In June, the Supreme Court, in a case called Lawrence v. Texas, ruled that statutes criminalizing homosexual sodomy are unconstitutional. Immediately lawyers began wondering whether this meant that homosexuals have a constitutional right to marry. (To marry persons of their sex, that is; there is no prohibition against a homosexual marrying a person of the opposite sex.) They were encouraged in their speculation by Justice Antonin Scalia’s suggestion, in his dissenting opinion, that the logic of the majority opinion so dictated. Evan Gerstmann has now produced a forceful and tough-minded brief for the result that Scalia dreads. But as he explains, the Lawrence decision came down just days before his book went to press, so he was unable to discuss the decision’s bearing on the homosexual-marriage issue except briefly in a preface; or to discuss the recent decision of the Supreme Judicial Court of Massachusetts in Goodridge v. Department of Public Health, holding that the Massachusetts constitution creates a right of homosexual marriage.

When I say that Gerstmann’s book is tough-minded, I mean that it rejects a number of the arguments made by advocates of a right of homosexual marriage, such as that homosexuals should be considered a “suspect class” (the sensible term would be “suspect classification”), the curious phrase for a group, such as blacks, that has historically been discriminated against; or that forbidding homosexual marriage is a form of sex discrimination, either because it uses a sexual classification (a man can marry a woman but not a man) or because it is part of an overall pattern of discrimination against women, insofar as homosexual marriage would undermine the norm of “patriarchal” marriage in which a man dominates a woman. Gerstmann argues that labeling a group of people as victims who deserve the special solicitude of the courts demeans them. He believes that the prohibition against homosexual marriage is not based on gender, since it gives men and women exactly the same rights; and that it is ridiculous to think that homosexual marriage is prohibited as a way of keeping women down rather than homosexuals down. These are powerful arguments, especially the last two. It might be added that to assimilate homosexual men to women reinforces rather than combats stereotypical thinking about homosexual orientation.

Gerstmann also points out that the fact that sterile people are permitted to marry is not the killer argument against distinguishing heterosexual marriage from homosexual marriage that it may seem to be. The law frequently and unexceptionably draws crude lines—for example, it forbids non-citizens to vote in our elections without insisting that citizens demonstrate that they are in fact loyal to the United States in order to be allowed to vote. He even acknowledges that, under current understandings of constitutional law, “society can oppose homosexuality as a moral matter without engaging in unconstitutional animus.” Government may not forbid abortion, but it can denounce it.

Gerstmann falters in places, as when he uses studies that show that married people live longer and are less promiscuous than unmarried people to argue that homosexuals would obtain comparable benefits if they were allowed to marry each other. All the studies are of heterosexual marriage. He also errs when he cites the estimate that eight to ten million American children are being raised by homosexuals. Homosexuals are at most 3 percent of the population; most of them are not raising children; and the under-eighteen American population is only about seventy million. So the estimate is off by at least an order of magnitude: if—and this, too, is undoubtedly an overestimate—one-third of homosexual Americans are raising children, then 1 percent of the nation’s children are being raised by homosexuals, which comes to seven hundred thousand.

But these are details, and anyway Gerstmann’s argument for a constitutional right of homosexual marriage is based not on empirical fact but on constitutional theory. His argument is that the Supreme Court—rightly, in his view—has ruled that marriage is a “fundamental right,” meaning that a state cannot take it away without a compelling reason; and the state does not have a compelling reason in the case of same-sex marriage. In a case called Turner v. Safly, the Court in 1987 ruled that a prison inmate could not be denied the right to marry, although the prison could forbid conjugal visits. If prisoners, why not law-abiding homosexuals?

But Gerstmann overlooks a “fundamental” difference. The Court in Turner was not expanding the basic right to marry as defined by marriage law and custom. It was not as if marriage laws forbade prisoners to marry. The question was whether a prisoner could be denied the same right (actually a greatly diminished right, since it did not permit the prisoner to live with his wife) enjoyed by non-prisoners. When Gerstmann describes the right to marry as fundamental, he means that any person who wants a marriage license has a strong presumptive right to it regardless of how the person defines marriage. He might be a man who wanted to marry his sister (both being sterile), or a very mature twelve-year-old boy (say, a freshman at MIT) who wanted to marry his twelve- year-old girlfriend (say, a freshman at Harvard), or a married man who wanted additional wives so that they might help out his current wife around the house, or a busy professional woman who wanted two husbands, the better to take care of the house and the kids, or a homosexual male who wanted three male spouses.

If the right to marry, irrespective of the conventional limitations on number, object, and so on, is fundamental in the portentous sense of putting on the state the burden of showing that the recognition of the right in the particular case would work some serious social harm, then it is doubtful that a marriage license could be refused in any of the cases that I have described. For what harm does polygamy do, exactly, and what harm does incest do when there is no possibility of children? Gerstmann’s approach thus has implications far beyond the question of homosexual marriage as it is ordinarily understood.

He is aware of the problem, but he cannot solve it. When he attempts to distinguish polygamy from homosexual marriage by saying that denying a right to marry several women does not deny the right to marry the person of your choice, he overlooks the fact that a woman who would like to be a polygamist’s second wife is denied the right to marry the person of her choice. Gerstmann also argues that if a man were allowed to have two wives, there could be no objection to his being allowed to have a thousand wives. Maybe not, but that is just an example of where the logic of his approach leads. (The Massachusetts court in the Goodridge case wisely did not try to explain, but instead merely asserted, that its decision would not disturb the state’s prohibition of polygamy.) And I have not even tasked him with explaining what the state’s compelling interest is in forbidding a man to marry his beloved dachshund.

Gerstmann criticizes the prohibition of homosexual marriage on the ground that it is a matter of the majority trying to take away other people’s right to marry; but that is also what is involved in prohibiting polygamy. He objects to incestuous marriage, even when the partners are sterile, on the ground that it is likely to involve the exploitation of children. This is certainly a reason for forbidding a father to marry his daughter, but not for forbidding marriage between siblings; and he is forced to entertain the possibility that such a prohibition might indeed be unconstitutional. It is a strange implication of Gerstmann’s approach that if a man wanted to marry his sterile sister, his eighty-year-old grandmother, three other women, two men, and his chihuahua, a court would have to turn somersaults to come up with a “compelling state interest” that would forbid these matches.

It should be apparent by now what the problem with Gerstmann’s approach is. Though he is a political scientist as well as a lawyer, his approach to the question of homosexual marriage is legalistic. Find a precedent (Turner v. Safly, or Zablocki v. Redhail, a case in 1978 that invalidated a law prohibiting a person who was under court order to support minor children to marry without the court’s permission), and analogize it to the present case, and use the analogy to put an impossible burden of proof on your opponent, and limit the scope of your rule by rejecting further analogies on however arbitrary a ground, so that the right of a prison inmate to marry is deemed analogous to a right of homosexual marriage but not to a right of polygamous marriage, because the polygamist, unlike the homosexual, is not denied the right to marry the person of his (first) choice.

This is what is called “legal reasoning,” and it is hard to take seriously. For one thing, there is nothing sacrosanct about precedent, especially in the Supreme Court. In Lawrence, the Court overruled a precedent that was not merely analogous to the case at hand, as Turner and Zablocki might be thought analogous to a case involving homosexual marriage, but identical to it. (The case was the notorious Bowers v. Hardwick, which had upheld the validity of criminalizing homosexual sodomy.) For another, it would be child’s play, as a matter of legal casuistry, to limit those two cases to conventional, monogamous, non-incestuous, heterosexual marriage.

Judges like to pretend that their decisions are dictated by “logic,” or by an authoritative text or precedent, because it downplays the element of judicial discretion, which worries people. The pretense wears particularly thin in constitutional cases about marriage and sex, because the Constitution does not say anything about these subjects, and the framers of the Constitution, and of the major amendments, in particular the Fourteenth Amendment, which is the principal source of constitutional rights against the states, were not thinking about marriage, sex, homosexuality, or related topics when they drafted these founding documents. (Neither were the ratifiers.) Decisions such as the four that I have mentioned, together with the Supreme Court’s other well-known sex- related decisions, such as Griswold v. Connecticut (holding that a state cannot forbid married couples to use contraceptives) and Roe v. Wade, are all “political” decisions—not in the narrow Democratic versus Republican sense, but in the sense of being motivated by values not dictated by the orthodox materials of judicial decision-making. Precedent and analogy operate as fig leaves in such cases.

So we have to dig much deeper to answer the question whether there should be a constitutional right of homosexual marriage. We have to consider why homosexuals want the right, what the consequences of giving it to them might be, why the right is so strongly opposed by the public, and whether the Supreme Court is the proper institution to authorize homosexual marriage. Gerstmann touches on almost all these questions, though he does not ask why there is so much public opposition to homosexual marriage, and he is surprisingly but revealingly incurious about what ordinary people think—he is doing constitutional law, after all, and constitutional law is about placing curbs on democratic preference. But he does not do the questions justice.

Regarding the first question, there are a number of material advantages of marriage—marriage confers rights that can be duplicated only imperfectly by contract—though there is also the well-known tax disadvantage (the “marriage tax”) for some married couples. The material advantages of marriage are important to homosexuals, though less so than to heterosexuals, because a homosexual couple is much less likely to be raising children. But probably a bigger issue is social legitimacy. Recognizing homosexual marriage would place a very public stamp of approval on homosexual relationships—which is the principal reason why heterosexuals tend by quite large margins to oppose not only homosexual marriage but even “civil unions” and other marriage simulacra for homosexuals.

In a recent Washington Post poll, 58 percent of the respondents opposed civil unions for homosexuals and only 37 percent favored them; oddly, this was greater than the margin—55 percent to 40 percent—by which homosexual marriage was opposed in another recent poll, this one by ABC and The New York Times. No doubt some parents fear, though almost certainly groundlessly, that the stamp of approval of homosexuality that a right of homosexual marriage would confer would make it more likely that their own children would become homosexuals. But I suspect that more object for the same reason they would object to incestuous or polygamous marriages, or allowing people to marry their pets or their SUVs— that it would impair the sanctity, degrade the institution, of marriage (their marriage) to associate marriage with homosexuality. The point is not that heterosexuality is part of the meaning of marriage. Meanings can change, as Gerstmann emphasizes. The point is that most heterosexuals do not want the meaning of marriage to change. Most of them, though willing to grant homosexuals equal rights in most spheres, do not approve of homosexuality (tolerance is not the same thing as approval) and do not want the state to allow the word “marriage” to be appropriated by homosexual couples.

The Massachusetts court in Goodridge missed this distinction. It thought the fact that the state prohibits employment discrimination against homosexuals and permits them to adopt children showed that the state’s refusal to recognize homosexual marriage was irrational. It is not irrational, because the stakes, including the symbolic stakes, are different. What is more, if a state’s laws must compose a consistent whole, which they never do, the courts’ power to invalidate laws of which they disapprove has no limits. And if it were irrational to permit homosexuals to adopt but not to marry, this would not establish which law should be disapproved.

The question is what weight the strong public disapproval of homosexual marriage should be given in balance with the interests of homosexuals. Gerstmann realizes that government, while it is not allowed to forbid abortion, is allowed to express its disapproval of it by refusing to fund abortion clinics. For that matter, it is permitted—and Gerstmann wants it to be continued to be permitted—to express disapproval of polygamy by outlawing polygamous marriage. So why, then, may government not be permitted to disapprove of homosexuality by forbidding homosexual marriage? The answer might be that this prohibition stigmatizes homosexuals, whereas the people who want to practice polygamy do not belong to a traditionally oppressed minority (this is not entirely true, but I will let it pass). Gerstmann does invoke stigma, but the invocation is in sharp tension with his rejection of “suspect class” status for homosexuals. If homosexuals do not constitute a suspect class, then disapproval of their sexual orientation, expressed in the prohibition of homosexual marriage, is no different from disapproval of polygamists, expressed in the prohibition of polygamy. Gerstmann points out that the courts protect a variety of unpopular groups, such as the Ku Klux Klan, so why not gays and lesbians, who are less unpopular than Klansmen or Nazis? But it is one thing to protect an unpopular group’s freedom of speech, and another to give it the Good Housekeeping seal of approval, or a tax deduction. And remember that Gerstmann had denied that moral disapproval of homosexuality is unconstitutional.

It seems that we have reached an impasse, with the desire of homosexual couples for the material, psychological, and symbolic benefits of marriage opposed by the desire of heterosexuals to withhold the official endorsement of homosexual relationships that would be implicit, or taken as implicit, in allowing them to claim the name and status of marriage. Goodridge cannot break the impasse, because it is the decision of one state, and the Massachusetts court was careful to base the decision on the Massachusetts constitution rather than on the Constitution of the United States. But maybe Lawrence can enable us to overcome the impasse.

Justice Anthony Kennedy’s majority opinion offers three grounds for invalidating criminal laws against homosexual sodomy. The first might be called (it is not Kennedy’s word, of course) the “bandwagon” argument, that there is increased toleration for homosexuals, as indicated by the steady decline in the number of states that criminalize homosexual sodomy and the virtual nonenforcement of such laws in the states that retain them. But the bandwagon argument will not carry the day for homosexual marriage, since no state recognizes such marriage (we have yet to see how the Massachusetts legislature and populace will respond to Goodridge) and only one (Vermont) recognizes even civil unions. The polls that I mentioned earlier understate the opposition to homosexual marriage, because they do not weight their numbers by intensity. It is apparent from the Defense of Marriage Act passed by an overwhelming vote in Congress and signed by President Clinton, which both denies any federal marital rights (such as Social Security survivor benefits) to members of a homosexual marriage and confirms that a state is not legally required to recognize a homosexual marriage contracted in another state, and from the subsequent passage by two-thirds of the states of statutes declaring that indeed their state will not recognize the validity of homosexual marriages wherever contracted, and from the fact that no mainstream politician supports homosexual marriage—it is apparent from all these things that the opponents of homosexual marriage feel much more strongly about the issue than the supporters.

The supporters of marriage, after all, are mainly heterosexuals—of the 40 percent of the population that according to the recent poll endorse homosexual marriage, the vast majority must be heterosexuals, since as I say homosexuals appear to be at most 3 percent of the population; and the only basis of their support, for most of them—an exception may be family members of homosexuals— is that homosexual marriage does not bother them. They do not feel passionate about the issue; it does not affect them. But many of the opponents are passionate in their opposition because they feel deeply threatened by the proposed change in the concept of marriage. And in a democratic society, powerful currents of public opinion deserve recognition even by the Supreme Court, at least in cases to which neither the Constitution nor any other authoritative legal text speaks with clarity.

Kennedy next pointed out that the criminalization of homosexual sodomy subjects those few homosexuals who are actually prosecuted, such as the two defendants in the Lawrence case, to significant disabilities; but those disabilities (such as having to register as a convicted sex offender) have no close counterpart in a prohibition of homosexual marriage. There is a difference between criminal punishment and the denial of access to a social institution. But Kennedy added that Lawrence and his co-defendant “are entitled to respect for their private lives. The State cannot demean their existence or control their destiny by making their private sexual conduct a crime.” When this remark is put together with Kennedy’s acknowledgment elsewhere in his opinion that the condemnation of homosexuality is actuated by “profound and deep convictions as ethical and moral principles,” there is at least a suggestion that moral disapproval of homosexuality is not a valid basis for legal action. (Scalia in his dissent picked up the suggestion, saying that the majority had written finis to any law based on moral disapproval with no accompanying proof of tangible harm, such as laws forbidding sex with animals.)

But it is just a suggestion. All that the decision actually holds is that moral disapproval of homosexuality is not a valid basis for criminalizing homosexual conduct; and, as I have said, the government can express its disapproval of what it regards as immoral activity by withholding a subsidy even when it cannot do so by criminal punishment; and it is a form of subsidy that Gerstmann is advocating, since marriage is a source of privileges. So I conclude that Lawrence will not overcome the impasse over whether there should be a constitutional right of homosexual marriage.

Where, then, are we? I am dubious about interpreting the Constitution to authorize the Supreme Court to make discretionary moral judgments that offend dominant public opinion. Nothing in the Constitution or its history suggests a constitutional right to homosexual marriage. If there is such a right, it will have to be manufactured by the justices out of whole cloth. The exercise of so freewheeling a judicial discretion in the face of adamantly opposed public opinion would be seriously undemocratic. It would be a matter of us judges, us enlightened ones, forcing our sophisticated views on a deeply unwilling population. It would be moral vanguardism. Gerstmann, when he is being tough- minded, inveighs against the Supreme Court’s engaging in “sociological freelancing,” but that is what he is advocating. He rightly does not suggest that moral philosophy can fill gaps in the constitutional text and enable the Court to declare moral disapproval of homosexuality unconstitutional.

There is an alternative to judicializing the issue. It is to submit it to social experimentation. A great advantage of our federal system is that it enables large-scale social experiments, as the Supreme Court recognized recently when it authorized public school vouchers. We can subject the issue of homosexual marriage to experimentation as well. Gerstmann is remarkably court- centric for a political scientist, stating that “the merits of same-sex marriage ... is too important an issue for the federal courts to ignore.” It could just as well be said that there are issues too important to be left to a committee of lawyers—which is what the Supreme Court is, after all.

Vermont has, since 2000, authorized civil unions, and we can learn from its experience with them, imperfect though it must be because Vermont cannot compel the federal government to extend the federal benefits of marriage to members of its civil unions. (This is where the Defense of Marriage Act bites.) Hawaii came close to authorizing homosexual marriage—and now Massachusetts, or at least its judicial branch, has authorized it. It looks as if we will be able to learn something about the consequences of such marriage after all; and, depending on what we learn, other states may follow suit, or the consequences may turn out to be so one-sided that the courts can step in with confidence.

Granted, there is a problem with the suggested experiment; it is the problem that lies behind the Defense of Marriage Act and the follow-on state statutes that I mentioned. If one state authorizes homosexual marriage, then (it might seem) homosexual couples from all over the country would flock there, marry there, and return home, so there would be plenty of experimental observations. But not so; for the states to which they returned would not recognize the validity of their marriages. The Constitution requires states to recognize the validity of legal judgments obtained in other states, but a marriage is not a judgment. Even so, the experiment would not be a failure, because homosexual couples planning to remain in the state would marry and we would learn from their experience how homosexual marriage works out. This seems to me the prudent as well as the democratic way to handle this explosive issue, an issue intractable to legal analysis and inappropriate for judicial resolution. So, paradoxically, although the Supreme Judicial Court of Massachusetts is not a democratic institution, it may contribute to a democratic resolution of the controversy. As Gerstmann remarks at one point in his careful, interesting, worthwhile, though ultimately unconvincing book, “good policy aims are not necessarily good constitutional or judicial aims.”