Once again American courts are at the center of bitterly divisive moral controversies, this time about euthanasia and homosexuality. Millions of people think that doctors are murderers if they help patients, even those dying slowly in great pain, to kill themselves; the American Medical Association has just confirmed its longstanding opposition to euthanasia, and most states have made assisting suicide a crime. But last March, in a decision that would have seemed incredible a few years ago, the Ninth Circuit of the US Court of Appeals declared that Washington State’s law forbidding assisted suicide was unconstitutional, and a month later the Second Circuit held New York’s similar law unconstitutional as well. The Supreme Court has not yet indicated whether it will hear appeals from these cases. It is by no means certain that the Court will ultimately confirm them. But if it does, judges will have brought about another social revolution.

In May the Supreme Court considered a different social problem: how far states are free to discriminate against homosexuals. Colorado voters had enacted a sweeping amendment to that state’s constitution, declaring that no law protecting homosexuals from discrimination hitherto or henceforth enacted in the state or any part of it would be valid. That struck many constitutional experts as an appalling interference with normal political practice—why shouldn’t homosexuals be as free as other groups to campaign for laws protecting their interests in local and state legislatures?—but the experts were doubtful that the Supreme Court would invalidate what Colorado had done. The Court had itself declared, in Bowers v. Hardwick in 1986, that states could make homosexual sex a crime, even between consenting adults. How could it now declare that a lesser and more speculative burden, which only prevented special legislation in favor of homosexuals, was unconstitutional? But in a 6-3 decision it did so declare, and Justice Scalia, in an outraged dissent, accused it of taking sides in a “Kulturkampf,” or culture war, that it had no business joining.

A common issue runs through these three recent decisions. May a “moral majority” limit the liberty of other citizens on no better ground than that it disapproves of the personal choices they make? That was the central question in the assisted suicide cases. Though almost all Americans agree that human life is sacred in some way, they disagree about whether it follows that people must never kill themselves, even to avoid terrible pain or crippling indignity and even when they will soon die anyway. Some think it degrades life to end it prematurely, even in those circumstances; others think it degrading not to die in dignity when further life would be appalling. Should that decision be made individually, each person deciding for his own life out of his own conviction? Or should it be made collectively, so that the convictions of the majority are imposed even on those whose most basic beliefs are thereby compromised?1

Sexual morality is also central to people’s lives and personalities. Should adults be free to make their own decisions about sex when these decisions have no direct impact on others? If so, how far should others then be free, as private persons, to express their disapproval of those decisions in their own choice of employees, associates, or teachers for their children? Americans have accepted that some forms of private discrimination are matters of public concern, and that the law should guarantee equality of treatment in many spheres for blacks, women, and the handicapped. Why not for homosexuals as well? Does it matter that sexual behavior, unlike race or gender or handicap, is finally a matter of choice? Scientists disagree about how far genetic factors fix sexuality, although it seems undeniable that they have at least a significant role. In any case, abstaining from homosexual sex would mean no sex at all for many people, or living a lie. Should society allow discrimination against people who refuse to make a choice with such costs?

These great questions of personal and political morality arise in any modern society, and lawmakers must confront them. In the United States—and in many other nations and international communities that have followed our lead in establishing constitutional rights—judges must face those issues as well. The American Constitution contains two pertinent provisions, and they both played a prominent part in the three recent cases. The Fourteenth Amendment provides that no state may constrain any citizen’s liberty without “due process of law.” The same amendment requires states to extend to everyone “the equal protection of the laws.” The essential difference between the two clauses lies in their rationale. The due process clause forbids compromising certain basic rights altogether, except for a particularly compelling reason. The equal protection clause is less stringent: it requires only that states not discriminate unfairly in the liberties and other privileges it chooses to allow.


But both clauses are exceedingly abstract. How should judges decide which liberties the due process clause treats as basic, and what kinds of discriminations the equal protection clause treats as unfair? For over a century two sharply opposed views have been fighting a constitutional War of the Roses, with first one side and then the other achieving temporary dominion. One party, anxious to restrict the power of judges to adjudicate moral issues, insists that the due process and equal protection clauses give legal protection only to a limited list of rights that have been recognized and enforced during America’s post—Civil-War history. In the Bowers v. Hardwick decision, Justice Byron White set out this view of the due process clause in a passage that has become a talisman for the party of history. He said that the clause protects only those rights that meet one of two tests that he treated as identical: the right must be such that “neither liberty nor justice would exist if [it] were sacrificed,” or it must be “deeply rooted in this Nation’s history and tradition.” He said it was a decisive argument against the alleged right of homosexuals to be free to practice sodomy that until 1961 all fifty American states outlawed that act; the suggestion that “a right to engage in such conduct” meets either of the two tests is therefore, he said, at best “facetious.”

On this view of constitutional interpretation, which has had immense influence on American law, logic and consistency in principle play little part in identifying constitutional rights. The fact that the court has recognized one right—a right to abortion, for example—provides no argument why it should also recognize any other right—the right of homosexuals to sexual freedom or of dying patients to control their own deaths, for instance—even if no principled reason can be given why people should have the former right but not the latter ones. The only issue is whether the particular right in question has been historically recognized, and that test must be applied independently to each suggested right, one by one. Only by such a procedure, as Judge White made plain, can the power of judges to expand constitutional rights in the name of consistency be curtailed. “Nor are we inclined,” he said, “to take a more expansive view of our authority to discover new fundamental rights imbedded in the Due Process Clause…. Otherwise, the Judiciary necessarily takes to itself further authority to govern the country without express constitutional authority.” It is better, on this view, to tolerate inconsistency in the rights the Court recognizes than to expand the list of those rights.

The opposite party in the constitutional wars—the party of principle—denies that order of priority. It insists on “integrity” in constitutional law: it argues that the abstract constitutional rights acknowledged for one group be extended to others if no moral ground distinguishes between them.2 In 1961, a conservative justice—John Harlan—offered one of the strongest judicial statements of this view, and just as White’s formulation in Bowers has become central for the party of history, Harlan’s has become central for the party of principle. The liberty protected by the due process clause, Harlan said, “is not a series of isolated points…. It is a rational continuum which, broadly speaking, includes a freedom from all substantial arbitrary imposition and purposeless restraints.”3 Though Harlan made that statement in a dissenting opinion, it has often been cited in later decisions of the Court. Justices Kennedy, O’Connor, and Souter relied on it, for example, in their crucial plurality opinion of 1992, in Planned Parenthood v. Casey, to explain why the Supreme Court was right to recognize a right to abortion in Roe v. Wade, even though most states had outlawed abortion for decades before that decision.

Both White’s and Harlan’s declarations were cited, by different sides, in each of the three recent court decisions, and though these cases are crucial because the particular issues they decided are matters of life, death, and freedom for so many people, they are also important because they are the latest battles in the long war over the character of our Constitution and therefore of our political society.


A growing number of people—according to a 1990 Roper poll, 64 percent of Americans—now favor laws allowing doctors to help mentally competent patients who choose to die at once if they face only terrible pain or an insensate, sedated existence or a life so diminished that they would prefer death. Juries have repeatedly refused to convict the Michigan doctor Jack Kevorkian for helping patients who seek him out to die, in spite of laws adopted specifically to stop him. In recent years, legislation authorizing doctors to help patients die, under careful control, was narrowly defeated in referenda in California and Washington, and then adopted by referendum in Oregon.4


Perhaps ordinary politics will produce comparable reform during the next decade or so in some other states.5 But it will not do so in all of them, and in any case thousands of patients will suffer needless agony in the meantime. So various groups now argue, on behalf of those patients, that though the Constitution allows states to limit and regulate assisted suicide in a variety of ways, it does not allow them to forbid it altogether. In the Ninth Circuit case, Compassion in Dying v. State of Washington, a group of doctors and patients challenged a law that declared “promoting a suicide” a felony punishable by a five year imprisonment, and defined someone guilty of that offense “when he knowingly causes or aids another person to attempt suicide.” In an elegant opinion, the trial judge, Chief District Judge Barbara Rothstein, held that the due process clause generates a constitutional right allowing competent dying patients to arrange, under suitable controls, an earlier death.6

She was overruled, however, two to one, by a three-judge panel of the Ninth Circuit, which sustained the law. Then the entire Ninth Circuit, citing the “extreme importance” of the case, decided to review its own panel’s decision en banc—that is, with a larger number of its members (in this instance, eleven) participating. In an opinion by Judge Stephen R. Reinhardt, the en banc court reinstated Rothstein’s original ruling and once more declared the law invalid. It recently refused a petition asking for yet another en banc review, with an even larger number of judges sitting, and it is expected that the state will soon appeal to the Supreme Court.

Judges characteristically decide due process challenges to legislation by asking two questions. First, does the law compromise a “liberty interest”—that is, a right that the Constitution in principle protects from state action? Second, if so, are the purposes and effects of the statute so important that they justify a state in nevertheless restricting that liberty interest? The first is the question that divides the parties of history and principle in the way I described. The second, which arises only if a constitutional right is recognized, requires a balance. A court must assess the strength of that right and consider whether the state’s alleged interests are sufficiently strong to justify compromising a right of that strength.

In 1990, in the Cruzan case, a majority of the Supreme Court agreed that citizens have a liberty interest in deciding for themselves when medical treatment that prolongs their lives, including respirators and other life support systems, should be terminated. And though the actual decision held that Missouri had a competing interest strong enough to require proof that a vegetative patient had made an unambiguous choice in advance, a majority made plain that a state could not deny the right altogether.7 Many commentators have assumed, however, that the question in the Compassion case in Washington State—whether the Constitution protects, even in principle, a dying patient’s right to the assistance of a willing doctor in suicide—is a distinct one, because both medical practice and ordinary moral opinion draw a sharp distinction between stopping treatment a patient does not want and administering drugs with the sole purpose of killing him. A right to prevent or stop treatment is part of a more general constitutionally protected right not to suffer unwanted invasions of one’s body. But that more general right does not include a right that invasions the patient desires, like lethal pills or injections, be provided—otherwise it would include a constitutional right for everyone to take narcotic drugs, for example.

Judge Rothstein’s due process argument therefore focused mainly not on the Cruzan precedent, but on the Supreme Court’s abortion decisions instead. She rejected the historical test Justice White applied in Bowers v. Hardwick, which would have denied any right to assisted suicide; she said, in the different spirit of Justice Harlan’s “rational continuum” test, that the crucial question was whether a principled distinction could be drawn between the right to an abortion and the right to control other critical features of one’s own life. She cited, from the key opinion on abortion in Planned Parenthood v. Casey, the pertinent claim that

matters involving the most intimate and personal choices a person may make in a lifetime…are central to the liberty protected by the Fourteenth Amendment. At the heart of liberty is the right to define one’s own concept of existence, of meaning, of the universe, and of the mystery of human life.

The choice of a terminally ill person about whether to die in dignity, Judge Rothstein said, falls within that description at least as plainly as a choice for or against abortion; she said her decision striking down the anti—assisted suicide statute was therefore “almost prescribed” by the precedent of Casey.

Rothstein was right. A right to abortion might have been justified in ways that would have entailed nothing about a right to suicide. Some lawyers argue, for example, that the abortion right flows from a more general right of women to gender equality. But the Supreme Court’s argument for the right, first in Roe v. Wade and then more explicitly in Casey, relied on a different idea: that personal and intimate decisions that reflect basic ethical or religious convictions should be made person by person, not by the state for everyone. That conception of individual dignity plainly applies to decisions about one’s own death.

The opinion for the Ninth Circuit panel that initially reversed Rothstein was written by Judge John Noonan, a noted Catholic theoretician who has been a firm opponent of legalized abortion. He condemned her principle-based reading of the due process clause, and relied on the contrary philosophy of Bowers to rebut it. Since assisted suicide is not the same thing as abortion, he said, Casey cannot be a precedent for a right to the former. Courts must ask, as an independent question, whether history has recognized a right to assisted suicide, and nothing is plainer than that it has not. On the contrary, that practice has been condemned by most American states from the beginning.

The en banc Ninth Circuit reversed both philosophy and decision once again. In a long opinion for the eight-strong majority, Judge Reinhardt declared for principle over Bowers, and for Harlan over White. In a historical survey he showed that suicide has not been universally condemned, particularly not when it has been seen as necessary to end great suffering. But he relied mainly on Rothstein’s argument based on judicial integrity, and took the opportunity to declare that the decision in Bowers “would appear to be aberrant.” “We also note,” he said, “without surprise, that in the decade since Bowers was handed down the [Supreme] Court has never cited its central holding approvingly.”

The three Ninth Circuit judges who dissented from the en banc ruling, on the other hand, made Bowers central to their argument. One of them—Judge Robert Beezer—said he accepted that the right to choose assisted suicide was a “liberty interest” of some force, but not one so fundamental that it could trump the State of Washington’s substantial interests in making such assistance a crime. The other two—Judge Ferdinand Fernandez and Judge Andrew Kleinfeld—doubted any constitutional right to assisted suicide at all, even a weak one, because it is not (in the words of Justice Scalia’s concurring opinion in Cruzan) “a right historically and traditionally protected against state interference.”


The Compassion in Dying case is therefore, at this stage of its history, a victory for the party of principle. So—though less clearly—is the Second Circuit decision in Quill v. Vacco. Three doctors challenged New York State’s penal laws which declare someone guilty of a crime who “intentionally…aids another person to attempt suicide.”8 The doctors supplied affidavits from dying cancer and AIDS patients who were in great pain that could no longer be relieved by drugs and who wanted, in the words of one, “to have drugs available for the purpose of hastening my death in a humane and certain manner.”

The trial judge in the Southern District of New York, Thomas P. Griesa, applied the Bowers test, and dismissed the complaint on the ground that “plaintiffs make no attempt to argue that physician-assisted suicide, even in the case of terminally ill patients, has any historic recognition as a legal right.” The doctors appealed. The Second Circuit, in an opinion of Judge Roger Miner for himself and Judge Milton Pollak, agreed with the trial judge that Bowers blocked any due process claim.9 “As in Bowers,” Miner said,

the right contended for here cannot be considered so implicit in our understanding of ordered liberty that neither justice nor liberty would exist if it were sacrificed. Nor can it be said that the right to assisted suicide claimed by plaintiffs is deeply rooted in the nation’s traditions and history. Indeed the very opposite is true.

But Judge Miner did claim to have found a way to declare the New York statutes unconstitutional anyway: he said that since New York allowed patients to order life-support systems removed (as the Cruzan decision said it must) it denied equal protection of the laws when it refused patients not on life support the aid of doctors in ending their lives. In order to assess that argument (and the arguments in the Supreme Court’s homosexuality decision we shall go on to consider) I must describe the complex test that courts now use to enforce the equal protection clause.

The Fourteenth Amendment language forbidding states to deny any person “the equal protection of the laws” might conceivably have been understood to impose only a very weak requirement on states: that they could only discriminate among their citizens if they first enacted laws describing and authorizing that discrimination. But that banal reading would leave a state free to create a caste system in which blacks (for example) were denied any civil or legal rights, so long as it did so through explicit legislation. Since the Fourteenth Amendment was enacted after the Civil War, with the expectation of preventing the most blatant forms of racial discrimination, that reading is unacceptable.

So is the opposite reading, however, which would declare that states must never enact laws that discriminate in any way among groups of citizens, awarding advantages to some at cost to others. For almost every national or state law has precisely that effect—the NAFTA treaty worked against the interests of some workers and in favor of others; environmental legislation injures some industries though not others; and state banking, securities, and professional regulations help some people but disadvantage others. So the Supreme Court has developed a more sophisticated interpretation of the equal protection clause that avoids either of these extreme and unacceptable readings. It has done this through a set of doctrinal rules and distinctions which, taken together, are calculated to serve an underlying rationale of political morality.

That underlying rationale is a theory which distinguishes between circumstances in which a democracy is working well, so that those who lose out in a political contest cannot complain of procedural inequality or unfairness, and when it is defective, so that losses to some groups cannot be accepted as fair. In the normal circumstances of ordinary politics, groups that lose—as the timber industry, for example, might lose through environmental legislation—have had a fair opportunity to present their case and exert an influence on the result in rough proportion to their numbers and the strength of their interests. The Court will therefore scrutinize ordinary legislation challenged on equal protection grounds only in a “relaxed” way. It will declare such legislation unconstitutional only if it finds that the distinction it draws, between those it benefits and those it harms, is plainly irrational, which means that those attacking it can show that it does not serve, even in a speculative or problematical way, any legitimate purpose of government. So the Court has approved, for example, a law subjecting oculists and optometrists to different regulatory schemes, even though no very impressive reason could be given why they should be treated differently. Only rarely, in fact, has any statute been found to violate this “relaxed” text of rationality.

In some circumstances, however, the general presumption that the political process has worked in a fair way is doubtful. That presumption cannot rescue legislation that deprives some group of the very political rights it needs in order to participate in the process on fair terms—when the legislation reduces the voting power of some group, for example, so that its political impact is made less than its numbers would otherwise justify. The Court has therefore created a different, “strict” or “heightened,” level of scrutiny for laws that have that effect. It declares such laws unconstitutional, even if they are rationally related to some legitimate state purpose, unless they can be shown to be necessary to prevent some grievous result that cannot be avoided in any other reasonable way. The Court relied on the strict test, for example, in its series of “reapportionment” decisions in which it struck down state schemes for drawing boundaries of electoral districts whose effect was to deny equal electoral impact, citizen by citizen, on a one-person, one-vote basis.10 Just as the “relaxed” test is rarely failed, so this “strict” test is rarely passed.

The presumption of a fair political process is also doubtful when the group that loses in one that has historically been the victim of a prejudice or stereotype that makes it likely that its interests will be discounted by other voters. Blacks have often lost out in politics, for example, not because their own interests were outweighed by those of others in a fair contest, but for one or both of two other reasons: because they were economically depressed and socially without influence, and lacked the training and means needed to command the attention of politicians and other voters, or because many white citizens voted for discriminatory laws not just to protect their own competing personal interests but because they held blacks in contempt and wanted them subjugated.

So the Court has created another special category attracting “heightened” scrutiny: it declared that blacks form a “suspect” class, and that any legislation that works to their special disadvantage must be struck down unless it can be defended as serving some absolutely compelling purpose.11 That strict test, too, has rarely been met. The Court has added other groups to the list of “suspect” classes deserving special protection: ethnic minorities and immigrants. It has, moreover, created a further category of “quasisuspect” classes—these now include women and illegitimate children—and declared that legislation working against them will also receive “heightened” (but not as “strict” as in the case of fully “suspect” classes) scrutiny.

We must consider Judge Miner’s claim in Quill v. Vacco—that laws forbidding assisted suicide deny equal protection—against that complex doctrinal background. The judge could not plausibly hold that dying patients form a suspect or quasi-suspect class—such patients are objects of sympathy, not prejudice—and Bowers prevented him, he said, from supposing that such patients have a fundamental right to a willing doctor’s aid in killing themselves. So he conceded that only relaxed scrutiny was appropriate, but he claimed that even on that test New York failed, because no rational ground could justify allowing some patients to terminate life support but not allowing others, who need a different kind of help in dying, to have it.

It is true that some of the grounds that have been urged for that distinction are dubious. It is sometimes said that doctors who do not use life-saving equipment are only failing to act to save life, while those who prescribe pills are positively helping patients to die. But though the distinction between acts and omissions is often valid—there is a moral difference between killing and failing to contribute to a life-saving charity—that distinction does not seem important in this context. On the contrary, it may seem more humane for a doctor to give a patient a lethal injection that kills at once than to terminate life support and allow a patient to suffocate or starve to death. In any case, moreover, removing life-support systems already in place, which the Cruzan case said states must allow, is as positive an act as an injection.

The relaxed-scrutiny standard requires only that there be some speculative ground for a distinction, however, and I have already mentioned one that seems strong enough to pass that weak test. A patient’s right to decline life support, or to order the removal of life support already in place, can plausibly be thought part of people’s undoubted right to control invasions of their own bodies. A right to a lethal injection or a lethal dose of pills cannot. So the State of New York can argue that its distinction is perfectly rational. It wants to preserve life in whatever way it can, and therefore to prevent as much suicide as is legally possible. It cannot forbid people from terminating their life support, because they have the right just described. But (if Miner is right in his judgment that patients have no independent liberty interest in assisted suicide) it can forbid them lethal injections or pills. So it does.

The Second Circuit’s argument, understood as a classic equal protection argument, is therefore fragile. It is much stronger, however, if we read it as the due-process argument Miner denied he was making—not the argument that New York must allow assisted suicide to one group of patients only because it allows others to direct that life support be removed, but the more fundamental argument that New York must respect a right that applies with equal force in both cases and could not be denied in either or both: the basic right of citizens to decide for themselves whether to die at once or after prolonged agony. For the most powerful passages in Miner’s opinion are not those in which he argues that New York cannot rationally distinguish between the two forms of hastening death, but those in which he denies that New York could have any legitimate interest in opposing either form. “But what interest can the state possibly have in requiring the prolongation of a life that is all but ended?” he asked. “And what business is it of the state to require the continuation of agony when the result is imminent and inevitable?” He then quoted exactly the passage from the plurality Casey opinion on which Judge Rothstein and the Ninth Circuit had relied, which asserted a due process “right to define [one’s] own concept of existence, of meaning, of the universe, and of the mystery of human life.”

Though press reports have emphasized that the Ninth and the Second Circuits relied on different constitutional provisions in striking down statutes forbidding assisted suicide, the difference was, in fact, only one of candor. The Ninth Circuit was able to justify its Compassion decision naturally—and persuasively—by refusing to follow the philosophy of Bowers. The Second Circuit, because it tried to defer to that philosophy, offered a bad argument first, and then, in thin disguise, the same argument the Ninth Circuit had offered openly.


Many Americans have become concerned about the legal, economic, and social disadvantages that homosexuals still suffer in this country, and in recent decades they have supported laws and industrial and academic regulations prohibiting or limiting such discrimination. The Colorado cities of Aspen, Boulder, and Denver recently enacted legislation that protected homosexuals, along with minority races and women, from discrimination in housing, education, employment, and health and welfare services. Other Colorado voters were outraged, however, by the suggestion, implicit in such legislation, that homosexuality is a legitimate way of life. In a 1992 statewide referendum, they adopted “Amendment 2” to their state constitution, which was titled, “No Protected Status Based on Homosexual, Lesbian, or Bisexual Orientation.” It declared that:

Neither the State of Colorado, through any of its branches or departments, nor any of its agencies, political subdivisions, municipalities or school districts, shall enact, adopt or enforce any statute, regulation, ordinance or policy whereby homosexual, lesbian or bisexual orientation, conduct, practices or relationships shall constitute or otherwise be the basis of or entitle any person or class of persons to have or claim any minority status, quota, preferences, protected status or claim of discrimination.

This provision, if valid, would have a catastrophic effect on the political situation of homosexuals in Colorado. It would annihilate the protection that some cities had already given, and forbid any political subdivision of the state, and indeed the state itself, from enacting any protective legislation in the future. Homosexuals could thereafter secure anti-discrimination legislation only by further amending the state constitution itself, to repeal or amend Amendment 2.

That struck many people, within the state and outside it, as monstrously unfair. They assumed that there must be some ground on which Amendment 2 violated the national constitution, and a group of Colorado homosexuals and others sued in a Denver court, in the case of Romer v. Evans, asking for a ruling to that effect. Many constitutional lawyers were dubious, however, that they could win. The doctrinal background I have already described, governing due process and equal protection, made it seem doubtful that Amendment 2 violated either of those clauses, and no other constitutional provision was pertinent.

For the Bowers precedent was now directly applicable. Justice White had explicitly declared that homosexuals do not have a “liberty interest” in freedom of sexual conduct that can prevent states from declaring homosexual sex a crime. It therefore seemed impossible to argue that they have a liberty interest that would bar the less serious disadvantage of Amendment 2, which merely prevents them from obtaining special legislation in their favor.12 As the District of Columbia Circuit had observed in 1987,

If the Court was unwilling to object to state laws that criminalize the behavior that defines the class, it is hardly open…to conclude that state sponsored discrimination against the class is invidious. After all, there can hardly be more palpable discrimination against a class than making the conduct that defines the class criminal.13

The equal protection clause might seem a more promising basis for objection, because Amendment 2 denied homosexuals a political opportunity—to attempt to secure local legislation protecting their basic interests—open to other groups. But, as we have seen, an equal protection challenge to any legislation must show either some reason why “heightened scrutiny” of that legislation is appropriate, or that the legislation is irrational because it does not bear even a speculative relation to a legitimate governmental purpose. Heightened scrutiny is appropriate if the group that is disadvantaged counts as a “suspect” or “quasi-suspect” class, and homosexuals are certainly targets of prejudice and irrational hatred. In the years following the Bowers decision, however, several federal courts held that homosexuals nevertheless do not count as a suspect or quasi-suspect class, for a reason it is important to explain. Suspect groups are those that lack the political power necessary to make the political process a fair and democratic one for them. But a group might lack that power for either of two different reasons that I distinguished, earlier, in discussing the case of blacks.

First, it might be so disadvantaged financially, socially, and politically that it lacks the means to attract the attention of politicians and other voters to its interests, and so cannot wield the power at the polls, or in alliances or horse-trading compromises with other groups, that its numbers could otherwise be expected to produce. Second, it might be the victim of bias, prejudice, hatred, or stereotype so serious that a majority wants it constrained or punished for that reason, even when this does not serve any other, more respectable or legitimate, interests of other groups.14

Blacks and the other groups the Supreme Court has hitherto treated as suspect or quasi-suspect suffer from both these disabilities. But (at least in the view of the judges who have spoken on the issue) homosexuals suffer only from the second.15 Justice Scalia, in his Evans dissent, insisted that homosexuals have at least the political power their numbers would warrant: “Because those who engage in homosexual conduct,” he said,

tend to reside in disproportionate numbers in certain communities, have high disposable income, and of course care about homosexual-rights issues much more ardently than the public at large, they possess political power much greater than their numbers, both locally and statewide.

If a state like Colorado rejects the homosexuals’ case, he said, it is not because they have not had a chance to organize their political efforts, or to speak effectively to their fellow citizens, but because, in spite of the fact that they have had those opportunities, the majority has decided against them.

It is therefore crucial to decide whether the second disadvantage—the prejudice and contempt of a potential majority—is an independent defect in the proper functioning of a democracy, a defect that is sufficiently serious to justify heightened scrutiny of legislation that harms those who suffer from such prejudice. The Bowers decision answered that question negatively. The groups who challenged the Georgia anti-sodomy law in that case argued that a state has no right to enact criminal legislation when its only reason is that the majority morally disapproves of those it makes criminals. Justice White replied that “the law, however, is constantly based on notions of morality, and if all laws representing essentially moral choices are to be invalidated under the Due Process Clause, the courts will be very busy indeed.”

White’s comment missed the point. Of course most of the criminal laws that a community enacts express a moral choice: laws against murder express a moral condemnation of that activity. The groups challenging the Georgia anti-sodomy law argued only for the narrower principle that a criminal law is unconstitutional if it is enacted only to condemn some people morally, and not to protect anyone else’s direct interests. Laws against murder do more than denounce murderers: they protect the most basic interests of innocent people. Making consensual adult sodomy a crime, on the other hand, serves no interests that are independent of the moral condemnation, and those who challenged Georgia’s anti-sodomy law argued that that is not a legitimate justification for a criminal penalty.

White clearly meant to reject that narrower principle as well, however, and Bowers therefore stands for the principle that it is permissible for government to prohibit freedom of choice in private sexual behavior even if that behavior harms no one in any direct way, so long as the condemnation expresses popular morality. Since the Bowers decision, judges have therefore uniformly rejected the suggestion that homosexuals form a suspect or quasi-suspect class for equal protection purposes, and every judge who expressed a clear opinion on that issue during the litigation of the Evans v. Romer case agreed.

There is, however, another kind of legislation that also attracts the “heightened scrutiny” test, as we noticed: legislation that compromises a fundamental political right. The claim that Amendment 2 violated a fundamental political right does not depend on supposing that homosexuals are a suspect or quasi-suspect class. On the contrary, it could be sustained only by showing that it would be unconstitutional to treat any group of citizens—people who rent houses, for example, as opposed to owning them—as Amendment 2 treated homosexuals. But, just for that reason, it is enormously difficult to define what the right in question is.

The Denver trial judge who first considered the case, H. Jeffrey Bayless, decided that the amendment did violate a fundamental right: the right “not to have the State endorse and give effect to private biases.” But that is exactly the right that Justice White, in Bowers, said that no one has. So when Colorado appealed the trial judge’s decision to the Colorado Supreme Court, that court upheld the trial judge’s decision but cited a different fundamental right—the right of all groups to “participate equally in the political process.”

There is an evident difficulty in that suggestion as well, however. Amendment 2 did not diminish or dilute anyone’s voting power. It did mean that one group—homosexuals—would have to amend the state constitution again to secure legislation of particular concern to it. But no group has a right not to have to amend the constitution to obtain legislation it favors or believes it needs. It would not be unconstitutional, in principle, for Colorado’s constitution to forbid municipalities to adopt rent-control legislation, for example, even though that would pose special problems for people who rent. Indeed, the US Constitution has parallel disabilities: groups who fervently want prayer back in their schools would have to repeal the First Amendment before petitioning the local school board.

Many lawyers were therefore fearful, when Colorado appealed from its own supreme court to the national one, that even if a majority of the justices wanted to strike down Amendment 2, they would not find room to do so within the network of doctrine and precedents they had themselves constructed. But a group of some of America’s most distinguished constitutional law scholars submitted an ingenious brief, as amici curiae, or friends of the court, which pointed a way out of these doctrinal difficulties.16 The academic brief bypassed the elaborate structure of categories and distinctions I described, about suspect classes, fundamental rights, and different levels of scrutiny, by insisting that it is an automatic (or “per se“) violation of equal protection for the state to declare that any group of citizens is simply ineligible for any protection whatever from any form of the harm of discrimination. Such a declaration, the brief said, in effect outlaws the group, and it was the central point of the equal protection clause to prevent that kind of caste distinction.

This argument, the brief insisted, assumes nothing about homosexuals being a suspect class. A state constitutional provision would violate equal protection, it said, if it forbade any legislation protecting any group of citizens against any form of discrimination. It offered this example: a state constitution would deny equal protection for people who rent homes (surely not a suspect class) if it provided that any local or state legislation that protects renters from any imaginable harm or loss is invalid. But this comparison reveals how limited an argument the brief was making, and how carefully it had tailored that argument to the facts of this particular case. It made the unconstitutionality of Amendment 2 depend on its great breadth and, as the example of home renters shows, that argument would permit a state to pass a narrower but still destructive constitutional amendment aimed at homosexuals.

The brief conceded, for example, that it would not be a per se violation of equal protection for a state constitution to forbid local rent-control legislation, even if such a law were necessary to protect renters from unjustified and oppressive rent increases. So, by a parity of reasoning, it would not be a per se violation for a constitution to forbid a single form of anti-discrimination legislation—local laws from banning discrimination against homosexuals in hiring, for example, or in hospital admissions. We might be tempted to say that these are different matters: a provision forbidding rent control might express an economic theory, while one forbidding even a specific and limited form of help to homosexuals would express only bias. But the academic brief’s argument was careful, for reasons I have tried to make clear, not to rely on that distinction, because that would have been tantamount to declaring homosexuals a suspect class, or to affirming the right Judge Bayless had cited but the state supreme court had feared to endorse—a right to be free from legislation motivated by bias or hatred.

When the Supreme Court delivered its long-awaited verdict in May, Justice Kennedy’s opinion for the six-justice majority was surprisingly bold—bolder than the opinions of either of the lower courts and bolder than even the academic brief. (As Justice Scalia wryly noted, the Court has the luxury, as other judges and lawyers do not, of ignoring its own precedents.) Kennedy did accept, in general terms, the argument of the academic brief. He emphasized, as that brief did, that Amendment 2 was wholly novel in the sheer breadth of the potential damage it worked on homosexuals, by depriving them of any possible opportunity to secure protection, except by constitutional amendment, against any form of discrimination no matter how harmful or wrongful.

But, perhaps understanding the limitations of that argument, he also made a much broader and potentially more reforming claim whose significance the press has not adequately reported. He said that Amendment 2 violates even the most relaxed form of scrutiny under equal protection doctrine, because it is not even rational. “In the ordinary case,” Kennedy said, “a law will be sustained if it can be said to advance a legitimate government interest, even if the law seems unwise or works to the disadvantage of a particular group, or if the rationale for it seems tenuous.” But, he said, “Amendment 2’s sheer breadth is so discontinuous with the reasons offered for it that the amendment seems inexplicable by anything but animus toward the class that it affects; it lacks a rational relationship to legitimate state interests.”

That statement is of crucial importance, because it flatly contradicts Justice White’s pivotal assumption in Bowers. White, you recall, declared that it is legitimate for a state to impose a disadvantage on a particular group just to express the majority’s moral contempt for that group’s practices, even when no other proper purpose, such as protecting anyone’s economic or security interests, is served. Kennedy, in the passage just quoted, said that this is not legitimate. It is true that White spoke in terms of moral disapproval and Kennedy in terms of “animus.” But there can be no difference in what these words mean in this context.17 For Colorado could certainly declare, in perfect good faith, that the amendment’s “sheer breath” was justified by the depth of its citizens’ moral disapproval of homosexuality. Nothing less than a complete ban on any law that suggests that homosexuality is an acceptable form of sexual union, it might say, would be enough to express the profoundness of the majority’s rejection of that moral opinion. By describing that justification as one appealing to “animus,” and declaring it illegitimate, Kennedy reached back to Judge Bayless’s original judgment in the trial court, and drew the sting from Bowers without even mentioning it.

Kennedy’s reasoning would apply even to a much narrower constitutional provision than Amendment 2, and to many other discriminatory laws as well. He did not put homosexuals in as secure a position as they would enjoy if they were designated a suspect or quasi-suspect class. They have the burden of proof in showing that a particular rule or law that harms them serves no legitimate purpose, but only the illegitimate one of expressing “animus,” and they might find that burden difficult to sustain in many cases—in challenging the military’s opposition to retaining openly homosexual soldiers, for example. But they would presumably not find it difficult to show that the outright criminalization of all homosexual sex serves no purpose beyond that illegitimate one.

Scalia was right, in his biting dissent, that the combination of the result in Evans and the result in Bowers is ludicrous: practicing homosexuals can be jailed but not put at an electoral disadvantage—having to amend a constitution to get legislation they want—that many other groups, including people who favor prayer in schools, suffer. But the inevitable resolution of that conflict may not be the one he would prefer. The members of the majority in Romer v. Evans may have done more than simply ignore Bowers: they may have begun the process of isolating and finally overruling it altogether, an event that would have an enormous impact not only on the civil liberties of homosexuals but, as I have tried to show throughout this essay, on constitutional theory generally.

In its ten-year life, the Bowers decision has frequently been damned by scholars and commentators for its cramped view of what a free society is and does. It was a 5-4 decision, and Justice Powell, who tipped the balance, said after his retirement that it was the worst mistake of his career. Justice O’Connor, another member of that slim majority, joined Kennedy’s opinion in Evans, which may suggest that she, too, now has doubts. Perhaps Bowers would win only three votes if it were directly challenged now: Justice Scalia and Chief Justice William Rehnquist, who were also members of the original Bowers majority, and Justice Clarence Thomas, who joined in Scalia’s dissent.

In any case, the three decisions discussed here, just in themselves, show a further strengthening of the judicial party of principle. Since the end of the period of great Supreme Court activity under the stewardship of Chief Justice Earl Warren in the 1950s and 1960s, when principle was firmly in control, commentators have repeatedly announced that its party was in decline, that the spirit of the Warren Court was dead. It is true that Reagan and Bush appointees now dominate the federal courts, including the Supreme Court, and that these judges are as a group less liberal in personal conviction, more suspicious of affirmative action and other racially sensitive programs designed to help minorities, and less persuaded of the need to protect the interests of the accused than were many of the judges they replaced.

But surprisingly many of the recent appointees have come to sense the importance of America’s great experiment in government under principle as well as law, and to recognize their responsibilities in that experiment. They realize that the Constitution is made of political morality, not historical accident, and that it is unfair to tell any group of Americans—dying patients or homosexuals or any other group—that because their convictions about how to live and die contradict the religious or moral convictions of others, or because others despise their “culture,” they must wait another generation to claim the rights that a principled reading of the Constitution would guarantee them now. Integrity of principle is not only the heart of democratic justice, but the soul of constitutional law as well.

July 11, 1996

This Issue

August 8, 1996