No judicial decision in our time has aroused as much sustained public outrage, emotion, and physical violence, or as much intemperate professional criticism, as the Supreme Court’s 1973 decision in Roe v. Wade, which declared, by a seven to two majority, that women have a constitutionally protected right to abortion in the early stages of pregnancy.1 For sixteen years anti-abortion groups and political conservatives have campaigned with single-minded conviction to reverse that decision. They proposed without success a series of constitutional amendments, sponsored unsuccessful bills asking Congress to declare that a fetus’s life begins at conception, persuaded President Reagan to appoint anti-abortion judges to the federal courts, waged single-issue political campaigns against candidates who support a right to abortion, and disrupted and bombed abortion clinics.2 The public at large is divided in different ways about different aspects of the abortion issue. A Los Angeles Times national survey reported that 61 percent of Americans think abortion morally wrong—57 percent think it murder—and yet 74 percent nevertheless believe that “abortion is a decision that has to be made by every woman for herself.”

The composition of the Supreme Court has changed dramatically since 1973,3 and now, in Webster v. Reproductive Health Services, the State of Missouri and the Bush administration ask the Court to reverse Roe v. Wade. The Missouri legislature had enacted a statute designed to discourage abortions in spite of that decision. The statute, among other things, declared that human life begins at conception; it required doctors, as part of determining whether a fetus is viable before undertaking an abortion, to perform expensive, often irrelevant, and sometimes dangerous tests; and it prohibited any abortion in hospitals or medical facilities that employ assets owned, leased, or controlled by the state. Lower federal courts declared all these provisions unconstitutional under Roe v. Wade. Missouri appealed to the Supreme Court, asking the Court to overrule that decision or, failing that, to curtail or restrict it in such a way that the Missouri statute would then be constitutional.

Oral argument was heard on April 26, while protesters on both sides of the issue organized long and noisy demonstrations outside the Court. Charles Fried, who was solicitor general in the Reagan administration and has now returned to the Harvard Law School faculty, defended in a brief and in oral argument the Bush administration’s claim that Roe v. Wade should now be discarded. Seventy-eight other briefs—more than in any previous case—were filed by a wide variety of concerned groups. These include, for example, briefs on various aspects of the litigation on behalf of 25 United States senators and 115 congressmen, the American Medical Association and other medical groups, 281 academic historians, 885 law professors, and a large number of anti-abortion groups. The Court is expected to hand down its decision before it adjourns in July, though it might wait until next year. Whatever decision it reaches will frustrate and anger millions of Americans.

Is the human fetus a person from the moment of conception? That question has been argued by theologians and moral philosophers and ordinary people for many centuries. It cannot be resolved by legal research or scientific evidence or conceptual analysis; it will continue to divide people, as it divides Americans now, so long as deep disagreements remain about God and morals and metaphysics. It therefore seems an exceptionally poor issue to ask any court, including the Supreme Court, to decide, and that fact best explains, I think, the immediate appeal many people feel in the idea that the abortion issue is best decided politically, through the ordinary processes of legislation. It seems offensive that a majority of judges on a single court should declare one answer for everyone. It seems more democratic, and also better suited to the inherent complexity of the issue, that different groups of Americans should be permitted to decide, in politics, state by state, which solution fits their own convictions and needs best.

That first impression is misguided in several ways, however. Leaving the abortion issue to state-by-state politics will not, of course, mean that each woman will be able to decide which solution best fits her convictions and needs. It means that if the anti-abortion lobby is sufficiently powerful in a particular state, the women of that state will be denied that opportunity, as they were before Roe v. Wade.4 It is doubtful that there were fewer abortions then, relative to the number of pregnancies, even though most of them were illegal. But there were many more deaths: abortion-related fatalities were 40 percent higher before Roe v. Wade.5 Blacks suffered most. In New York, for example, a black woman was nine times as likely to die in an illegal abortion as a white one. Of course, if Roe v. Wade were reversed women who were rich and knowledgeable enough could still decide to have an abortion by traveling to the nearest or most convenient place where it was legal, as thousands did before 1973 by traveling to Britain, for example. But a poor woman who found herself pregnant might have to choose between the danger of illegal abortion and the misery imposed by and on a child she could not support or raise.


The first impression is misguided not just practically, but legally and logically as well. The key question in the debate over Roe v. Wade is not a metaphysical question about the concept of personhood or a theological question about whether a fetus has a soul, but a legal question about the correct interpretation of the Constitution which in our political system must be settled one way or the other judicially, by the Supreme Court, rather than politically. It is the question whether the fetus is a constitutional person, that is, a person whose rights and interests must be ranked equally important with those of other people in the scheme of individual rights the Constitution establishes. That is a complex and difficult question, and it does involve moral issues. But it is nevertheless different from the metaphysical question philosophers and theologians debate; it is entirely consistent to think, for example, that a fetus is just as much a human being as an adult, or that it has a soul from the moment of conception, and yet that the Constitution, on the best interpretation, does not grant a fetus rights competitive with the rights it grants other people.

Courts cannot avoid deciding the legal question whether a fetus is a constitutional person because it makes no sense to consider what constitutional rights some people do or do not have, in any area of constitutional law, without first deciding who else has rights a state must or may also recognize. The Supreme Court has held, for example, that the citizens of each state have a constitutional right that state elections be conducted under districting arrangements that ensure one person one vote, and a state could not undermine that principle by counting as people whole classes of entities that the Constitution, properly interpreted, does not. A state could not declare corporations persons, for instance, by providing separate votes for them, and cut down the voting power of real people. The question of whether and in what sense corporations are constitutional persons, with rights of their own, has been much debated throughout constitutional history. But it has never been doubted that because that question affects the rights of everyone else, it must be decided judicially, at the national constitutional level. Of course a state may promote the interests of its corporations in a wide variety of ways. But it cannot endow them with rights whose force is to curtail the constitutional rights enjoyed by others. Only the Constitution can do that.6

So the question of who is a constitutional person must be settled at the constitutional level, by the Supreme Court, as part of deciding what constitutional rights anyone has, and the question whether a fetus is a constitutional person is pivotal to the abortion debate. In Roe v. Wade the Court decided that a fetus is not a constitutional person before birth, and though its opinion has been criticized by several academic lawyers, it is largely persuasive once that premise is accepted. Earlier Supreme Court decisions had established that a person has a fundamental constitutional right to control his or her own role in procreation—the Court had decided, for example, that for this reason a state may not prohibit the sale of contraceptives. If a fetus is not a constitutional person, then a fetus’s right to live cannot be cited as a justification for denying that right after pregnancy begins, though of course a state can nevertheless protect the fetus’s interests in a great variety of other ways.

But if the fetus is a constitutional person then Roe v. Wade is plainly wrong, as the Court’s opinion in that case conceded. The Fourteenth Amendment declares that no state may deny any person “equal protection of the laws.” If the fetus is protected by that clause, then of course a state is entitled to protect its life in the same way it protects the lives of other people under its care, and for that reason is entitled to say that a woman’s right to control the use of her body for procreation ends, at least when her health is not at stake, when pregnancy begins. Indeed it would be difficult to resist a very much stronger conclusion: that a state is not only entitled but required to take that view, so that states like New York, which decided to permit abortion in early pregnancy even before Roe v. Wade was decided, would be constitutionally prohibited from doing so.


The equal protection clause requires states to extend the protection of their laws against murder and assault equally to all persons, and if fetuses were constitutional persons any state legislation that discriminated against them in that respect, by permitting abortion, would be “suspect,” under equal protection principles, and the Supreme Court would have an obligation to review such legislation to determine whether the state’s justification for that discrimination was “compelling.” In some cases it would be: when a state permitted abortion to protect the health of a mother, for example, or perhaps in cases of rape or incest. But if a woman is well aware of the physical and emotional consequences of pregnancy and voluntarily has sexual intercourse knowing that she risks becoming pregnant, a state that permits her or her doctor to abort her fetus has no compelling justification for doing so if the fetus is entitled to equal protection of the laws. For a state fails to show equal concern for both mother and fetus when it allows the mother to regain the freedom of her body at the expense of the fetus’s life.

It is true, as a number of legal scholars have pointed out, that the law does not generally require people to make any sacrifice at all to save the life of another person who needs their aid. A person ordinarily has no legal duty to save a stranger from drowning even if he can do so at no risk to himself and with minimal effort.7 But abortion normally requires a physical attack on a fetus, not just the failure to come to its aid. And in any case parents are invariably made an exception to the general doctrine under which people are not required to save others. Parents have a legal duty to care for their children, and if a fetus is a constitutional person from conception a state would not be justified in discriminating between fetuses and infants. If it did not permit killing infants or abandoning them in circumstances in which they would inevitably die, it could not permit abortion either.8 The physical and emotional and economic burdens of pregnancy are intense, of course, but so are the parallel burdens of parenthood.

I stress this point because it is important to notice that those who urge the Supreme Court to leave the question of abortion to the states, to decide as their politics dictate, have in effect conceded that a fetus is not a constitutional person. In oral argument, Justice White asked Charles Fried whether in his view there is “some problem about the state permitting abortion.” Fried replied, “Oh, no,” and said very firmly that it would be a serious mistake for the Court to “constitutionalize” the issue at any “point in the spectrum” by requiring constitutional scrutiny of permissive abortion legislation by a state. That position is preposterous except on the assumption that the Constitution itself offers a fetus’s life no protection at all.9

But Fried could hardly have given White’s question any other answer. It would be political madness for the Court to try to force unwilling states to outlaw abortion; and neither the government nor any other responsible group has asked it to do so. The damage to the community, to the Court’s authority, and to the Constitution would be far greater if it did try to force the states to outlaw abortion, than if it simply left the law where it stands. But the Court can avoid that inconceivable decision, legitimately, only by confirming Roe v. Wade’s explicit decision that a fetus is not a constitutional person. So the most complex and difficult of the legal issues in the abortion dispute has been removed from the controversy by a kind of practical necessity. I do not mean to suggest, by emphasizing these practical arguments for that view, that it is not the correct view in law. On the contrary, I think that it is.

The question is one of legal interpretation. The principle that the fetus is not a constitutional person fits better with other parts of our law and also with our sense of how related issues would and should be decided if they arose than the rival principle that it is. Even if the fetus is a human being, it is in a unique situation politically as well as biologically for a reason that could properly be thought sufficient to deny it constitutional status. The state can take action that affects it, in order to protect or advance its interests, only through its mother, and only through means that would necessarily restrict her freedom in ways no man or other woman’s freedom could constitutionally be limited: by dictating her diet and other personal and intimate behavior, for example. Apart from anti-abortion statutes, there are few signs in our law of the kind of regulation of pregnancy that would be appropriate if the fetus were a constitutional person, and the Supreme Court has never suggested any constitutional requirement of such protection.

The best historical evidence shows, moreover, that even anti-abortion laws, which were not prevalent in the United States before the middle of the nineteenth century, were adopted to protect the health of the mother and the privileges of the medical profession, not out of any recognition of a fetus’s rights.10 Even states that had the most stringent anti-abortion laws before Roe v. Wade, moreover, did not punish abortion as severely as murder, as they should have done if they thought a fetus a constitutional person. Nor did they try to outlaw or penalize a woman procuring an abortion in another state or abroad.

So the better interpretation of our constitutional law and practice holds that a fetus is not a constitutional person. That conclusion could be accepted, as I suggested, even by someone who thinks abortion a heinous sin: not every sin is or could be punished by law. But it will of course be easier to accept for someone who believes that a human being has no moral right to life until it has developed self-consciousness as a being whose life extends over time.11 On the assumption that this condition is not reached until some time after birth the interpretive conclusion, that a human being becomes a constitutional person no earlier than at that point, seems even sounder.


It is therefore not an acceptable argument, against the claim that women have a constitutional right to choose an abortion in early pregnancy, that the fetus is a constitutional person whose competing right to live would overcome any such right. I have already mentioned the argument for the claim that women do have a constitutional right of that character. In a series of previous cases relating to sterilization, marriage, and contraception, the Supreme Court recognized that all citizens have a general right, based in the Fourteenth Amendment’s guarantee of due process of law, to decide for themselves ethical and personal issues arising from marriage and procreation.12 Justice Blackmun relied heavily on these previous decisions, which have come to be called “privacy” decisions, in his opinion for the Court in Roe v. Wade. He argued that though abortion raises questions different from those raised by these other issues, the general principle that people have a right to control their own role in procreation plainly applied to abortion as well.13 Neither Missouri nor the Bush administration has argued that these precedents should be overruled.14 Fried, in oral argument, said that the case of Griswold v. Connecticut, which upheld a right of contraception, was correct and should not be disturbed. He said that Roe v. Wade could be overruled without affecting that case or the other privacy precedents, that Roe v. Wade could be pulled from the fabric woven by these past decisions like a single thread.

Abortion cannot be disentangled from contraception even medically, however, because both the IUD and the most popular and safest birth-control pills act as abortifacients, that is, they destroy fertilized ova. So the Court could not hold that a woman’s right to control her role in procreation ends with fertilization without permitting states to outlaw the contraceptives now in use. That would be in effect to overrule Griswold, which Fried said was a correct decision. Even if contraception and abortion did not overlap medically in that way, they could not be distinguished in principle, once it is assumed that a fetus is not a constitutional person.

The Court’s previous privacy decisions can be justified only on the assumption that decisions affecting marriage and childbirth are so important, so intimate and personal, so crucial to the development of personality and sense of moral responsibility, and so closely tied to religious and ethical convictions protected by the First Amendment, that people must be allowed to make these decisions for themselves, consulting their own conscience, rather than allowing society to thrust its collective decision on them. The abortion decision is at least as much a private decision in that sense as any other the Court has protected. In many ways it is more private, because the decision involves a woman’s control not just of her connections to others, but of the use of her own body, and the Constitution recognizes in a variety of ways the special intimacy of a person’s connection to her own physical integrity.15

If a fetus were a constitutional person, then abortion could of course be distinguished from at least contraception that did not involve abortifacients, because a state could properly cite a compelling interest in protecting the fetus’s right to life and to be treated with equal concern. But given the assumption that a fetus is not a constitutional person, that reason for distinguishing abortion from contraception, and from the other activities permitted by decisions protecting privacy, fails. Fried tried to distinguish the contraception cases on the ground that Griswold v. Connecticut was based not on any general right to control one’s own procreation, but on the different basis that the police could enforce a prohibition on the use of contraceptives only by searching the marital bedroom, which would be offensive. It is true that one opinion in Griswold v. Connecticut mentioned that reason for invalidating a prohibition on married couples using contraceptives. But it is a silly reason, not only because prohibitions on the use of contraceptives could be enforced without breaking down bedroom doors, but because the Court has upheld other criminal statutes that might be thought just as difficult to enforce without offensive and impermissible searches.16

In any case, the later contraception cases rejected that interpretation of Griswold v. Connecticut, and they are inconsistent with it. In Eisenstadt v. Baird, Justice Brennan, for the Court, stated the point of the past privacy cases this way:

If the right of privacy means anything, it is the right of the individual, married or not, to be free from government intrusion into matters so fundamentally affecting a person as the decision whether to bear or beget a child.

And one of the justices who dissented in the Griswold case, Potter Stewart, joined the majority in Roe v. Wade on the ground that if one accepts the Griswold decision, as he then did on grounds of precedent, one has to accept Roe v. Wade as well. Fried’s claim that the privacy decisions were really only about searching bedrooms proved too bizarre for him to defend with any confidence. When Justice O’Connor asked a direct question, “Do you say there is no fundamental right to decide whether to have a child or not?” he could only answer, “I would hesitate to formulate the right in such abstract terms.”

So the argument from precedent in favor of Roe v. Wade seems a strong one: Supreme Court precedents established a constitutional right of control over one’s own role in childbirth, and, if a fetus is not a constitutional person, that right naturally extends to abortion. But we must now consider the opposing arguments made by those lawyers, including the justices dissenting in that case, who insist it was wrong and should now be discarded. They say that the right to abortion is “judge-made” and has “little or no cognizable roots in the language or design of the Constitution.”17 Or that the right has “no moorings in the text of our Constitution or in familiar constitutional doctrine,” and cannot be sustained by “the interpretive tradition of the legal community.”18 Or that the right does not exist because the subject of abortion is “one upon which the Constitution is silent.”19

But these various complaints beg the question. Of course, if the judges who decided Roe v. Wade made up the constitutional rights they announced, or if those rights have no roots in the language or design of the Constitution, or if they cannot be established as drawn from the Constitution by interpretive methods traditional to legal reasoning, then the decision was certainly wrong. But we cannot decide whether these complaints are justified without some theory of how judges should interpret the abstract provisions of the Constitution, such as the provision that requires due process of law. How should judges decide which rights do and which do not have “roots” in the abstract language?

The various government briefs in Webster sometimes suggest an answer to that question which our legal tradition has decisively rejected: that abstract language should never be interpreted to yield a right that the historical framers who enacted the abstract provision did not accept themselves. The briefs argue that the Fourteenth Amendment cannot be thought to include a right to abortion because abortion laws were being enacted by states throughout the country when that amendment was added to the Constitution.20 But the Congress that enacted the Fourteenth Amendment itself segregated the public schools of the District of Columbia, and no one now argues that Brown v. Board of Education, which held that segregation violated the rights provided by that amendment, was wrong.

The briefs of the Bush administration and the state of Missouri also rely on a variety of other interpretative suggestions. They propose that the Constitution should be understood to contain only “enumerated” rights, that is, rights explicitly mentioned in the text. But that ignores the fact that the same legal situation can be described in different ways. The Supreme Court decided, in 1952, that the police may not pump out a suspect’s stomach for evidence. Shall we say that the Court decided that the right to due process of law, which is mentioned in the text of the Constitution, applied to the particular facts of that case? Or that it decided that people have a right not to have their stomachs pumped, which is derived from the due process clause but which is not itself mentioned in the text? There is only a verbal difference between the two formulations and neither is more accurate than the other.

In any case, if we must reject the right to an abortion because abortion is not mentioned in the Constitution, then we must also reject a great number of other, unquestioned constitutional rights that lawyers frequently describe in language not to be found there either. These include the right to use contraceptives, which the government now argues is part of the Constitution in spite of the fact that contraception is not mentioned. They also include the right to vote, to marry, to travel between states, to live with one’s extended family, to educate one’s children privately in schools meeting educational standards, and to attend racially desegregated schools. If these are all “unenumerated” rights, and so “judge-made” constitutional law, it hardly counts against Roe v. Wade that it falls into the same category.

One of the government briefs replies to that objection with a metaphor. It says that the supposed right of abortion “travels further from its point of departure in the text” than these other rights. But how do we measure the distance between a right and the constitutional language from which it is drawn? How can we tell whether the distance between abortion and the constitutional language of due process is greater than the distance between contraception or stomach pumping and that language? Or the distance between the other “unenumerated” rights I listed and the constitutional language in which these were rooted?

Our legal tradition gives a very different, less metaphorical and superficial, answer to the question how abstract constitutional provisions should be interpreted. Judges should seek to identify the principles latent in the Constitution as a whole, and in past judicial decisions applying the Constitution’s abstract language, in order to enforce the same principles in new areas and so make the law steadily more coherent. In that way, the principles that have been relied on to justify rights for one group or in one situation are extended, so far as that is possible, to everyone else to whom they equally apply. That common-law process was used in Roe v. Wade to argue that the principles latent in the earlier privacy decisions about sterilization and family and contraception must be applied to the abortion case as well. These earlier privacy decisions can themselves be defended in a similar way, as part of a broader project of the Court, begun earlier in the century, to identify and enforce the principles implicit in what the Court called “the concept of ordered liberty,” which means the principle a society truly committed to individual liberty and dignity must recognize. A right to control one’s part in procreation finds support in that general project, as well as in the more closely related decisions protecting privacy, because that right is crucially important to the moral, social, and economic freedom of women.

These are the arguments that the opponents of Roe v. Wade must meet, and they should try to meet them in the traditional way, by explaining why different principles from those mentioned, which do not yield a right to abortion, provide a more satisfactory interpretation of the Constitution as a whole and the Court’s past decisions under it. Of course different judges will come to very different conclusions about which principles provide the best interpretation of the Constitution, and since there is no neutral standpoint from which it can be proved which side is right, each justice must in the end rely on his or her convictions about which argument is best. But that is an inevitable feature of a political system like ours, which conceives of its constitution as a charter of principle rather than a particular collection of political settlements.

Certainly the present critics of Roe v. Wade offer no alternative. Since their question-begging rhetoric about “judge-made law” and “new rights” rests on no reasoned intellectual basis, it provides even less discipline than the traditional interpretive method, because the latter does demand coherent and extended argument, not just name-calling. The question-begging rhetoric, on the contrary, leaves lawyers free to accept constitutional rights now popular in the community, such as the right to legally integrated education and to use contraceptives, and to oppose rights politically more troublesome, such as the right to abortion, without having to explain what the difference between the constitutional standing of these rights actually is.


Though Roe v. Wade held that women have a right in principle to control their part in procreation, it added that states have a legitimate interest in protecting “potential life,” and that any statement of a woman’s constitutional right to an abortion must take that interest into account. It decided that the state’s interest becomes compelling enough in late pregnancy, when the fetus has become viable, to permit the state to regulate or prohibit abortions after that point, except as necessary to protect the mother’s health. Unfortunately, the Court did not satisfactorily explain what kind of interest a state is permitted to take in “potential life,” or why its concern grows stronger or more legitimate after a fetus becomes viable.

The Court did not mean, of course, that a state has a legitimate interest in increasing the birthrate, because that interest would apply with equal strength at all times in pregnancy and, indeed, would justify a state’s opposing contraception as vigorously as abortion. Nor did the Court mean that a state may legitimately decide that a being with potential life has rights of its own which the state may take an interest in protecting. As we saw, the Court rightly held that the question whether a fetus is a constitutional person, and thus a person whose rights are competitive with the constitutional rights of others, must be settled at the constitutional level, not by state legislation, and it then held that the fetus is not such a person. What else could a state’s interest in “potential life” mean?

The most persuasive answer, which takes the Court’s subsequent decisions into account, is, I believe, the following. Even though a fetus is not a constitutional person, it is nevertheless an entity of considerable moral and emotional significance in our culture, and a state may recognize and try to protect that significance in ways that fall short of any substantial abridgment of a woman’s constitutional right over the use of her own body. A state might properly fear the impact of widespread abortion on its citizens’ instinctive respect for the value of human life and their instinctive horror at human destruction or suffering, which are values essential for the maintenance of a just and decently civil society. A political community in which abortion became commonplace and a matter of ethical indifference, like appendectomy, would certainly be a more callous and insensitive community, and it might be a more dangerous one as well.

A state’s concern for the moral significance of a fetus increases as pregnancy advances, and it is particularly intense after viability when the fetus has assumed a postnatal baby’s form. This is a matter of resemblance.21 People’s instinctive respect for life is unlikely to be lessened significantly if they come to regard the abortion of a just-fertilized ovum as permissible, any more than it is lessened when they accept contraception. But the assault on instinctive values is likely to be almost as devastating when a nearly full-term baby is aborted as when a week-old child is killed.

So the state’s concern is greatest after the point at which a fetus, under present technology, is viable, and a prohibition on elective abortion after that time will not significantly burden or compromise a woman’s constitutional right.22 Her right is a right to make fundamental decisions for herself, and that right is satisfied when she has had ample time after discovering her pregnancy to consider whether she wishes to continue it, and to arrange a safe and convenient abortion if she does not.23 Roe v. Wade, understood in that way, did not balance a woman’s rights against the competing rights of a fetus or of anyone else. Rather it identified a scheme of regulation that could meet a state’s most powerful needs without substantially compromising a woman’s rights at all.

The Court had to pick a particular event or period of pregnancy in constructing that scheme in order to make it clear enough to be administered by officials and judges.24 If the Court had said simply that a state must allow a woman “ample” or “reasonable” time after the discovery of pregnancy to decide about abortion, it would have faced a succession of test cases provoked by state legislatures defining the cutoff line earlier and earlier, so that it would eventually have been forced to draw a line in any case. The Court’s decision to make the crucial event viability, which occurs at approximately twenty-three or twenty-four weeks, has much to recommend it. Viability marks a distinct stage of pregnancy after which the difference between a fetus and a premature infant is not a matter of development but only of environment. Since viability follows “quickening,” or the point at which a pregnant woman feels movement in her womb, it is late enough to provide her a reasonable opportunity for an abortion after pregnancy is discovered. (Teen-age women, particularly, may easily be unaware of pregnancy before quickening; their periods may have been erratic or missing before pregnancy and they may not “show,” or look pregnant, before then.)

Some critics feared that advances in medical technology would make fetuses viable much earlier, and thus require the Court to change its standard; in an earlier case Justice O’Connor said that Roe v. Wade was for that reason on “a collision course” with itself. But a consensus of medical opinion now declares that fear unfounded: there is, according to the brief filed in the Webster case by the American Medical Association and other medical groups, an “anatomical threshhold for fetal survival of about twenty-three to twenty-four weeks of gestation…because the fetal lung does not mature sufficiently to permit normal or even mechanically-assisted respiration before [that time].”

An established Supreme Court decision, particularly one that recognizes individual constitutional rights, should not be overruled unless it is clearly wrong or has proved thoroughly unworkable.25 Roe v. Wade is not wrong, and it certainly is not clearly wrong. Justice Blackmun’s opinion might have been clearer in some respects, and the Court might have chosen an event in pregnancy other than viability but which occurs at roughly the same time, such as neocortical functioning, to mark the point at which abortion might be prohibited.26 But these are hardly reasons to tear apart constitutional law by overturning the decision now. The Court should refuse to nourish the cynical view, already popular among its critics, that constitutional law is only a matter of which president appointed the last few justices.


If the Court declines to overrule or substantially restrict Roe v. Wade, as it should, it must decide the more limited constitutional issues raised by the Webster case. As I said, the lower courts declared unconstitutional a variety of clauses in Missouri’s statute. The state does not now contest some of these rulings, and urges implausible but benign interpretations of others in order to save them from unconstitutionality. The important remaining controversy concerns the state’s ban on the use of public facilities in connection with abortion even when the abortion is performed by a private doctor and paid for by private funds.

The statute defines public facilities very broadly as “any public institution, public facility, public equipment, or any physical asset owned, leased, or controlled by this state or any agency or political subdivision thereof.” So it would forbid abortion in the Truman Medical Center in Kansas City—where 97 percent of all hospital abortions at sixteen weeks or later in Missouri were performed in 1985—in spite of the fact that the center is a private hospital staffed mainly by private doctors, and administered by private corporations, just because that hospital is located on ground leased from a political subdivision of the state.27

Missouri defends the provision by appealing to earlier decisions of the Supreme Court. In Maber v. Roe,28 the Court sustained a state’s right to provide medical assistance funds for childbirth but not for abortion, and in Poelker v. Doe29 it allowed a state to provide childbirth but not abortion facilities in a public city hospital. The Court said that although a state may not forbid abortions, it need not go into the abortion business itself. It might constitutionally adopt a preference for childbirth to abortion, and provide funds only for the former.

The decisions in the Maber and Poelker cases have been criticized because they permit states to take action to discourage people from exercising their constitutional rights. But even if we accept these decisions as sound they do not support Missouri’s broad prohibition. Of course a state need not subsidize or support the exercise of every constitutional right, and it may pursue policies of its own choice in the benefits it awards. It may without violating anyone’s rights to free speech publish literature encouraging conservation while refusing to distribute other political material.

But Missouri’s argument overlooks a crucial distinction. It is one thing for a state to decline to participate in some act it disapproves in circumstances in which it would itself be the author of the act, or would plausibly be taken to be, if it did. A state, for example, may refuse to distribute political criticism of its own government without violating anyone’s rights to free speech. It is quite another thing for a state to use its economic power or control of crucial resources to discourage citizens from exercising their constitutional rights when there is no question of the state being seen as the author of, or as in any way supporting, what they do. A city cannot force news-stands in shopping centers built on public land to sell only papers it approves. It cannot force theaters it supplies with water and power and police protection to perform only plays it likes.

Perhaps a state that itself pays for abortions, or provides them in free public hospitals, will in effect have declared itself neutral between abortion and childbirth, or will be understood to have done so. For the state is necessarily the author of its own public funding and public medical provision. But it is preposterous that a state should be understood as itself performing abortions carried out by private doctors on their own initiative and paid for with private funds, just because the hospital in which this is done is in other ways state-supported, or because it is on land the state, as it happens, owns.

The true explanation of why Missouri adopted its stringent prohibition is not, of course, that it wants to avoid declaring itself neutral about abortion, but that it wants to make abortion as difficult and as expensive as possible, in order to discourage its residents from exercising their constitutional rights. It enacts whatever measures to that end its officials can devise and the federal courts have not yet condemned, including measures so obviously unconstitutional that its lawyers do not seriously defend them when they are challenged. That is impermissible: a state must not declare war on its own people because it is angry that the law is on their side.

Unhappily, if the Court in any way now signals itself more ready to accept constraints on abortion than it has been in the past, that dismal spectacle will continue. Other states will adopt more and more restrictive statutes to provoke more and more test cases to see how far the Court will actually go. Charles Fried anticipated exactly that at the close of his oral argument. He asked the justices, even if they did not overrule Roe v. Wade, at least not to say anything “that would further entrench this decision as a secure premise for reasoning in future cases.” The justices would do best for constitutional order and decorum, as well as principle, if they refused to take that bad advice.

This Issue

June 29, 1989