal Protection Clause, provides a partial reply to one standard ‘conservative’ argument against recognizing same-sex marriages. Some ‘conservatives’ like to point out that legally recognizing only opposite-sex marriages is no violation of the Equal Protection Clause because the same standard is applied to everyone: men can marry only women, and women can marry only men. But Loving rejected an analogous argument as to race, and held that the Equal Protection Clause requires an analysis of whether the classifications drawn by any statute are themselves discriminatory. If the classifications themselves fail to pass Fourteenth Amendment muster, the law is struck down no matter how even-handedly its consequences are applied.
However, prior to Romer, this reply wouldn’t have provided much traction in the present argument; in order to make our case, we would still have had to show that sexual orientation was a ‘suspect classification’. (Perhaps it should be — but the Court has never held that it is.) But following Romer, we may not need to show that any more. Here’s a little background on the case.
Some jurisdictions in Colorado were passing ordinances forbidding discrimination based on sexual orientation. In order to get rid of these ordinances, the people of Colorado passed a (state) constitutional amendment prohibiting them — forbidding, that is, ordinances that protected gays, lesbians, and bisexuals from discrimination. (Romer, the governor, strongly opposed the amendment.) The Supreme Court struck down the amendment on a rational-basis analysis — similar to the one it later employed in Lawrence, but here applied to the Equal Protection Clause.
The Equal Protection Clause essentially forbids state laws that are unduly discriminatory. Here’s a short and simplified summary of how the pre-Romer analysis worked.
Suppose a state law seems to be discriminatory in its intent or effect. The first step in a pre-Romer Fourteenth Amendment Equal Protection analysis is to determine whether the discrimination is based on, or otherwise affects, a ‘suspect classification’ (like race). If it does, then it is subjected to strict scrutiny; if not, it needs only to survive a rational-basis test. Alternatively, a law may affect a ‘fundamental interest’ (like voting), in which case strict scrutiny applies. (Generally, the Court doesn’t recognize very many such ‘fundamental interests’.)
Romer skipped most of this analysis and held simply that the Colorado amendment did not survive even a rational-basis test. Justice Kennedy, writing for the majority, held that the object of the law was merely to make homosexuals ‘unequal to everyone else’ and declared that there is no legitimate government interest in setting out to harm a politically unpopular group out of simple animosity. Indeed, Kennedy went further and declared that the Colorado amendment was ‘a classification of persons undertaken for its own sake, something [the Equal Protection Clause] does not permit’.
Here again, the Court has put teeth into a previously feckless rational-basis test. And the Court didn’t even reach the question whether sexual orientation is a ‘suspect classification’; it declared broadly that any state law based on any ‘classification . . . for its own sake’ simply didn’t make it out of the gate. (Strictly speaking, to be unconstitutional such a law may have to involve some sort of difference in treatment for that classification; a law that said ‘Black people have the right of free speech, and so do non-black people’ would be silly and devoid of any practical significance, but it might not be a Fourteenth Amendment violation. On the other hand, would it? Some of the language in Romer suggests that invoking such a classification at all might be enough to violate the Equal Protection Clause.)
As in Lawrence, if a law cannot survive an initial rational-basis test, the rest of the usual analysis is unnecessary; a law with no rational basis would not survive strict scrutiny either. However, it’s less clear whether Romer transfers the initial burden to the proponent of the law, so in what follows we’ll assume that the challenger has toshow that a challenged law is based on a ‘classification . . . for its own sake’.
But counterarguments are still available. For one thing, it would still be possible to argue in reply that a one-man-one-woman definition of ‘marriage’ doesn’t involve any sexual-orientation-based classification at all: strictly speaking, laws recognizing only opposite-sex unions as marriages make no reference to the sexual orientations of the spouses. (Gay men are allowed to marry women just as surely as are straight men; the legal definition of marriage doesn’t distinguish between them.) This argument obviously misses the point — but unless that point can be characterized in Equal Protection terms, we can’t get all the way to the conclusion that same-sex marriages merit equal recognition with opposite-sex ones.
And even if such a definition of marriage is found to depend implicitly on a classification by sexual orientation, it’s still possible for opponents of same-sex marriage to argue that this classification isn’t just ‘for its own sake’. They could even rely on Supreme Court precedent (Reynolds, which we discuss below) to argue that prohibition of other forms of marriage really does promote a legitimate government interest.
That’s why it’s crucial that we completed our Due Process analysis first. We’re not starting from scratch; our Due Process analysis has already shown that each jurisdiction is Constitutionally required to provide some kind and degree of recognition to same-sex unions. All that is still at issue is whether such recognition must be fully equivalent to civil marriage.
We can, that is, assume the existence of laws that recognize same-sex unions in some way — and of course the Equal Protection Clause applies to those laws, whatever they may be. As the Court noted in the Equal Protection portion of Loving: ‘While the state court is no doubt correct in asserting that marriage is a social relation subject to the State’s police power . . . the State does not contend that its powers to regulate marriage are unlimited notwithstanding the commands of the Fourteenth Amendment. Nor could it do so in the light of Meyer v. Nebraska, 262 U.S. 390 (1923), and Skinner v. Oklahoma, 316 U.S. 535 (1942).’ In other words, although Loving itself dealt only with statutes criminalizing, outlawing, or otherwise prohibiting marriages based on race (or, by extension, other suspect classifications), it reminds us that all state marriage laws are subject to the Equal Protection Clause (in the sense that the Equal Protection Clause sets a minimum standard that all such laws must meet).
Let’s suppose, then, that a state has enacted laws recognizing same-sex unions as ‘civil unions’ or ‘domestic partnerships’, and created whatever associated privileges are needed in order to satisfy the Due Process requirements we discussed above. On what basis could that state then differentiate same-sex civil unions from civil marriages?
We’ve noted that sexual orientation isn’t a currently ‘suspect classification’ — but under Romer, we don’t need to argue that it should be. All we need to show is that any law distinguishing between same-sex and opposite-sex unions is based on a ‘classification . . . for its own sake’, whether ‘suspect’ or not, and Romer will declare the law unconstitutional.
This seems to be a pretty light burden. It would be a different matter if we hadn’t already argued that same-sex unions merited any sort of recognition; then we’d face the problems we noted above — namely, that a legal definition of marriage might not need to rely on sexual orientation at all, and that even if it did, it might be for a legitimate reason. Those counterarguments make at least some sense if the law doesn’t grant any legal recognition to other forms of committed union. But given that a body of law that recognizes opposite-sex unions must also recognize same-sex unions, it’s hard to see how that body of law could distinguish them without relying in any way on sexual orientation. (And we’ll be assisted here by the usual rulings that ‘separate but equal’ just doesn’t cut it as far as the Fourteenth Amendment is concerned.) The consequence is that any state law granting recognition to same-sex unions but distinguishing between such unions and civil marriages will probably fail to satisfy the Equal Protection Clause.
(Even supposing that some such distinctions are found to have an independent justification, it’s still possible to argue that the Equal Protection Clause requires that such unions be regarded as ‘variant’ marriages rather than as something else. But I won’t try to make that argument here.)
This is, I think, the right way to invoke the Equal Protection Clause: establish a fundamental right first, and then argue that the law can’t differentiate in respecting or protecting it based on any sort of ‘classification . . . for its own sake’. That’s the real purpose of the Equal Protection Clause anyway: not to secure ‘special rights’ against discrimation for various protected ‘classifications’, but to secure everyone’s (independently established) rights across the board by forbidding states to act as though certain ‘classifications’ lack those rights.
This makes better political and strategical sense as well. Many people, when they hear an argument based solely on the Equal Protection Clause, think: ‘Aw, geez, another group of “victims” whining about “discrimination” and demanding special rights’. The argument we’ve outlined here makes clear that the right to marry whom one chooses is not a ‘special right’ that one gets simply by being discriminated against; it’s a fundamental liberty right that is being systematically denied to people who happen not to be heterosexual.
It’s easy enough to dismiss discrimination against an abstract ‘classification’, but it’s much more difficult to deny the rights of individual human beings. When faced with an actual human being, say a gay man who isn’t able to marry the man he loves because the law doesn’t recognize such a marriage, it’s a hard soul indeed who doesn’t at least understand what the problem is. And Constitutionally this is as it should be: as we’ve said, the ‘equal protection’ of individual liberty rights is what the Equal Protection Clause is really about.
Moreover, reliance on ‘classifications’ tends to reinforce a sense that gays and lesbians are somehow fundamentally ‘different’ from straight people: they must be if they have a legal classification of their own, right? Putting the emphasis on fundamental rights, where it really belongs, takes the focus off ‘classifications’ and focuses instead on the important ways in which all of us are alike.
Anyway, under our analysis here, the Defense of Marriage Act (DOMA) is at best irrelevant as far as states are concerned: since state jurisdictions have to recognize same-sex marriages anyway, there’s not much point in a law that says they don’t have to recognize each other’s. (The DOMA may already be unconstitutional under the Full Faith and Credit Clause because it relieves states of the obligation to recognize marriages performed in other states’ jurisdictions.)
The federal level
Lawrence and Romer apply to state laws via the Fourteenth Amendment. What about federal law? Could Congress pass a law establishing a national definition of marriage as ‘one man and one woman’?
The logic of Lawrence suggests not. Although its argument applies strictly to Fourteenth Amendment analysis, its broad holding that the mere promotion of morality isnever rationally related to a legitimate government interest seems to apply to all levels of government. Even if this conclusion does not strictly follow from the holding itself, the holding is at least a clear indication of what the Court would hold if a case came before it under federal law.
In that case any federal statute defining ‘marriage’ as a heterosexual union, or in any way forbidding the inclusion of same-sex unions as civil marriages for federal purposes, would also be unconstitutional. (There are lots of federal rights associated with marriage that state-level recognition alone wouldn’t protect.) The argument here simply parallels the Due Process argument we gave above, and to make it we would invoke the Due Process Clause of the Fifth Amendment instead of the Fourteenth.
Before we leave this subject, we need to say a few words about a case that might appear to stand in our way: Reynolds v. U.S., 98 U.S. 145 (1878), in which the Supreme Court upheld a federal law prohibiting polygamy in U.S. territories including Utah and Idaho. (The law was the Morril Act, and it was passed during the height of the nineteenth-century controversy over the Mormons.)
Essentially, Reynolds held that the federal government is within its Constitutional bounds in regulating the definition of marriage, and that the First Amendment’s Free Exercise Clause doesn’t provide some sort of ‘religious exception’ to such laws: if the government makes polygamy a crime, then it’s a crime even if the polygamist in question is ‘polygamizing’ out of a sincere religious belief.
Several replies to Reynolds are available.
First of all, its conclusion that the federal government may constitutionally regulate marriage seems to be founded on the view that chaos could result if more than one form of marriage were permissible. This view can surely be challenged factually in the case of same-sex marriage (and perhaps also for polygamy as well, but that’s a bit beyond our scope here). For example, there are lots of same-sex couples currently cohabiting in committed relationships, and society hasn’t collapsed into chaos as a result; it’s hard to see why granting their relationships legal recognition would suddenly plunge us into the Outer Darkness. So the Court’s reasoning might not justify government regulation after all.
Second, the Court’s main argument relies heavily on the fact that polygamy has long been regarded as ‘odious’ in Western civilization. This portion of its argument bears an uncomfortable resemblance to Bowers and may already have been overturned in Lawrence.
Third, the Court seems to examine only the question whether such federal regulation is a violation of the Free Exercise Clause; it never explicitly reaches the question whether any other part of the Constitution might prohibit such regulation. In particular, it never considers whether a right to marry might be a liberty/privacy right entitled to Constitutional protection.
Fourth, the Court doesn’t actually hold that Congress must choose monogamy; it simply holds that Congress has the power to decide the matter. Even if Congress still has that power, it doesn’t follow that Congress should decide against same-sex marriage.
And fifth, the holding in Reynolds is, after all, about polygamy and not about same-sex marriage. Strictly speaking, the entire decision could be left intact without granting Congress the power to forbid same-sex marriages too.
To the extent to which Reynolds is relevant to same-sex marriage, its legal holdings have probably been overruled and its factual basis is subject to dispute. And even if it is still good law with respect to polygamy, the arguments we have summarized here tend to show that there is no reason to extend it to cover same-sex marriage.
A federal Constitutional amendment
If the arguments we have given are correct, then opponents of same-sex marriage have only one option (other than acquiescing to the Constitutionally inevitable): to pass a Constitutional amendment giving a national definition of marriage. (Even this might not be the end of the story; some Constitutional scholars believe that the Supreme Court could strike down even a duly ratified amendment if it conflicted with the essential spirit of the Constitution. Whether the Court would do so is another question.)
I’m firmly opposed to such an amendment, but I won’t discuss the subject further in this essay. In closing, though, I will remark briefly on what I think is a better solution.
Civil marriage is not the only form of marriage there is. The Roman Catholic Church, for example, doesn’t recognize civil marriages as satisfying the requirements of its own canon law, and legally it doesn’t have to. Other religious organizations have their own standards as well, and there’s nothing in the law that requires them to accept civil marriages as sufficient for their own purposes.
‘Legalizing’ same-sex marriages doesn’t alter that situation. If a church or other organization doesn’t want to recognize same-sex civil marriages as ‘marriages’ under its own definition and for its own purposes, it doesn’t have to. And at any rate, surely those various religious groups don’t think that civil marriages will suddenly become religiously adequate just because a national one-man-one-woman definition is adopted.
So why not stop using the power of the State to fight a culture war? Let civil marriages be available to everyone on the terms we’ve described here, and let religious organizations refuse to recognize them for religious purposes as they think necessary.
What do you think of this article? Discuss it on Scholardarity’s message board.
PARENT PAGE: Philosophy