The bad decision that started it all: Griswold at 40.

FORTY years ago, in Griswold v. Connecticut, the Supreme Court of the United States struck down state laws forbidding the sale, distribution, and use of contraceptives on the basis of a novel constitutional doctrine known as the "right to marital privacy."

At the time, the decision appeared to be harmless. After all, Griswold simply allowed married couples to decide whether to use contraceptives. But the Supreme Court soon transformed the "right to privacy" (the reference to marriage quickly disappeared) into a powerful tool for making public policy. In Eisenstadt v. Baird (1972), the Court changed a right of spouses--justified in Griswold precisely by reference to the importance of marriage--into a right of unmarried adults to buy and use contraceptives. Then, in a move that plunged the United States into a "culture war," the Court ruled in Roe v. Wade and Doe v. Bolton (1973) that this generalized "right to privacy" also encompassed a woman's virtually unrestricted right to have an abortion.

No one doubts that there are true privacy rights in the Constitution, especially in the Fourth Amendment, which protects against unreasonable searches and seizures and ensures that warrants issue only upon a showing of probable cause that a crime has been committed. (Indeed, these rights prevented any kind of aggressive enforcement of the laws struck down in Griswold.) But the justices in Griswold produced a non-text-based and generalized right. "Privacy" functioned as a euphemism for immunity from those public-morals laws deemed by the justices to reflect benighted moral views.

The privacy decisions that sprang from Griswold have been widely criticized, and in the last 20 years there have been two notable efforts to silence and stigmatize that criticism. The first occurred in 1987, when a coalition of liberal interest groups helped to scotch Judge Robert Bork's nomination to the Supreme Court, partly because of Bork's misgivings about this novel doctrine. The second occurred in 1992, when the Supreme Court decided Planned Parenthood v. Casey, which reaffirmed the "central holding" of Roe v. Wade.

Neither of these efforts succeeded. To this day, millions of Americans cannot accept Roe v. Wade as constitutionally legitimate. And thanks to recent developments, public suspicion of the Court's "privacy" doctrine is now greater than ever.

Two years ago, in Lawrence v. Texas, the Supreme Court pushed the doctrine into new territory by overruling Bowers v. Hardwick (1986), a decision that had upheld a state's authority to prohibit homosexual sodomy. But in Lawrence, Justice Anthony Kennedy provocatively remarked that Bowers was wrong the day it was decided. Criticism of the ruling in Lawrence intensified a few months later when the supreme judicial court of Massachusetts promulgated a right to same-sex marriage in that state. In Goodridge v. Department of Public Health (2003), the court cited Lawrence to support this newly minted right. It evidently mattered little to these judges that the majority opinion in Lawrence expressly denied that the case involved the issue of marriage.

As the courts push the "privacy" doctrine further and further, public criticism keeps pace. Griswold, however, has received little attention. Even harsh critics of Roe and Lawrence are loath to say that Griswold was wrongly decided. Most of those who worry about the judicial abuse of the right to privacy do not want or expect the Supreme Court to revisit the case. Yet the cogency of any serious critique of "privacy" may depend on the willingness to reexamine the roots of the doctrine in Griswold.

Consider abortion. Conceding the correctness of Griswold gives a huge advantage to the defenders of Roe and Casey. They benefit because so many influential jurists and scholars say that the "inner logic" of the contraception cases must yield something like Roe. Outsiders may regard this argument with skepticism, but its purpose is clear: It tries to smooth the road from Griswold to Eisenstadt to Roe--and beyond.

But one point is rarely mentioned. Even though Griswold was less consequential than Roe, the two cases suffer from similar flaws. The many shortcomings of Griswold are less well known, because the case is enveloped in myths.

In American law schools, decisions such as Roe, Casey, and Lawrence are widely praised--not because of their legal merits (which are dubious), but because they comport with the ideology of "lifestyle liberalism" that enjoys hegemony there. Consequently, since 1973 most legal scholars have had no incentive to reassess Griswold. But if Griswold was wrongly decided, Roe--intellectually shaky on any account--loses even the meager jurisprudential support on which it rests.

The lack of scholarly engagement with Griswold partly explains the myths now surrounding it. Exposing those myths further undermines the arguments for a generalized right to privacy.

Myth #1: The Connecticut laws were "purposeless restraints," serving no social interest.

Supreme Court justice David Souter is one of several jurists to make this assertion. The confusion arises from Griswold, whose majority opinion nowhere identifies a legislative purpose.

For anyone who cares to look, the purposes of the laws are apparent in the record of the case: Connecticut sought to promote marital fidelity and stable families by discouraging attempts to avoid the possible consequences of non-marital sexual relations through the use of contraceptives. Prominent judges in Connecticut recognized the legitimacy of these purposes, and the state's supreme court upheld the laws against several constitutional challenges from 1940 to 1964.

Did Connecticut's policy go too far in its efforts to promote marital fidelity? Many thought so. But roughly 30 states regulated contraceptives in the early 1960s, and the uniqueness of Connecticut's statutory scheme was long recognized as its constitutional prerogative.

Myth #2: The decision in Griswold rested on some overarching or time-honored constitutional principle.

Ostensibly, that principle was "privacy." But the Griswold doctrine would have been unrecognizable to the Supreme Court even a few years earlier. In Gardner v. Massachusetts (1938), for example, the Court dismissed a similar challenge, noting that the suit failed to present "a substantial federal question."

In the majority opinion in Griswold, Justice William O. Douglas referred--as comically metaphysical as it sounds--to "penumbras formed by emanations" of specific constitutional guarantees as the source of the new right. He had nothing else to go on.

Other jurists have since argued that the right to marital privacy could be derived from cases before 1965 involving the rights of parents to direct the upbringing of their children. But the cases they cite have little in common with Griswold.

What, then, was the operative "principle" in Griswold? Nothing other than the Court's desire to place its imprimatur on "enlightened" views about human sexuality. This project continued beyond Griswold and culminated in Lawrence, where the Court essentially said that all adults in America have a right to engage in consenting, non-marital sexual relations. Consistently missing from the Court's discourse on privacy, however, has been any discussion of parental duties, public health, and the welfare of children.

Myth #3: No sensible jurist or commentator would say that the case was wrongly decided.

In fact, two widely respected and sensible jurists, Justices Hugo Black and Potter Stewart, dissented in Griswold. Black was a noted liberal and, like Stewart, recorded his opposition to Connecticut's policy as a political matter. Yet both jurists insisted that the policy was a valid exercise of the state's power to promote public health, safety, and morals.

To Justices Black and Stewart, the "right to privacy" cloaked a naked policy preference. Justices in the majority were, without constitutional warrant, substituting their own judgments for those of the elected representatives in Connecticut. This, according to jurists across the political spectrum, is precisely what had brought shame on the Court during the "Lochner era," from roughly 1890 to 1937, when in the name of an unwritten "liberty of contract" the justices invalidated state social-welfare and worker-protection laws. But the crucial distinction underscored by Black and Stewart between the desirability or justice of a policy and the state's constitutional authority to enact it lost much of its currency as the right to privacy expanded.

Myth #4: The legislation invalidated in Griswold might be widely used again if the case was overturned.

This line was often repeated in 1987 when Robert Bork was nominated to the Supreme Court. Meant to frighten ordinary citizens who approve of contraceptive use, this scenario simply fails to acknowledge changes in public opinion since 1965. Laws like those struck down in Griswold clearly have little chance of passing today even in the most conservative states.

Myth #5: The widespread use of contraceptives in the United States today provides a post hoc justification for Griswold.

When Griswold was decided, adults could buy and use contraceptives in almost every state (despite various regulations on their sale and distribution). Given the social ferment of the 1960s and '70s, the Connecticut policy would sooner or later have been modified. But the ubiquity of contraceptives in America today does not justify Griswold--any more than the widespread use of abortion justifies Roe.

It might seem fanciful to say that the idea of a generalized constitutional right to "privacy" could now be repudiated; many believe that it has become an integral part of American law. But no one should accept this conclusion. The struggle against usurpations by the Supreme Court committed under the pretext of giving effect to unwritten constitutional rights has a historical precedent. As noted, from roughly 1890 to 1937, the Supreme Court invalidated worker-protection and social-welfare legislation on the basis of an unenumerated right to "liberty of contract." After much criticism, the Court relented and in 1937 announced that it would defer to legislative judgment where policies did not run afoul of constitutional principles. They promised, in short, to halt the practice of reading into the Constitution their own personal judgments about social and economic policy and the morality of economic relations.

The Supreme Court will not revisit the question of state or federal laws banning contraceptives. Yet the Court can and should find an occasion to admit that the manipulation of constitutional law that began with Griswold has been a colossal mistake. Such an admission would hardly be radical or, as we have observed, unprecedented. The Court's confession of error in repudiating its Griswold jurisprudence, far from harming its reputation, would enhance its prestige. We have no doubt that the same good effect would redound to the Court if the justices were candidly to speak the truth: The idea of a generalized right to privacy floating in penumbras formed by emanations was a pure judicial invention--one designed to license the judicial usurpation of democratic legislative authority.

Mr. George is a professor of jurisprudence and director of the James Madison Program in American Ideals and Institutions at Princeton. Mr. Tubbs is a fellow at the Witherspoon Institute. In the fall, he will be an assistant professor of political science at King's College in New York City.