133*133 Mr. Matthew Hale Carpenter, for the plaintiff in error.

No opposing counsel.

137*137 Mr. Justice MILLER delivered the opinion of the court.

The record in this case is not very perfect, but it may be fairly taken that the plaintiff asserted her right to a license on the grounds, among others, that she was a citizen of the United States, and that having been a citizen of Vermont at one time, she was, in the State of Illinois, entitled to any right granted to citizens of the latter State.

The court having overruled these claims of right founded on the clauses of the Federal Constitution before referred 138*138 to, those propositions may be considered as properly before this court.

As regards the provision of the Constitution that citizens of each State shall be entitled to all the privileges and immunities of citizens in the several States, the plaintiff in her affidavit has stated very clearly a case to which it is inapplicable.

The protection designed by that clause, as has been repeatedly held, has no application to a citizen of the State whose laws are complained of. If the plaintiff was a citizen of the State of Illinois, that provision of the Constitution gave her no protection against its courts or its legislation.

The plaintiff seems to have seen this difficulty, and attempts to avoid it by stating that she was born in Vermont.

While she remained in Vermont that circumstance made her a citizen of that State. But she states, at the same time, that she is a citizen of the United States, and that she is now, and has been for many years past, a resident of Chicago, in the State of Illinois.

The fourteenth amendment declares that citizens of the United States are citizens of the State within which they reside; therefore the plaintiff was, at the time of making her application, a citizen of the United States and a citizen of the State of Illinois.

We do not here mean to say that there may not be a temporary residence in one State, with intent to return to another, which will not create citizenship in the former. But the plaintiff states nothing to take her case out of the definition of citizenship of a State as defined by the first section of the fourteenth amendment.

In regard to that amendment counsel for the plaintiff in this court truly says that there are certain privileges and immunities which belong to a citizen of the United States as such; otherwise it would be nonsense for the fourteenth amendment to prohibit a State from abridging them, and he proceeds to argue that admission to the bar of a State of a person who possesses the requisite learning and character is one of those which a State may not deny.

139*139 In this latter proposition we are not able to concur with counsel. We agree with him that there are privileges and immunities belonging to citizens of the United States, in that relation and character, and that it is these and these alone which a State is forbidden to abridge. But the right to admission to practice in the courts of a State is not one of them. This right in no sense depends on citizenship of the United States. It has not, as far as we know, ever been made in any State, or in any case, to depend on citizenship at all. Certainly many prominent and distinguished lawyers have been admitted to practice, both in the State and Federal courts, who were not citizens of the United States or of any State. But, on whatever basis this right may be placed, so far as it can have any relation to citizenship at all, it would seem that, as to the courts of a State, it would relate to citizenship of the State, and as to Federal courts, it would relate to citizenship of the United States.

The opinion just delivered in the Slaughter-House Cases[*] renders elaborate argument in the present case unnecessary; for, unless we are wholly and radically mistaken in the principles on which those cases are decided, the right to control and regulate the granting of license to practice law in the courts of a State is one of those powers which are not transferred for its protection to the Federal government, and its exercise is in no manner governed or controlled by citizenship of the United States in the party seeking such license.

It is unnecessary to repeat the argument on which the judgment in those cases is founded. It is sufficient to say they are conclusive of the present case.

JUDGMENT AFFIRMED.

Mr. Justice BRADLEY:

I concur in the judgment of the court in this case, by which the judgment of the Supreme Court of Illinois is affirmed, but not for the reasons specified in the opinion just read.

140*140 The claim of the plaintiff, who is a married woman, to be admitted to practice as an attorney and counsellor-at-law, is based upon the supposed right of every person, man or woman, to engage in any lawful employment for a livelihood. The Supreme Court of Illinois denied the application on the ground that, by the common law, which is the basis of the laws of Illinois, only men were admitted to the bar, and the legislature had not made any change in this respect, but had simply provided that no person should be admitted to practice as attorney or counsellor without having previously obtained a license for that purpose from two justices of the Supreme Court, and that no person should receive a license without first obtaining a certificate from the court of some county of his good moral character. In other respects it was left to the discretion of the court to establish the rules by which admission to the profession should be determined. The court, however, regarded itself as bound by at least two limitations. One was that it should establish such terms of admission as would promote the proper administration of justice, and the other that it should not admit any persons, or class of persons, not intended by the legislature to be admitted, even though not expressly excluded by statute. In view of this latter limitation the court felt compelled to deny the application of females to be admitted as members of the bar. Being contrary to the rules of the common law and the usages of Westminster Hall from time immemorial, it could not be supposed that the legislature had intended to adopt any different rule.

The claim that, under the fourteenth amendment of the Constitution, which declares that no State shall make or enforce any law which shall abridge the privileges and immunities of citizens of the United States, the statute law of Illinois, or the common law prevailing in that State, can no longer be set up as a barrier against the right of females to pursue any lawful employment for a livelihood (the practice of law included), assumes that it is one of the privileges and immunities of women as citizens to engage in any and every profession, occupation, or employment in civil life.

141*141 It certainly cannot be affirmed, as an historical fact, that this has ever been established as one of the fundamental privileges and immunities of the sex. On the contrary, the civil law, as well as nature herself, has always recognized a wide difference in the respective spheres and destinies of man and woman. Man is, or should be, woman’s protector and defender. The natural and proper timidity and delicacy which belongs to the female sex evidently unfits it for many of the occupations of civil life. The constitution of the family organization, which is founded in the divine ordinance, as well as in the nature of things, indicates the domestic sphere as that which properly belongs to the domain and functions of womanhood. The harmony, not to say identity, of interests and views which belong, or should belong, to the family institution is repugnant to the idea of a woman adopting a distinct and independent career from that of her husband. So firmly fixed was this sentiment in the founders of the common law that it became a maxim of that system of jurisprudence that a woman had no legal existence separate from her husband, who was regarded as her head and representative in the social state; and, notwithstanding some recent modifications of this civil status, many of the special rules of law flowing from and dependent upon this cardinal principle still exist in full force in most States. One of these is, that a married woman is incapable, without her husband’s consent, of making contracts which shall be binding on her or him. This very incapacity was one circumstance which the Supreme Court of Illinois deemed important in rendering a married, woman incompetent fully to perform the duties and trusts that belong to the office of an attorney and counsellor.

It is true that many women are unmarried and not affected by any of the duties, complications, and incapacities arising out of the married state, but these are exceptions to the general rule. The paramount destiny and mission of woman are to fulfil the noble and benign offices of wife and mother. This is the law of the Creator. And the rules of civil society 142*142 must be adapted to the general constitution of things, and cannot be based upon exceptional cases.

The humane movements of modern society, which have for their object the multiplication of avenues for woman’s advancement, and of occupations adapted to her condition and sex, have my heartiest concurrence. But I am not prepared to say that it is one of her fundamental rights and privileges to be admitted into every office and position, including those which require highly special qualifications and demanding special responsibilities. In the nature of things it is not every citizen of every age, sex, and condition that is qualified for every calling and position. It is the prerogative of the legislator to prescribe regulations founded on nature, reason, and experience for the due admission of qualified persons to professions and callings demanding special skill and confidence. This fairly belongs to the police power of the State; and, in my opinion, in view of the peculiar characteristics, destiny, and mission of woman, it is within the province of the legislature to ordain what offices, positions, and callings shall be filled and discharged by men, and shall receive the benefit of those energies and responsibilities, and that decision and firmness which are presumed to predominate in the sterner sex.

For these reasons I think that the laws of Illinois now complained of are not obnoxious to the charge of abridging any of the privileges and immunities of citizens of the United States.

Mr. Justice SWAYNE and Mr. Justice FIELD concurred in the foregoing opinion of Mr. Justice BRADLEY.

The CHIEF JUSTICE dissented from the judgment of the court, and from all the opinions.

[*] Supra, p. 36.

  Rehearing Denied Feb. 22, 1977. See 429 U.S. 1124, 97 S.Ct. 1161.

          Appellant Craig, a male then between 18 and 21 years old, and appellant Whitener, a licensed vendor of 3.2% beer, brought this action for declaratory and injunctive relief, claiming that an Oklahoma statutory scheme prohibiting the sale of "nonintoxicating" 3.2% beer to males under the age of 21 and to females under the age of 18 constituted a gender-based discrimination that denied to males 18-20 years of age the equal protection of the laws. Recognizing that Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225, and later cases establish that classification by gender must substantially further important governmental objectives, a three-judge District Court held that appellees' statistical evidence regarding young males' drunk-driving arrests and traffic injuries demonstrated that the gender-based discrimination was substantially related to the achievement of traffic safety on Oklahoma roads. Held :

          1. Since only declaratory and injunctive relief against enforcement of the gender-based differential was sought, the controversy has been mooted as to Craig, who became 21 after this Court had noted probable jurisdiction. See, e. g., DeFunis v. Odegaard, 416 U.S. 312, 94 S.Ct. 1704, 40 L.Ed.2d 164. P. 192.

          2. Whitener has standing to make the equal protection challenge. Pp. 192-197.

          (a) No prudential objective thought to be served by limitations of jus tertii standing can be furthered here, where the lower court already has entertained the constitutional challenge and the parties have sought resolution of the constitutional issue. Pp. 193-194.

          (b) Whitener in any event independently has established third-party standing. She suffers "injury in fact" since the challenged statutory provisions are addressed to vendors like her, who either must obey the statutory provisions and incur economic injury or disobey the statute and suffer sanctions. In such circumstances, vendors may resist efforts to restrict their operations by advocating the rights of third parties seeking access to their market. See, e. g., Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349. Pp. 194-197.

          3. Oklahoma's gender-based differential constitutes an invidious discrimination against males 18-20 years of age in violation of the Equal Protection Clause. Appellees' statistics (the most relevant of which

Page 191

show only that .18% of females and 2% of males in the 18-20-year-old age group were arrested for driving while under the influence of liquor) do not warrant the conclusion that sex represents an accurate proxy for the regulation of drinking and driving. Pp. 199-204.

          4. The operation of the Twenty-first Amendment does not alter the application of equal protection standards that otherwise govern this case. The Court has never recognized that application of that Amendment can defeat an otherwise established claim under the Equal Protection Clause, the principles of which cannot be rendered inapplicable here by reliance upon statistically measured but loose-fitting generalities concerning the drinking tendencies of aggregate groups. Pp. 204-210.

          399 F.Supp. 1304, reversed.

          Frederick P. Gilbert, Tulsa, Okl., for appellants.

          James H. Gray, Oklahoma City, Okl., for appellees.

           Mr. Justice BRENNAN delivered the opinion of the Court.

          The interaction of two sections of an Oklahoma statute, Okla.Stat., Tit. 37, §§ 241 and 245 (1958 and Supp.1976),¹

Page 192

prohibits the sale of "nonintoxicating" 3.2% beer to males under the age of 21 and to females under the age of 18. The question to be decided is whether such a gender-based differential constitutes a denial to males 18-20 years of age of the equal protection of the laws in violation of the Fourteenth Amendment.

          This action was brought in the District Court for the Western District of Oklahoma on December 20, 1972, by appellant Craig, a male then between 18 and 21 years of age, and by appellant Whitener, a licensed vendor of 3.2% beer. The complaint sought declaratory and injunctive relief against enforcement of the gender-based differential on the ground that it constituted invidious discrimination against males 18-20 years of age. A three-judge court convened under 28 U.S.C. § 2281 sustained the constitutionality of the statutory differential and dismissed the action. 399 F.Supp. 1304 (1975). We noted probable jurisdiction of appellants' appeal, 423 U.S. 1047, 96 S.Ct. 771, 46 L.Ed.2d 635 (1976). We reverse.

          We first address a preliminary question of standing. Appellant Craig attained the age of 21 after we noted probable jurisdiction. Therefore, since only declaratory and injunctive relief against enforcement of the gender-based differential is sought, the controversy has been rendered moot as to Craig. See, e. g., DeFunis v. Odegaard, 416 U.S. 312, 94 S.Ct. 1704, 40 L.Ed.2d 164 (1974).² The question thus arises whether appellant Whitener, the licensed vendor of 3.2% beer, who has a live controversy against enforcement of the statute, may rely upon the equal protection objections of males 18-20 years of age to establish her claim of

Page 193

unconstitutionality of the age-sex differential. We conclude that she may.

          Initially, it should be noted that, despite having had the opportunity to do so,³ appellees never raised before the District Court any objection to Whitener's reliance upon the claimed unequal treatment of 18-20-year-old males as the premise of her equal protection challenge to Oklahoma's 3.2% beer law. See 399 F.Supp., at 1306 n. 1. Indeed, at oral argument Oklahoma acknowledged that appellees always "presumed" that the vendor, subject to sanctions and loss of license for violation of the statute, was a proper party in interest to object to the enforcement of the sex-based regulatory provision. Tr. of Oral Arg. 41. While such a concession certainly would not be controlling upon the reach of this Court's constitutional authority to exercise jurisdiction under Art. III, see, e. g., Sierra Club v. Morton, 405 U.S. 727, 732 n. 3, 92 S.Ct. 1361, 1365, 31 L.Ed.2d 636 (1972); cf. Association of Data Processing Service Organizations, Inc. v. Camp, 397 U.S. 150, 151, 90 S.Ct. 827, 829, 25 L.Ed.2d 184 (1970), our decisions have settled that limitations on a litigant's assertion of jus tertii are not constitutionally mandated, but rather stem from a salutary "rule of self-restraint" designed to minimize unwarranted intervention into controversies where the applicable constitutional questions are ill-defined and speculative. See, e. g., Barrows v. Jackson, 346 U.S. 249, 255, 257, 73 S.Ct. 1031, 1034, 1035, 97 L.Ed. 1586 (1953); see also Singleton v. Wulff, 428 U.S. 106, 123-124, 96 S.Ct. 2868, 2877, 49 L.Ed.2d 826 (1976) (Powell, J., dissenting). These prudential objectives, thought to be enhanced by restrictions on third-party standing, cannot be furthered here, where the lower court already has entertained the relevant constitutional challenge and the parties have sought or at least have never resisted an authoritative constitutional determination. In such circumstances, a decision by us to forgo

Page 194

consideration of the constitutional merits in order to await the initiation of a new challenge to the statute by injured third parties would be impermissibly to foster repetitive and time-consuming litigation under the guise of caution and prudence. Moreover, insofar as the applicable constitutional questions have been and continue to be presented vigorously and "cogently," Holden v. Hardy, 169 U.S. 366, 397, 18 S.Ct. 383, 390, 42 L.Ed. 780 (1898), the denial of jus tertii standing in deference to a direct class suit can serve no functional purpose. Our Brother Blackmun's comment is pertinent: "(I)t may be that a class could be assembled, whose fluid membership always included some (males) with live claims. But if the assertion of the right is to be 'representative' to such an extent anyway, there seems little loss in terms of effective advocacy from allowing its assertion by" the present jus tertii champion. Singleton v. Wulff, supra, at 117-118, 96 S.Ct. at 2875.

          In any event, we conclude that appellant Whitener has established independently her claim to assert jus tertii standing. The operation of §§ 241 and 245 plainly has inflicted "injury in fact" upon appellant sufficient to guarantee her "concrete adverseness," Baker v. Carr, 369 U.S. 186, 204, 82 S.Ct. 691, 703, 7 L.Ed.2d 663 (1962), and to satisfy the constitutionally based standing requirements imposed by Art. III. The legal duties created by the statutory sections under challenge are addressed directly to vendors such as appellant. She is obliged either to heed the statutory discrimination, thereby incurring a direct economic injury through the constriction of her buyers' market, or to disobey the statutory command and suffer, in the words of Oklahoma's Assistant Attorney General, "sanctions and perhaps loss of license." Tr. of Oral Arg. 41. This Court repeatedly has recognized that such injuries establish the threshold requirements of a "case or controversy" mandated by Art. III. See, e. g., Singleton v. Wulff, supra, at 113, 96 S.Ct., at 2873 (doctors who receive payments for their abortion services are "classically adverse" to government as payer); Sullivan v. Little Hunting

Page 195

Park, 396 U.S. 229, 237, 90 S.Ct. 400, 404, 24 L.Ed.2d 386 (1969); Barrows v. Jackson, supra, 346 U.S., at 255-256, 73 S.Ct., at 1034-1035.

          As a vendor with standing to challenge the lawfulness of §§ 241 and 245, appellant Whitener is entitled to assert those concomitant rights of third parties that would be "diluted or adversely affected" should her constitutional challenge fail and the statutes remain in force. Griswold v. Connecticut, 381 U.S. 479, 481, 85 S.Ct. 1678, 1679, 14 L.Ed.2d 510 (1965); see Note, Standing to Assert Constitutional Jus Tertii, 88 Harv.L.Rev. 423, 432 (1974). Otherwise, the threatened imposition of governmental sanctions might deter appellant Whitener and other similarly situated vendors from selling 3.2% beer to young males, thereby ensuring that "enforcement of the challenged restriction against the (vendor) would result indirectly in the violation of third parties' rights." Warth v. Seldin, 422 U.S. 490, 510, 95 S.Ct. 2197, 2211, 45 L.Ed.2d 343 (1975). Accordingly, vendors and those in like positions have been uniformly permitted to resist efforts at restricting their operations by acting as advocates of the rights of third parties who seek access to their market or function. See, e. g., Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349 (1972); Sullivan v. Little Hunting Park, supra ; Barrows v. Jackson, supra.⁴

Page 196

          Indeed, the jus tertii question raised here is answered by our disposition of a like argument in Eisenstadt v. Baird, supra. There, as here, a state statute imposed legal duties and disabilities upon the claimant, who was convicted of distributing a package of contraceptive foam to a third party.⁵ Since the statute was directed at Baird and penalized his conduct, the Court did not hesitate again as here to conclude that the "case or controversy" requirement of Art. III was satisfied. 405 U.S., at 443, 92 S.Ct., at 1033. In considering Baird's constitutional objections, the Court fully recognized his standing to defend the privacy interests of third parties. Deemed crucial to the decision to permit jus tertii standing was the recognition of "the impact of the litigation on the third-party interests." Id., at 445, 92 S.Ct., at 1034. Just as the defeat of Baird's suit and the "(e)nforcement of the Massachusetts statute will materially impair the ability of single persons to obtain contraceptives," id., at 446, 92 S.Ct., at 1034, so too the failure of Whitener to prevail in this suit and the continued enforcement of §§ 241 and 245 will "materially impair the ability of" males 18-20 years of age to purchase 3.2% beer despite their classification by an overt gender-based criterion. Similarly, just as the Massachusetts law in Eisenstadt "prohibit(ed),

Page 197

not use, but distribution," 405 U.S., at 446, 92 S.Ct., at 1034, and consequently the least awkward challenger was one in Baird's position who was subject to that proscription, the law challenged here explicitly regulates the sale rather than use of 3.2% beer, thus leaving a vendor as the obvious claimant.

          We therefore hold that Whitener has standing to raise relevant equal protection challenges to Oklahoma's gender-based law. We now consider those arguments.

          Before 1972, Oklahoma defined the commencement of civil majority at age 18 for females and age 21 for males. Okla.Stat., Tit. 15, § 13 (1972 and Supp.1976). In contrast, females were held criminally responsible as adults at age 18 and males at age 16. Okla.Stat., Tit. 10, § 1101(a) (Supp.1976). After the Court of Appeals for the Tenth Circuit held in 1972, on the authority of Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed. 225 (1971), that the age distinction was unconstitutional for purposes of establishing criminal responsibility as adults, Lamb v. Brown, 456 F.2d 18, the Oklahoma Legislature fixed age 18 as applicable to both males and females. Okla.Stat., Tit. 10, § 1101(a) (Supp.1976). In 1972, 18 also was established as the age of majority for males and females in civil matters, Okla.Stat., Tit. 15, § 13 (1972 and Supp.1976), except that §§ 241 and 245 of the 3.2% beer statute were simultaneously codified to create an exception to the gender-free rule.

          Analysis may appropriately begin with the reminder that Reed emphasized that statutory classifications that distinguish between males and females are "subject to scrutiny under the Equal Protection Clause." 404 U.S., at 75, 92 S.Ct., at 253. To withstand constitutional challenge, previous cases establish that classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives. Thus, in Reed, the ob-

Page 198

jectives of "reducing the workload on probate courts," id., at 76, 92 S.Ct., at 254, and "avoiding intrafamily controversy," id., at 77, 92 S.Ct., at 254, were deemed of insufficient importance to sustain use of an overt gender criterion in the appointment of administrators of intestate decedents' estates. Decisions following Reed similarly have rejected administrative ease and convenience as sufficiently important objectives to justify gender-based classifications. See, e. g., Stanley v. Illinois, 405 U.S. 645, 656, 92 S.Ct. 1208, 1215, 31 L.Ed.2d 551 (1972); Frontiero v. Richardson, 411 U.S. 677, 690, 93 S.Ct. 1764, 1772, 36 L.Ed.2d 583 (1973); cf. Schlesinger v. Ballard, 419 U.S. 498, 506-507, 95 S.Ct. 572, 576-577, 42 L.Ed.2d 610 (1975). And only two Terms ago, Stanton v. Stanton, 421 U.S. 7, 95 S.Ct. 1373, 43 L.Ed.2d 688 (1975), expressly stating that Reed v. Reed was "controlling," 421 U.S., at 13, 95 S.Ct., at 1377, held that Reed required invalidation of a Utah differential age-of-majority statute, notwithstanding the statute's coincidence with and furtherance of the State's purpose of fostering "old notions" of role typing and preparing boys for their expected performance in the economic and political worlds. 421 U.S., at 14-15, 95 S.Ct., at 1378.⁶

          Reed v. Reed has also provided the underpinning for decisions that have invalidated statutes employing gender as an inaccurate proxy for other, more germane bases of classification. Hence, "archaic and overbroad" generalizations, Schlesinger v. Ballard, supra, 419 U.S., at 508, 95 S.Ct., at 577, concerning the financial position of servicewomen, Frontiero v. Richardson, supra, 411 U.S., at 689 n. 23, 93 S.Ct., at 1772, and working women, Weinberger v. Wiesenfeld, 420 U.S. 636, 643, 95 S.Ct. 1225, 1230, 43 L.Ed.2d 514 (1975), could not justify use of a gender line in determining eligibility for certain governmental entitlements. Similarly, increasingly outdated

Page 199

misconceptions concerning the role of females in the home rather than in the "marketplace and world of ideas" were rejected as loose-fitting characterizations incapable of supporting state statutory schemes that were premised upon their accuracy. Stanton v. Stanton, supra ; Taylor v. Louisiana, 419 U.S. 522, 535 n. 17, 95 S.Ct. 692, 700, 42 L.Ed.2d 690 (1975). In light of the weak congruence between gender and the characteristic or trait that gender purported to represent, it was necessary that the legislatures choose either to realign their substantive laws in a gender-neutral fashion, or to adopt procedures for identifying those instances where the sex-centered generalization actually comported with fact. See, e. g., Stanley v. Illinois, supra, 405 U.S., at 658, 92 S.Ct., at 1216; cf. Cleveland Board of Education v. LaFleur, 414 U.S. 632, 650, 94 S.Ct. 791, 801, 39 L.Ed.2d 52 (1974).

          In this case, too, "Reed, we feel is controlling . . .," Stanton v. Stanton, supra, 421 U.S., at 13, 95 S.Ct., at 1377. We turn then to the question whether, under Reed, the difference between males and females with respect to the purchase of 3.2% beer warrants the differential in age drawn by the Oklahoma statute. We conclude that it does not.

          The District Court recognized that Reed v. Reed was controlling. In applying the teachings of that case, the court found the requisite important governmental objective in the traffic-safety goal proffered by the Oklahoma Attorney General. It then concluded that the statistics introduced by the appellees established that the gender-based distinction was substantially related to achievement of that goal.

          We accept for purposes of discussion the District Court's identification of the objective underlying §§ 241 and 245 as the enhancement of traffic safety. ⁷ Clearly, the protection

Page 200

of public health and safety represents an important function of state and local governments. However, appellees' statistics in our view cannot support the conclusion that the gender-based distinction closely serves to achieve that objective and therefore the distinction cannot under Reed withstand equal protection challenge.

          The appellees introduced a variety of statistical surveys. First, an analysis of arrest statistics for 1973 demonstrated that 18-20-year-old male arrests for "driving under the influence" and "drunkenness" substantially exceeded female arrests for that same age period.⁸ Similarly, youths aged 17-21 were found to be overrepresented among those killed

Page 201

or injured in traffic accidents, with males again numerically exceeding females in this regard.⁹ Third, a random roadside survey in Oklahoma City revealed that young males were more inclined to drive and drink beer than were their female counterparts. ¹⁰ Fourth, Federal Bureau of Investigation nationwide statistics exhibited a notable increase in arrests for "driving under the influence." ¹¹ Finally, statistical evidence gathered in other jurisdictions, particularly Minnesota and Michigan, was offered to corroborate Oklahoma's experience by indicating the pervasiveness of youthful participation in motor vehicle accidents following the imbibing of alcohol. Conceding that "the case is not free from doubt," 399 F.Supp., at 1314, the District Court nonetheless concluded that this statistical showing substantiated "a rational basis for the legislative judgment underlying the challenged classification." Id., at 1307.

          Even were this statistical evidence accepted as accurate, it nevertheless offers only a weak answer to the equal protection question presented here. The most focused and relevant of the statistical surveys, arrests of 18-20-year-olds for alcohol-related driving offenses, exemplifies the ultimate unpersuasiveness of this evidentiary record. Viewed in terms of the correlation between sex and the actual activity that Oklahoma seeks to regulate driving while under the influence of alcohol the statistics broadly establish that .18% of females and 2% of males in that age group were arrested for that offense. While such a disparity is not trivial in a statistical sense, it hardly can form the basis for employment of a gender line as a classifying device. Certainly if male-

Page 202

ness is to serve as a proxy for drinking and driving, a correlation of 2% must be considered an unduly tenuous "fit." ¹² Indeed, prior cases have consistently rejected the use of sex as a decisionmaking factor even though the statutes in question certainly rested on far more predictive empirical relationships than this.¹³

          Moreover, the statistics exhibit a variety of other shortcomings that seriously impugn their value to equal protection analysis. Setting aside the obvious methodological problems,¹⁴ the surveys do not adequately justify the salient

Page 203

features of Oklahoma's gender-based traffic-safety law. None purports to measure the use and dangerousness of 3.2% beer as opposed to alcohol generally, a detail that is of particular importance since, in light of its low alcohol level, Oklahoma apparently considers the 3.2% beverage to be "nonintoxicating." Okla.Stat., Tit. 37, § 163.1 (1958); see State ex rel. Springer v. Bliss, 199 Okl. 198, 185 P.2d 220 (1947). Moreover, many of the studies, while graphically documenting the unfortunate increase in driving while under the influence of alcohol, make no effort to relate their findings to age-sex differentials as involved here.¹⁵ Indeed, the only survey that explicitly centered its attention upon young drivers and their use of beer albeit apparently not of the diluted 3.2% variety reached results that hardly can be viewed as impressive in justifying either a gender or age classification.¹⁶

Page 204

          There is no reason to belabor this line of analysis. It is unrealistic to expect either members of the judiciary or state officials to be well versed in the rigors of experimental or statistical technique. But this merely illustrates that proving broad sociological propositions by statistics is a dubious business, and one that inevitably is in tension with the normative philosophy that underlies the Equal Protection Clause.¹⁷ Suffice to say that the showing offered by the appellees does not satisfy us that sex represents a legitimate, accurate proxy for the regulation of drinking and driving. In fact, when it is further recognized that Oklahoma's statute prohibits only the selling of 3.2% beer to young males and not their drinking the beverage once acquired (even after purchase by their 18-20-year-old female companions), the relationship between gender and traffic safety becomes far too tenuous to satisfy Reed's requirement that the gender-based difference be substantially related to achievement of the statutory objective.

          We hold, therefore, that under Reed, Oklahoma's 3.2% beer statute invidiously discriminates against males 18-20 years of age.

          Appellees argue, however, that §§ 241 and 245 enforce state policies concerning the sale and distribution of alcohol and by force of the Twenty-first Amendment should therefore be held to withstand the equal protection challenge. The District Court's response to this contention is unclear. The court assumed that the Twenty-first Amendment "strengthened" the State's police powers with respect to alcohol regulation, 399 F.Supp. at 1307, but then said that "the standards of review that (the Equal Protection Clause) mandates are not relaxed." Id., at 1308. Our view is, and we hold, that the Twenty-first Amendment does not save the

Page 205

invidious gender-based discrimination from invalidation as a denial of equal protection of the laws in violation of the Fourteenth Amendment.

          The history of state regulation of alcoholic beverages dates from long before adoption of the Eighteenth Amendment. In the License Cases, 5 How. 504, 579, 12 L.Ed. 256 (1847), the Court recognized a broad authority in state governments to regulate the trade of alcoholic beverages within their borders free from implied restrictions under the Commerce Clause. Later in the century, however, Leisy v. Hardin, 135 U.S. 100, 10 S.Ct. 681, 34 L.Ed. 128 (1890), undercut the theoretical underpinnings of License Cases. This led Congress, acting pursuant to its powers under the Commerce Clause, to reinvigorate the State's regulatory role through the passage of the Wilson ¹⁸ and Webb-Kenyon Acts.¹⁹ See, e. g., Clark Distilling Co. v. Western Maryland R. Co., 242 U.S. 311, 37 S.Ct. 180, 61 L.Ed. 326 (1917) (upholding Webb-Kenyon Act); In re Rahrer, 140 U.S. 545, 11 S.Ct. 865, 35 L.Ed. 572 (1891) (upholding Wilson Act). With passage of the Eighteenth Amendment, the uneasy tension between the Commerce Clause and state police power temporarily subsided.

          The Twenty-first Amendment repealed the Eighteenth Amendment in 1933. The wording of § 2 of the Twenty-first Amendment ²⁰ closely follows the Webb-Kenyon and Wil-

Page 206

son Acts, expressing the framers' clear intention of constitutionalizing the Commerce Clause framework established under those statutes. This Court's decisions since have confirmed that the Amendment primarily created an exception to the normal operation of the Commerce Clause. See, e. g., Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 330, 84 S.Ct. 1293, 1297, 12 L.Ed.2d 350 (1964); Carter v. Virginia, 321 U.S. 131, 139-140, 64 S.Ct. 464, 469-470, 88 L.Ed. 605 (1944) (Frankfurter, J., concurring); Finch & Co. v. McKittrick, 305 U.S. 395, 398, 59 S.Ct. 256, 257, 83 L.Ed. 246 (1939). Even here, however, the Twenty-first Amendment does not pro tanto repeal the Commerce Clause, but merely requires that each provision "be considered in the light of the other, and in the context of the issues and interests at stake in any concrete case." Hostetter v. Idlewild Bon Voyage Liquor Corp., supra, 377 U.S., at 332, 84 S.Ct., at 1298; cf. Department of Revenue v. James Beam Distilling Co., 377 U.S. 341, 84 S.Ct. 1247, 12 L.Ed.2d 362 (1964); Collins v. Yosemite Park & Curry Co., 304 U.S. 518, 58 S.Ct. 1009, 82 L.Ed. 1502 (1938).

          Once passing beyond consideration of the Commerce Clause, the relevance of the Twenty-first Amendment to other constitutional provisions becomes increasingly doubtful. As one commentator has remarked: "Neither the text nor the history of the Twenty-first Amendment suggests that it qualifies individual rights protected by the Bill of Rights and the Fourteenth Amendment where the sale or use of liquor is concerned." P. Brest, Processes of Constitutional Decisionmaking, Cases and Materials 258 (1975). Any departures from this historical view have been limited and sporadic. Two States successfully relied upon the Twenty-first Amendment to respond to challenges of major liquor importers to state authority to regulate the importation and manufacture of alcoholic beverages on Commerce Clause and Fourteenth Amendment grounds. See Mahoney v. Joseph Triner Corp., 304 U.S. 401, 58 S.Ct. 952, 82 L.Ed. 1424 (1938); State Board of Equalization of California v. Young's Market Co.,

Page 207

299 U.S. 59, 64, 57 S.Ct. 77, 79, 81 L.Ed. 38 (1936). In fact, however, the arguments in both cases centered upon importation of intoxicants, a regulatory area where the State's authority under the Twenty-first Amendment is transparently clear, Hostetter v. Idlewild Bon Voyage Liquor Corp., supra, 377 U.S., at 330, and n.9, 84 S.Ct., at 1297, and touched upon purely economic matters that traditionally merit only the mildest review under the Fourteenth Amendment, see, e. g., Joseph E. Seagram & Sons v. Hostetter, 384 U.S. 35, 47-48, 50-51, 86 S.Ct. 1254, 1262-1263, 1263-1264, 16 L.Ed.2d 336 (1966) (rejecting Fourteenth Amendment objections to state liquor laws on the strength of Ferguson v. Skrupa, 372 U.S. 726, 729-730, 83 S.Ct. 1028, 1030-1031, 10 L.Ed.2d 93 (1963) and Williamson v. Lee Optical Co., 348 U.S. 483, 75 S.Ct. 461, 99 L.Ed. 563 (1955) ).²¹ Cases involving individual rights protected by the Due Process Clause have been treated in sharp contrast. For example, when an individual objected to the mandatory "posting" of her name in retail liquor establishments and her characterization as an "excessive drink(er)," the Twenty-first Amendment was held not to qualify the scope of her due process rights. Wisconsin v. Constantineau, 400 U.S. 433, 436, 91 S.Ct. 507, 509, 27 L.Ed.2d 515 (1971).

          It is true that California v. LaRue, 409 U.S. 109, 115, 93 S.Ct. 390, 395, 34 L.Ed.2d 342 (1972), relied upon the Twenty-first Amendment to "strengthen" the State's authority to regulate live entertainment at establishments licensed to dispense liquor, at least when the performances "partake more of gross sexuality than of communication," id., at 118, 93 S.Ct., at 397. Nevertheless, the Court has never recognized sufficient "strength" in the Amendment to defeat an otherwise established claim of invidious discrimination in violation of the Equal Protection Clause.

Page 208

Rather, Moose Lodge No. 107 v. Irvis, 407 U.S. 163, 178-179, 92 S.Ct. 1965, 1974, 32 L.Ed.2d 627 (1972), establishes that state liquor regulatory schemes cannot work invidious discriminations that violate the Equal Protection Clause.

          Following this approach, both federal and state courts uniformly have declared the unconstitutionality of gender lines that restrain the activities of customers of state-regulated liquor establishments irrespective of the operation of the Twenty-first Amendment. See, e. g., White v. Fleming, 522 F.2d 730 (CA7 1975); Women's Liberation Union of R.I. v. Israel, 512 F.2d 106 (CA1 1975); Daugherty v. Daley, 370 F.Supp. 338 (ND Ill.1974) (three-judge court); Seidenberg v. McSorleys' Old Ale House, Inc., 317 F.Supp. 593 (SDNY 1970); Commonwealth Alcoholic Beverage Control Bd. v. Burke, 481 S.W.2d 52 (Ky.1972); cf. Sail'er Inn, Inc. v. Kirby, 5 Cal.3d 1, 95 Cal.Rptr. 329, 485 P.2d 529 (1971); Paterson Tavern & G. O. A. v. Hawthorne, 57 N.J. 180, 270 A.2d 628 (1970). Even when state officials have posited sociological or empirical justifications for these gender-based differentiations, the courts have struck down discriminations aimed at an entire class under the guise of alcohol regulation. In fact, social science studies that have uncovered quantifiable differences in drinking tendencies dividing along both racial and ethnic lines strongly suggest the need for application of the Equal Protection Clause in preventing discriminatory treatment that almost certainly would be perceived as invidious.²² In sum, the principles embodied in the Equal

Page 209

Protection Clause are not to be rendered inapplicable by statistically measured but loose-fitting generalities concerning the drinking tendencies of aggregate groups. We thus hold that the operation of the Twenty-first Amendment does not alter the application of equal protection standards that otherwise govern this case.

Page 210

          We conclude that the gender-based differential contained in Okla.Stat., Tit. 37, § 245 (1976 Supp.) constitutes a denial of the equal protection of the laws to males aged 18-20 ²³ and reverse the judgment of the District Court. ²⁴

          It is so ordered.

           Mr. Justice POWELL, concurring.

          I join the opinion of the Court as I am in general agreement with it. I do have reservations as to some of the discussion concerning the appropriate standard for equal protection analysis and the relevance of the statistical evidence. Accordingly, I add this concurring statement.

          With respect to the equal protection standard, I agree that Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971), is the most relevant precedent. But I find it unnecessary, in deciding this case, to read that decision as broadly as some of the Court's language may imply. Reed and subsequent cases involving gender-based classifications make clear that the Court subjects such classifications to a more critical examination than is normally applied when "fundamental" constitutional rights and "suspect classes" are not present. ^*

Page 211

          I view this as a relatively easy case. No one questions the legitimacy or importance of the asserted governmental objective: the promotion of highway safety. The decision of the case turns on whether the state legislature, by the classification it has chosen, had adopted a means that bears a " 'fair and substantial relation' " to this objective. Id., at 76, 92 S.Ct., at 254, quoting Royster Guano Co. v. Virginia, 253 U.S. 412, 415, 40 S.Ct. 560, 561, 64 L.Ed. 989 (1920).

          It seems to me that the statistics offered by appellees and relied upon by the District Court do tend generally to support the view that young men drive more, possibly are inclined to drink more, and for various reasons are involved in more accidents than young women. Even so, I am not persuaded that these facts and the inferences fairly drawn from them justify this classification based on a three-year age differential between the sexes, and especially one that it so easily circumvented as to be virtually meaningless. Putting it differently, this gender-based classification does not bear a fair and substantial relation to the object of the legislation.

           Mr. Justice STEVENS, concurring.

          There is only one Equal Protection Clause. It requires every State to govern impartially. It does not direct the

Page 212

courts to apply one standard of review in some cases and a different standard in other cases. Whatever criticism may be leveled at a judicial opinion implying that there are at least three such standards applies with the same force to a double standard.

          I am inclined to believe that what has become known as the two-tiered analysis of equal protection claims does not describe a completely logical method of deciding cases, but rather is a method the Court has employed to explain decisions that actually apply a single standard in a reasonably consistent fashion. I also suspect that a careful explanation of the reasons motivating particular decisions may contribute more to an identification of that standard than an attempt to articulate it in all-encompassing terms. It may therefore be appropriate for me to state the principal reasons which persuaded me to join the Court's opinion.

          In this case, the classification is not as obnoxious as some the Court has condemned,¹ nor as inoffensive as some the Court has accepted. It is objectionable because it is based on an accident of birth,² because it is a mere remnant of the now almost universally rejected tradition of discriminating against males in this age bracket,³ and because, to the extent it reflects any physical difference between males and

Page 213

females, it is actually perverse. ⁴ The question then is whether the traffic safety justification put forward by the State is sufficient to make an otherwise offensive classification acceptable.

          The classification is not totally irrational. For the evidence does indicate that there are more males than females in this age bracket who drive and also more who drink. Nevertheless, there are several reasons why I regard the justification as unacceptable. It is difficult to believe that the statute was actually intended to cope with the problem of traffic safety,⁵ since it has only a minimal effect on access to a not very intoxicating beverage and does not prohibit its consumption.⁶ Moreover, the empirical data submitted by

Page 214

the State accentuate the unfairness of treating all 18-21-year-old males as inferior to their female counterparts. The legislation imposes a restraint on 100% of the males in the class allegedly because about 2% of them have probably violated one or more laws relating to the consumption of alcoholic beverages. ⁷ It is unlikely that this law will have a significant deterrent effect either on that 2% or on the law-abiding 98%. But even assuming some such slight benefit, it does not seem to me that an insult to all of the young men of the State can be justified by visiting the sins of the 2% on the 98%.

           Mr. Justice BLACKMUN, concurring in part.

          I join the Court's opinion except Part II-D thereof. I agree, however, that the Twenty-first Amendment does not save the challenged Oklahoma statute.

           Mr. Justice STEWART, concurring in the judgment.

          I agree that the appellant Whitener has standing to assert the equal protection claims of males between 18 and 21 years old. Eisenstadt v. Baird, 405 U.S. 438, 443-446, 92 S.Ct. 1029, 1033-1035, 31 L.Ed.2d 349; Griswold v. Connecticut, 381 U.S. 479, 481, 85 S.Ct. 1678, 1679, 14 L.Ed.2d 510; Barrows v. Jackson, 346 U.S. 249, 255-260, 73 S.Ct. 1031, 1034-1037, 97 L.Ed. 1586; Buchanan v. Warley, 245 U.S. 60, 72-73, 38 S.Ct. 16, 17-18, 62 L.Ed. 149; see Note, Standing To Assert Constitutional Jus Tertii, 88 Harv.L.Rev. 423, 431-436 (1974). I also concur in the Court's judgment on the merits of the constitutional issue before us.

Page 215

          Every State has broad power under the Twenty-first Amendment to control the dispensation of alcoholic beverages within its borders. E. g., California v. LaRue, 409 U.S. 109, 93 S.Ct. 390, 34 L.Ed.2d 342; Joseph E. Seagram & Sons v. Hostetter, 384 U.S. 35, 86 S.Ct. 1254, 16 L.Ed.2d 336; Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 330, 84 S.Ct. 1293, 1296, 12 L.Ed.2d 350; Mahoney v. Joseph Triner Corp., 304 U.S. 401, 58 S.Ct. 952, 82 L.Ed. 1424; State Board of Equalization v. Young's Market Co., 299 U.S. 59, 57 S.Ct. 77, 81 L.Ed. 38. But "(t)his is not to say that the Twenty-first Amendment empowers a State to act with total irrationality or invidious discrimination in controlling the distribution and dispensation of liquor . . . ." California v. LaRue, supra, 409 U.S., at 120, 93 S.Ct., at 398 n. (concurring opinion).

          The disparity created by these Oklahoma statutes amounts to total irrationality. For the statistics upon which the State now relies, whatever their other shortcomings, wholly fail to prove or even suggest that 3.2% beer is somehow more deleterious when it comes into the hands of a male aged 18-20 than of a female of like age. The disparate statutory treatment of the sexes here, without even a colorably valid justification or explanation, thus amounts to invidious discrimination. See Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225.

           Mr. Chief Justice BURGER, dissenting.

          I am in general agreement with Mr. Justice REHNQUIST's dissent, but even at the risk of compounding the obvious confusion created by those voting to reverse the District Court, I will add a few words.

          At the outset I cannot agree that appellant Whitener has standing arising from her status as a saloonkeeper to assert the constitutional rights of her customers. In this Court "a litigant may only assert his own constitutional rights or immunities." United States v. Raines, 362 U.S. 17, 22, 80 S.Ct. 519, 523, 4 L.Ed.2d 524 (1960). There are a few, but strictly limited exceptions to that rule; despite the most creative efforts, this case fits within none of them.

Page 216

          This is not Sullivan v. Little Hunting Park, 396 U.S. 229, 90 S.Ct. 400, 24 L.Ed.2d 386 (1969), or Barrows v. Jackson, 346 U.S. 249, 73 S.Ct. 1031, 97 L.Ed. 1586 (1953), for these is here no barrier whatever to Oklahoma males 18-20 years of age asserting, in an appropriate forum, any constitutional rights they may claim to purchase 3.2% beer. Craig's successful litigation of this very issue was prevented only by the advent of his 21st birthday. There is thus no danger of interminable dilution of those rights if appellant Whitener is not permitted to litigate them here. Cf. Eisenstadt v. Baird, 405 U.S. 438, 445-446, 92 S.Ct. 1029, 1034-1035, 31 L.Ed.2d 349 (1972).

          Nor is this controlled by Griswold v. Connecticut, 381 U.S. 479, 85 S.Ct. 1678, 14 L.Ed.2d 510 (1965). It borders on the ludicrous to draw a parallel between a vendor of beer and the intimate professional physician-patient relationship which undergirded relaxation of standing rules in that case.

          Even in Eisenstadt, the Court carefully limited its recognition of third-party standing to cases in which the relationship between the claimant and the relevant third party "was not simply the fortuitous connection between a vendor and potential vendees, but the relationship between one who acted to protect the rights of a minority and the minority itself." 405 U.S., at 445, 92 S.Ct., at 1034. This is plainly not the case here. See also McGowan v. Maryland, 366 U.S. 420, 429-430, 81 S.Ct. 1101, 1106-1108, 6 L.Ed.2d 393 (1961); Brown v. United States, 411 U.S. 223, 230, 93 S.Ct. 1565, 1569, 36 L.Ed.2d 208 (1973).

          In sum, permitting a vendor to assert the constitutional rights of vendees whenever those rights are arguably infringed introduces a new concept of constitutional standing to which I cannot subscribe.

          On the merits, we have only recently recognized that our duty is not "to create substantive constitutional rights in the name of guaranteeing equal protection of the laws." San Antonio School Dist. v. Rodriguez, 411 U.S. 1, 33, 93 S.Ct. 1278, 1297, 36 L.Ed.2d 16 (1973). Thus, even interests of such importance in our society as public education and housing do not qualify as "fundamental rights" for equal protection purposes because they have no

Page 217

textually independent constitutional status. See id., at 29-39, 93 S.Ct., at 1294-1300 (education); Lindsey v. Normet, 405 U.S. 56, 92 S.Ct. 862, 31 L.Ed.2d 36 (1972) (housing). Though today's decision does not go so far as to make gender-based classifications "suspect," it makes gender a disfavored classification. Without an independent constitutional basis supporting the right asserted or disfavoring the classification adopted, I can justify no substantive constitutional protection other than the normal McGowan v. Maryland, supra, at 425-426, 81 S.Ct., at 1104-1105, protection afforded by the Equal Protection Clause.

          The means employed by the Oklahoma Legislature to achieve the objectives sought may not be agreeable to some judges, but since eight Members of the Court think the means not irrational, I see no basis for striking down the statute as violative of the Constitution simply because we find it unwise, unneeded, or possibly even a bit foolish.

          With Mr. Justice REHNQUIST, I would affirm the judgment of the District Court.

           Mr. Justice REHNQUIST, dissenting.

          The Court's disposition of this case is objectionable on two grounds. First is its conclusion that men challenging a gender-based statute which treats them less favorably than women may invoke a more stringent standard of judicial review than pertains to most other types of classifications. Second is the Court's enunciation of this standard, without citation to any source, as being that "classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives." Ante at 197 (emphasis added). The only redeeming feature of the Court's opinion, to my mind, is that it apparently signals a retreat by those who joined the plurality opinion in Frontiero v. Richardson, 411 U.S. 677, 93 S.Ct. 1764, 36 L.Ed.2d 583 (1973), from their view that sex is a "suspect" classification for purposes of equal protection analysis. I think the Oklahoma statute challenged here need pass only the "rational basis" equal

Page 218

protection analysis expounded in cases such as McGowan v. Maryland, 366 U.S. 420, 81 S.Ct. 1101, 6 L.Ed.2d 393 (1961), and Williamson v. Lee Optical Co., 348 U.S. 483, 75 S.Ct. 461, 99 L.Ed. 563 (1955), and I believe that it is constitutional under that analysis.

          In Frontiero v. Richardson, supra, the opinion for the plurality sets forth the reasons of four Justices for concluding that sex should be regarded as a suspect classification for purposes of equal protection analysis. These reasons center on our Nation's "long and unfortunate history of sex discrimination," 411 U.S., at 684, 93 S.Ct., at 1769, which has been reflected in a whole range of restrictions on the legal rights of women, not the least of which have concerned the ownership of property and participation in the electoral process. Noting that the pervasive and persistent nature of the discrimination experienced by women is in part the result of their ready identifiability, the plurality rested its invocation of strict scrutiny largely upon the fact that "statutory distinctions between the sexes often have the effect of invidiously relegating the entire class of females to inferior legal status without regard to the actual capabilities of its individual members." Id., at 686-687, 93 S.Ct., at 1770. See Stanton v. Stanton, 421 U.S. 7, 14-15, 95 S.Ct. 1373, 1377-1378, 43 L.Ed.2d 688 (1975).

          Subsequent to Frontiero, the Court has declined to hold that sex is a suspect class, Stanton v. Stanton, supra, at 13, 95 S.Ct., at 1377, and no such holding is imported by the Court's resolution of this case. However, the Court's application here of an elevated or "intermediate" level scrutiny, like that invoked in cases dealing with discrimination against females, raises the question of why the statute here should be treated any differently from countless legislative classifications unrelated to sex which have been upheld under a minimum rationality standard. Jefferson v. Hackney, 406 U.S. 535, 546-547, 92 S.Ct. 1724, 1731-1732, 32 L.Ed.2d 285 (1972); Richardson v. Belcher, 404 U.S. 78, 81-84, 92 S.Ct. 254, 257-259, 30 L.Ed.2d 231 (1971); Dandridge v. Williams, 397 U.S. 471, 484-485, 90 S.Ct. 1153, 1161-1162, 25 L.Ed.2d 491 (1970);

Page 219

McGowan v. Maryland, supra, 366 U.S., at 425-426, 81 S.Ct., at 1104-1105; Flemming v. Nestor, 363 U.S. 603, 611, 80 S.Ct. 1367, 1373, 4 L.Ed.2d 1435 (1960); Williamson v. Lee Optical Co., supra, 348 U.S., at 488-489, 75 S.Ct., at 464-465.

          Most obviously unavailable to support any kind of special scrutiny in this case, is a history or pattern of past discrimination, such as was relied on by the plurality in Frontiero to support its invocation of strict scrutiny. There is no suggestion in the Court's opinion that males in this age group are in any way peculiarly disadvantaged, subject to systematic discriminatory treatment, or otherwise in need of special solicitude from the courts.

          The Court does not discuss the nature of the right involved, and there is no reason to believe that it sees the purchase of 3.2% beer as implicating any important interest, let alone one that is "fundamental" in the constitutional sense of invoking strict scrutiny. Indeed, the Court's accurate observation that the statute affects the selling but not the drinking of 3.2% beer, ante, at 204, further emphasizes the limited effect that it has on even those persons in the age group involved. There is, in sum, nothing about the statutory classification involved here to suggest that it affects an interest, or works against a group, which can claim under the Equal Protection Clause that it is entitled to special judicial protection.

          It is true that a number of our opinions contain broadly phrased dicta implying that the same test should be applied to all classifications based on sex, whether affecting females or males. E. g., Frontiero v. Richardson, supra, 411 U.S., at 688, 93 S.Ct., at 1771; Reed v. Reed, 404 U.S. 71, 76, 92 S.Ct. 251, 254, 30 L.Ed.2d 225 (1971). However, before today, no decision of this Court has applied an elevated level of scrutiny to invalidate a statutory discrimination harmful to males, except where the statute impaired an important personal interest protected by the Constitution.¹ There being no such interest

Page 220

here, and there being no plausible argument that this is a discrimination against females,² the Court's reliance on our previous sex-discrimination cases is ill-founded. It treats gender classification as a talisman which without regard to the rights involved or the persons affected calls into effect a heavier burden of judicial review.

          The Court's conclusion that a law which treats males less favorably than females "must serve important governmental objectives and must be substantially related to achievement of those objectives" apparently comes out of thin air. The Equal Protection Clause contains no such language, and none of our previous cases adopt that standard. I would think we have had enough difficulty with the two standards of review which our cases have recognized the

Page 221

norm of "rational basis," and the "compelling state interest" required where a "suspect classification" is involved so as to counsel weightily against the insertion of still another "standard" between those two. How is this Court to divine what objectives are important? How is it to determine whether a particular law is "substantially" related to the achievement of such objective, rather than related in some other way to its achievement? Both of the phrases used are so diaphanous and elastic as to invite subjective judicial preferences or prejudices relating to particular types of legislation, masquerading as judgments whether such legislation is directed at "important" objectives or, whether the relationship to those objectives is "substantial" enough.

          I would have thought that if this Court were to leave anything to decision by the popularly elected branches of the Government, where no constitutional claim other than that of equal protection is invoked, it would be the decision as to what governmental objectives to be achieved by law are "important," and which are not. As for the second part of the Court's new test, the Judicial Branch is probably in no worse position than the Legislative or Executive Branches to determine if there is any rational relationship between a classification and the purpose which it might be thought to serve. But the introduction of the adverb "substantially" requires courts to make subjective judgments as to operational effects, for which neither their expertise nor their access to data fits them. And even if we manage to avoid both confusion and the mirroring of our own preferences in the development of this new doctrine, the thousands of judges in other courts who must interpret the Equal Protection Clause may not be so fortunate.

          The applicable rational-basis test is one which

          "permits the States a wide scope of discretion in enacting laws which affect some groups of citizens differently than

Page 222

          others. The constitutional safeguard is offended only if the classification rests on grounds wholly irrelevant to the achievement of the State's objective. State legislatures are presumed to have acted within their constitutional power despite the fact that, in practice, their laws result in some inequality. A statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it." McGowan v. Maryland, supra, 366 U.S., at 425-426, 81 S.Ct., at 1105 (citations omitted).

          Our decisions indicate that application of the Equal Protection Clause in a context not justifying an elevated level of scrutiny does nor demand "mathematical nicety" or the elimination of all inequality. Those cases recognize that the practical problems of government may require rough accommodations of interests, and hold that such accommodations should be respected unless no reasonable basis can be found to support them. Dandridge v. Williams, 397 U.S., at 485, 90 S.Ct., at 1161. Whether the same ends might have been better or more precisely served by a different approach is no part of the judicial inquiry under the traditional minimum rationality approach. Richardson v. Belcher, 404 U.S., at 84, 92 S.Ct., at 258.

          The Court "accept(s) for purposes of discussion" the District Court's finding that the purpose of the provisions in question was traffic safety, and proceeds to examine the statistical evidence in the record in order to decide if "the gender-based distinction closely serves to achieve that objective." Ante, at 199, 200 (emphasis added). (Whether there is a difference between laws which "closely serv(e)" objectives and those which are only "substantially related" to their achievement, ante, at 197, we are not told.) I believe that a more traditional type of scrutiny is appropriate in this case, and I think that the Court would have done well here to heed its own warning that "(i)t is unrealistic to expect . . . members of the judiciary . . . to be well versed in the rigors of experimental or statistical technique." Ante, at 204. One

Page 223

need not immerse oneself in the fine points of statistical analysis, however, in order to see the weaknesses in the Court's attempted denigration of the evidence at hand.

          One survey of arrest statistics assembled in 1973 indicated that males in the 18-20 age group were arrested for "driving under the influence" almost 18 times as often as their female counterparts, and for "drunkenness" in a ratio of almost 10 to 1.³ Accepting, as the Court does, appellants' comparison of the total figures with 1973 Oklahoma census data, this survey indicates a 2% arrest rate among males in the age group, as compared to a .18% rate among females.

          Other surveys indicated (1) that over the five-year period from 1967 to 1972, nationwide arrests among those under 18 for drunken driving increased 138%, and that 93% of all persons arrested for drunken driving were male; ⁴ (2) that youths in the 17-21 age group were overrepresented among those killed or injured in Oklahoma traffic accidents, that male casualties substantially exceeded female, and that deaths in this age group continued to rise while overall traffic deaths declined; ⁵ (3) that over three-fourths of the drivers under 20 in the Oklahoma City area are males, and that each of them, on average, drives half again as many miles per year as their female counterparts; ⁶ (4) that four-fifths of male drivers

Page 224

under 20 in the Oklahoma City area state a drink preference for beer, while about three-fifths of female drivers of that age state the same preference; ⁷ and (5) that the percentage of male drivers under 20 admitting to drinking within two hours of driving was half again larger than the percentage for females, and that the percentage of male drivers of that age group with a blood alcohol content greater than .01% was almost half again larger than for female drivers.⁸

          The Court's criticism of the statistics relied on by the District Court conveys the impression that a legislature in enacting a new law is to be subjected to the judicial equivalent of a doctoral examination in statistics. Legislatures are not held to any rules of evidence such as those which may govern courts or other administrative bodies, and are entitled to draw factual conclusions on the basis of the determination of probable cause which an arrest by a police officer normally represents. In this situation, they could reasonably infer that the incidence of drunk driving is a good deal higher than the incidence of arrest.

          And while, as the Court observes, relying on a report to a Presidential Commission which it cites in a footnote, such statistics may be distorted as a result of stereotyping, the legislature is not required to prove before a court that its statistics are perfect. In any event, if stereotypes are as pervasive as the Court suggests, they may in turn influence the conduct of the men and women in question, and cause the young men to conform to the wild and reckless image which is their stereotype.

          The Court also complains of insufficient integration of the various surveys on several counts that the injury and death figures are in no way directly correlated with intoxication, ante, at 201 n. 9; that the national arrest figures for drunk driving contain no breakdown for the 18 to 21-year-old group,

Page 225

ante, at 459, n. 11; and that the arrest records for intoxication are not tied to the consumption of 3.2% beer, ante, at 201-202, nn. 11 and 12. But the State of Oklahoma and certainly this Court for purposes of equal protection review can surely take notice of the fact that drunkenness is a significant cause of traffic casualties, and that youthful offenders have participated in the increase of the drunk-driving problem. On this latter point, the survey data indicating increased driving casualties among 18-21-year-olds, while overall casualties dropped, are not irrelevant.

          Nor is it unreasonable to conclude from the expressed preference for beer by four-fifths of the age-group males that that beverage was a predominant source of their intoxication-related arrests. Taking that as the predicate, the State could reasonably bar those males from any purchases of alcoholic beer, including that of the 3.2% variety. This Court lacks the expertise or the data to evaluate the intoxicating properties of that beverage, and in that posture our only appropriate course is to defer to the reasonable inference supporting the statute that taken in sufficient quantity this beer has the same effect as any alcoholic beverage.

          Quite apart from these alleged methodological deficiencies in the statistical evidence, the Court appears to hold that that evidence, on its face, fails to support the distinction drawn in the statute. The Court notes that only 2% of males (as against .18% of females) in the age group were arrested for drunk driving, and that this very low figure establishes "an unduly tenuous 'fit' " between maleness and drunk driving in the 18 to 20-year-old group. On this point the Court misconceives the nature of the equal protection inquiry.

          The rationality of a statutory classification for equal protection purposes does not depend upon the statistical "fit" between the class and the trait sought to be singled out. It turns on whether there may be a sufficiently higher in-

Page 226

cidence of the trait within the included class than in the excluded class to justify different treatment. Therefore the present equal protection challenge to this gender-based discrimination poses only the question whether the incidence of drunk driving among young men is sufficiently greater than among young women to justify differential treatment. Notwithstanding the Court's critique of the statistical evidence, that evidence suggests clear differences between the drinking and driving habits of young men and women. Those differences are grounds enough for the State reasonably to conclude that young males pose by far the greater drunk-driving hazard, both in terms of sheer numbers and in terms of hazard on a per-driver basis. The gender-based difference in treatment in this case is therefore not irrational.

          The Court's argument that a 2% correlation between maleness and drunk driving is constitutionally insufficient therefore does not pose an equal protection issue concerning discrimination between males and females. The clearest demonstration of this is the fact that the precise argument made by the Court would be equally applicable to a flat bar on such purchases by anyone, male or female, in the 18-20 age group; in fact it would apply a fortiori in that case given the even more "tenuous 'fit' " between drunk-driving arrests and femaleness. The statistics indicate that about 1% of the age group population as a whole is arrested. What the Court's argument is relevant to is not equal protection, but due process whether there are enough persons in the category who drive while drunk to justify a bar against purchases by all members of the group.

          Cast in those terms, the argument carries little weight, in light of our decisions indicating that such questions call for a balance of the State's interest against the harm resulting from any overinclusiveness or underinclusiveness. Vlandis v. Kline, 412 U.S. 441, 448-452, 93 S.Ct. 2230, 2234-2236, 37 L.Ed.2d 63 (1973). The personal interest harmed

Page 227

here is very minor the present legislation implicates only the right to purchase 3.2% beer, certainly a far cry from the important personal interests which have on occasion supported this Court's invalidation of statutes on similar reasoning. Cleveland Board of Education v. LaFleur, 414 U.S. 632, 640, 94 S.Ct. 791, 796, 39 L.Ed.2d 52 (1974); Stanley v. Illinois, 405 U.S. 645, 651, 92 S.Ct. 1208, 1212, 31 L.Ed.2d 551 (1972). And the state interest involved is significant the prevention of injury and death on the highways.

          This is not a case where the classification can only be justified on grounds of administrative convenience. Vlandis v. Kline, supra, 412 U.S., at 451, 93 S.Ct., at 2236; Stanley v. Illinois, supra, 405 U.S., at 656, 92 S.Ct., at 1215. There being no apparent way to single out persons likely to drink and drive, it seems plain that the legislature was faced here with the not atypical legislative problem of legislating in terms of broad categories with regard to the purchase and consumption of alcohol. I trust, especially in light of the Twenty-first Amendment, that there would be no due process violation if no one in this age group were allowed to purchase 3.2% beer. Since males drink and drive at a higher rate than the age group as a whole, I fail to see how a statutory bar with regard only to them can create any due process problem.

          The Oklahoma Legislature could have believed that 18-20-year-old males drive substantially more, and tend more often to be intoxicated than their female counterparts; that they prefer beer and admit to drinking and driving at a higher rate than females; and that they suffer traffic injuries out of proportion to the part they make up of the population. Under the appropriate rational-basis test for equal protection, it is neither irrational nor arbitrary to bar them from making purchases of 3.2% beer, which purchases might in many cases be made by a young man who immediately returns to his vehicle with the beverage in his possession. The record does not give any good indication of the true proportion of males in the age group who drink and drive (ex-

Page 228

cept that it is no doubt greater than the 2% who are arrested), but whatever it may be I cannot see that the mere purchase right involved could conceivably raise a due process question. There being no violation of either equal protection or due process, the statute should accordingly be upheld.

^{1. Sections 241 and 245 provide in pertinent part:}

§ 241. "It shall be unlawful for any person who holds a license to sell and dispense beer . . . to sell, barter or give to any minor any beverage containing more than one-half of one per cent of alcohol measured by volume and not more than three and two-tenths (3.2) per cent of alcohol measured by weight.

§ 245. "A 'minor,' for the purposes of Section . . . 241 . . . is defined as a female under the age of eighteen (18) years, and a male under the age of twenty-one (21) years."

^2. Appellants did not seek class certification of Craig as representative of other similarly situated males 18-20 years of age. See, e. g., Sosna v. Iowa, 419 U.S. 393, 401, 95 S.Ct. 553, 558, 42 L.Ed.2d 532 (1975).

^3. The District Court's opinion confirms that Whitener from the outset has based her constitutional challenge on gender-discrimination grounds, 399 F.Supp., at 1306, and "(n)o challenge is made to (her) standing and requisite interest in the controversy . . . ." Id., at 1306 n. 1.

^4. The standing question presented here is not answered by the principle stated in United States v. Raines, 362 U.S. 17, 21, 80 S.Ct. 519, 522, 4 L.Ed.2d 524 (1960), that "one to whom application of a statute is constitutional will not be heard to attack the statute on the ground that impliedly it might also be taken as applying to other persons or other situations in which its application might be unconstitutional." In Raines, the Court refused to permit certain public officials of Georgia to defend against application of the Civil Rights Act to their official conduct on the ground that the statute also might be construed to encompass the "purely private actions" of others. The Raines rule remains germane in such a setting, where the interests of the litigant and the rights of the proposed third parties are in no way mutually interdependent. Thus, a successful suit against Raines did not threaten to impair or diminish the independent private rights of others, and consequently, consideration of those third-party rights properly was deferred until another day.

Of course, the Raines principle has also been relaxed where legal action against the claimant threatens to "chill" the First Amendment rights of third parties. See, e. g., Lewis v. New Orleans, 415 U.S. 130, 94 S.Ct. 970, 39 L.Ed.2d 214 (1974).

^5. The fact that Baird chose to disobey the legal duty imposed upon him by the Massachusetts, anticontraception statute, resulting in his criminal conviction, 405 U.S., at 440, 92 S.Ct., at 1031, does not distinguish the standing inquiry from that pertaining to the anticipatory attack in this case. In both Eisenstadt and here, the challenged statutes compel jus tertii claimants either to cease their proscribed activities or to suffer appropriate sanctions. The existence of Art. III "injury in fact" and the structure of the claimant's relationship to the third parties are not altered by the litigative posture of the suit. And, certainly, no suggestion will be heard that Whitener's anticipatory challenge offends the normal requirements governing such actions. See generally Steffel v. Thompson, 415 U.S. 452, 94 S.Ct. 1209, 39 L.Ed.2d 505 (1974); Samuels v. Mackell, 401 U.S. 66, 91 S.Ct. 764, 27 L.Ed.2d 688 (1971); Younger v. Harris, 401 U.S. 37, 91 S.Ct. 746, 27 L.Ed.2d 669 (1971).

^6. Kahn v. Shevin, 416 U.S. 351, 94 S.Ct. 1734, 40 L.Ed.2d 189 (1974) and Schlesinger v. Ballard, 419 U.S. 498, 95 S.Ct. 572, 42 L.Ed.2d 610 (1975), upholding the use of gender-based classifications, rested upon the Court's perception of the laudatory purposes of those laws as remedying disadvantageous conditions suffered by women in economic and military life. See 416 U.S., at 353-354, 94 S.Ct. at 1736-1737; 419 U.S., at 508, 95 S.Ct., at 577. Needless to say, in this case Oklahoma does not suggest that the age-sex differential was enacted to ensure the availability of 3.2% beer for women as compensation for previous deprivations.

^7. That this was the true purpose is not at all self-evident. The purpose is not apparent from the face of the statute and the Oklahoma Legislature does not preserve statutory history materials capable of clarifying the objectives served by its legislative enactments. The District Court acknowledged the nonexistence of materials necessary "to reveal what the actual purpose of the legislature was," but concluded that "we feel it apparent that a major purpose of the legislature was to promote the safety of the young persons affected and the public generally." 399 F.Supp., at 1311 n. 6. Similarly, the attorney for Oklahoma, while proposing traffic safety as a legitimate rationale for the 3.2% beer law, candidly acknowledged at oral argument that he is unable to assert that traffic safety is "indeed the reason" for the gender line contained in § 245. Tr. of Oral Arg. 27. For this appeal we find adequate the appellee's representation of legislative purpose, leaving for another day consideration of whether the statement of the State's Assistant Attorney General should suffice to inform this Court of the legislature's objectives, or whether the Court must determine if the litigant simply is selecting a convenient, but false, post hoc rationalization.

^8. The disparities in 18-20-year-old male-female arrests were substantial for both categories of offenses: 427 versus 24 for driving under the influence of alcohol, and 966 versus 102 for drunkenness. Even if we assume that a legislature may rely on such arrest data in some situations, these figures do not offer support for a differential age line, for the disproportionate arrests of males persisted at older ages; indeed, in the case of arrests for drunkenness, the figures for all ages indicated "even more male involvement in such arrests at later ages." 399 F.Supp., at 1309. See also n. 14, infra.

^9. This survey drew no correlation between the accident figures for any age group and levels of intoxication found in those killed or injured.

^10. For an analysis of the results of this exhibit, see n. 16, infra.

^11. The FBI made no attempt to relate these arrest figures either to beer drinking or to an 18-21 age differential, but rather found that male arrests for all ages exceeded 90% of the total.

^12. Obviously, arrest statistics do not embrace all individuals who drink and drive. But for purposes of analysis, this "underinclusiveness" must be discounted somewhat by the shortcomings inherent in this statistical sample, see n. 14, infra. In any event, we decide this case in light of the evidence offered by Oklahoma and know of no way of extrapolating these arrest statistics to take into account the driving and drinking population at large, including those who avoided arrest.

^13. For example, we can conjecture that in Reed, Idaho's apparent premise that women lacked experience in formal business matters (particularly compared to men) would have proved to be accurate in substantially more than 2% of all cases. And in both Frontiero and Wiesenfeld, we expressly found appellees' empirical defense of mandatory dependency tests for men but not women to be unsatisfactory, even though we recognized that husbands are still far less likely to be dependent on their wives than vice versa. See, e. g., 411 U.S., at 688-690, 93 S.Ct., at 1771-1772.

^14. The very social stereotypes that find reflection in age-differential laws, see Stanton v. Stanton, 421 U.S. 7, 14-15, 95 S.Ct., at 1378 (1975), are likely substantially to distort the accuracy of these comparative statistics. Hence "reckless" young men who drink and drive are transformed into arrest statistics, whereas their female counterparts are chivalrously escorted home. See, e. g., W. Reckless & B. Kay, The Female Offender 4, 7, 13, 16-17 (Report to Presidential Commission on Law Enforcement and Administration of Justice, 1967). Moreover, the Oklahoma surveys, gathered under a regime where the age-differential law in question has been in effect, are lacking in controls necessary for appraisal of the actual effectiveness of the male 3.2% beer prohibition. In this regard, the disproportionately high arrest statistics for young males and, indeed, the growing alcohol-related arrest figures for all ages and sexes simply may be taken to document the relative futility of controlling driving behavior by the 3.2% beer statute and like legislation, although we obviously have no means of estimating how many individuals, if any, actually were prevented from drinking by these laws.

^15. See, e. g., nn. 9 and 11, supra. See also n. 16, infra.

^16. The random roadside survey of drivers conducted in Oklahoma City during August 1972 found that 78% of drivers under 20 were male. Turning to an evaluation of their drinking habits and factoring out nondrinkers, 84% of the males versus 77% of the females expressed a preference for beer. Further 16.5% of the men and 11.4% of the women had consumed some alcoholic beverage within two hours of the interview. Finally, a blood alcohol concentration greater than .01% was discovered in 14.6% of the males compared to 11.5% of the females. "The 1973 figures, although they contain some variations, reflect essentially the same pattern." 399 F.Supp., at 1309. Plainly these statistical disparities between the sexes are not substantial. Moreover, when the 18-20 age boundaries are lifted and all drivers analyzed, the 1972 roadside survey indicates that male drinking rose slightly whereas female exposure to alcohol remained relatively constant. Again, in 1973, the survey established that "compared to all drivers interviewed, . . . the under-20 age group generally showed a lower involvement with alcohol in terms of having drunk within the past two hours or having a significant BAC (blood alcohol content)." Ibid. In sum, this survey provides little support for a gender line among teenagers and actually runs counter to the imposition of drinking restrictions based upon age.

^17. See, e. g., n. 22, infra.

^18. The Wilson Act, enacted in 1890, reads in pertinent part: "All . . . intoxicating liquors or liquids transported into any State or Territory . . . shall upon arrival in such State or Territory be subject to the operation and effect of the laws of such State or Territory enacted in the exercise of its police powers, to the same extent and in the same manner as though such liquids or liquors had been produced in such State or Territory . . . ." 27 U.S.C. § 121.

^19. The Webb-Kenyon Act of 1913 prohibits "(t)he shipment or transportation . . . of any . . . intoxicating liquor of any kind from one State, Territory, or District . . . into any other State, Territory, or District . . . (for the purpose of being) received, possessed, sold, or in any manner used . . . in violation of any law of such State, Territory, or District . . .." 27 U.S.C. § 122.

^20. "The transportation or importation into any State, Territory, or possession of the United States for delivery or use therein of intoxicating liquors, in violation of the laws thereof, is hereby prohibited."

^21. The dictum contained in State Board of Equalization of California v. Young's Market Co., 299 U.S. 59, 64, 57 S.Ct. 77, 79, 81 L.Ed. 38 (1936), that "(a) classification recognized by the Twenty-first Amendment cannot be deemed forbidden by the Fourteenth," is inapplicable to this case. The Twenty-first Amendment does not recognize, even indirectly, classifications based upon gender. And, as the accompanying text demonstrates, that statement has not been relied upon in recent cases that have considered Fourteenth Amendment challenges to state liquor regulation.

^22. Thus, if statistics were to govern the permissibility of state alcohol regulation without regard to the Equal protection Clause as a limiting principle, it might follow that States could freely favor Jews and Italian Catholics at the expense of all other Americans, since available studies regularly demonstrate that the former two groups exhibit the lowest rates of problem drinking. See, e. g., Haberman & Sheinberg, Implicative Drinking Reported in a Household Survey: A Corroborative Note on Subgroup Differences, 28 Q. J. Studies on Alcohol 538 (1967); Wechsler, Thum, Demone, & Dwinnell, Social Characteristics and Blood Alcohol Level, 33 Q. J. Studies on Alcohol 132, 141-142 (1972); Wechsler,

Demone, Thum, & Kasey, Religious-Ethnic Differences In Alcohol Consumption, 11 J. Health & Soc. Behavior 21, 28 (1970); Schmidt & Popham, Impressions of Jewish Alcoholics, 37 J. Studies on Alcohol 931 (1976). Similarly, if a State were allowed simply to depend upon demographic characteristics of adolescents in identifying problem drinkers, statistics might support the conclusion that only black teenagers should be permitted to drink, followed by Asian-Americans and Spanish-Americans. "Whites and American Indians have the lowest proportions of abstainers and the highest proportions of moderate/heavy and heavy drinkers." Summary of Final Report of a National Study of Adolescent Drinking Behavior, Attitudes and Correlates 147-148 (Center for the Study of Social Behavior, Research Triangle Inst., Apr. 1975) (percentage of moderate/heavy and heavy adolescent drinkers by race: black 15.2%; Asian-American 18.3%; Spanish-American 22.7%; white 25.3%; American Indian 28.1%).

In the past, some States have acted upon their notions of the drinking propensities of entire groups in fashioning their alcohol policies. The most typical recipient of this treatment has been the American Indian; indeed, several States established criminal sanctions for the sale of alcohol to an Indian or "half or quarter breed Indian." See, e. g., Fla.Stat.Ann. § 569.07 (1962 and 1976 Supp.) (repealed in 1972); Iowa Code Ann. § 732.5 (1950 and 1976 Supp.) (repealed in 1967); Minn.Stat.Ann. § 340.82 (1957) (repealed in 1969); Neb.Rev.Stat. 53-181 (1944) (repealed in 1955); Utah Code Ann. § 76-34-1 (1953 and 1975 Supp.) (repealed in 1955). Other statutes and constitutional provisions proscribed the introduction of alcoholic beverages onto Indian reservations. See, e. g., Act of June 10, 1910, § 2, 36 Stat. 558; Ariz.Const., Art. XX, § 3; N.M.Const., Art. XXI, § 8; Okla.Const., Art. I, § 7. While Indian-oriented provisions were the most common, state alcohol beverage prohibitions also have been directed at other groups, notably German, Italian, and Catholic immigrants. See, e. g., J. Higham, Strangers in the Land 25, 267-268, 295 (1975). The repeal of most of these laws signals society's perception of the unfairness and questionable constitutionality of singling out groups to bear the brunt of alcohol regulation.

^* As is evident from our opinions, the Court has had difficulty in agreeing upon a standard of equal protection analysis that can be applied consistently to the wide variety of legislative classifications. There are valid reasons for dissatisfaction with the "two-tier" approach that has been prominent in the Court's decisions in the past decade. Although viewed by many as a result-oriented substitute for more critical analysis, that approach with its narrowly limited "upper-tier" now has substantial precedential support. As has been true of Reed and its progeny, our decision today will be viewed by some as a "middle-tier" approach. While I would not endorse that characterization and would not welcome a further subdividing of equal protection analysis, candor compels the recognition that the relatively deferential "rational basis" standard of review normally applied takes on a sharper focus when we address a gender-based classification. So much is clear from our recent cases. For thoughtful discussions of equal protection analysis, see, e. g., Gunther, The Supreme Court, 1971 Term Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for A Newer Equal Protection, 86 Harv.L.Rev. 1 (1972); Wilkinson, The Supreme Court, the Equal Protection Clause, and the Three Faces of Constitutional Equality, 61 Va.L.Rev. 945 (1975).

----------

^23. Insofar as Goesaert v. Cleary, 335 U.S. 464, 69 S.Ct. 198, 93 L.Ed. 163 (1948), may be inconsistent, that decision is disapproved. Undoubtedly reflecting the view that Goesaert's equal protection analysis no longer obtains, the District Court made no reference to that decision in upholding Oklahoma's statute. Similarly, the opinions of the federal and state courts cited earlier in the text invalidating gender lines with respect to alcohol regulation uniformly disparaged the contemporary vitality of Goesaert.

^24. As noted in Stanton v. Stanton, 421 U.S., at 17-18, 95 S.Ct., at 1379-1380, the Oklahoma Legislature is free to redefine any cutoff age for the purchase and sale of 3.2% beer that it may choose, provided that the redefinition operates in a gender-neutral fashion.

^1. Men as a general class have not been the victims of the kind of historic, pervasive discrimination that has disadvantaged other groups.

^2. "(S)ince sex, like race and national origin, is an immutable characteristic determined solely by the accident of birth, the imposition of special disabilities upon the members of a particular sex because of their sex would seem to violate 'the basic concept of our system that legal burdens should bear some relationship to individual responsibility . . .,' Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 175, 92 S.Ct. 1400, 1406, 31 L.Ed.2d 768." Frontiero v. Richardson, 411 U.S. 677, 686, 93 S.Ct. 1764, 1770, 36 L.Ed.2d 583.

^3. Apparently Oklahoma is the only State to permit this narrow discrimination to survive the elimination of the disparity between the age of majority for males and females.

^4. Because males are generally heavier than females, they have a greater capacity to consume alcohol without impairing their driving ability than do females.

^5. There is no legislative history to indicate that this was the purpose, and several features of the statutory scheme indicate the contrary. The statute exempts license holders who dispense 3.2% beer to their own children, and a related statute makes it unlawful for 18-year-old men (but not women) to work in establishments in which 3.2% beer accounts for over 25% of gross sales. Okla.Stat., Tit. 37, §§ 241, 243, 245 (1953 and Supp.1976).

There is, of course, no way of knowing what actually motivated this discrimination, but I would not be surprised if it represented nothing more than the perpetuation of a stereotyped attitude about the relative maturity of the members of the two sexes in this age bracket. If so, the following comment is relevant:

"(A) traditional classification is more likely to be used without pausing to consider its justification than is a newly created classification. Habit, rather than analysis, makes it seem acceptable and natural to distinguish between male and female, alien and citizen, legitimate and illegitimate; for too much of our history there was the same inertia in distinguishing between black and white. But that sort of stereotyped reaction may have no rational relationship other than pure prejudicial discrimination to the stated purpose for which the classification is being made." Mathews v. Lucas, 427 U.S. 495, at 520-521, 96 S.Ct. 2755, at 2769, 49 L.Ed.2d 651 (Stevens, J., dissenting).

^6. It forbids the sale of 3.2% beer to 18-21-year-old men without forbidding possession, or preventing them from obtaining it from other sources, such as friends who are either older or female. Thus, the statute only slightly impedes access to 3.2% beer.

^7. The only direct evidence submitted by the State concerning use of beer by young drivers indicates that there is no substantial difference between the sexes. In a random roadside survey of drivers, 16.5% of the male drivers under 20 had consumed alcohol within two hours of the interview as opposed to 11.4% of the women. Over three-fourths of the nonabstainers in both groups expressed a preference for beer. And 14.6% of the men, as opposed to 11.5% of the women, had blood alcohol concentrations over .01%. See ante, at 203 n. 16.

^1. In Stanley v. Illinois, 405 U.S. 645, 92 S.Ct. 1208, 31 L.Ed.2d 551 (1972), the Court struck down a statute allowing separation of illegitimate children from a surviving father but not a surviving mother, without any showing of parental unfitness. The Court stated that "the interest of a parent in the companionship, care, custody, and management of his or her children 'come(s) to this Court with a momentum for respect lacking when appeal is made to liberties which derive merely from shifting economic arrangements.' "

In Kahn v. Shevin, 416 U.S. 351, 94 S.Ct. 1734, 40 L.Ed.2d 189 (1974), the Court upheld Florida's $500 property tax exemption for widows only. The opinion of the Court appears to apply a rational-basis test, id., at 355, 94 S.Ct., at 1737, and is so understood by the dissenters. Id., at 357, 94 S.Ct., at 1738 (BRENNAN, J., joined by MARSHALL, J., dissenting).

In Weinberger v. Wiesenfeld, 420 U.S. 636, 95 S.Ct. 1225, 43 L.Ed.2d 514 (1975), the Court invalidated § 202(g) of the Social Security Act, which allowed benefits to mothers but not fathers of minor children, who survive the wage earner. This statute was treated, in the opinion of the Court, as a discrimination against females wage earners, on the ground that it minimizes the financial security which their work efforts provide for their families. 420 U.S., at 645, 95 S.Ct., at 1231.

^2. I am not unaware of the argument from time to time advanced, that all discriminations between the sexes ultimately redound to the detriment of females, because they tend to reinforce "old notions" restricting the roles and opportunities of women. As a general proposition applying equally to all sex categorizations, I believe that this argument was implicitly found to carry little weight in our decisions upholding gender-based differences. See Schlesinger v. Ballard, 419 U.S. 498, 95 S.Ct. 572, 42 L.Ed.2d 610 (1975); Kahn v. Shevin, supra, 416 U.S. 351, 94 S.Ct. 1734, 40 L.Ed.2d 189 (1974). Seeing no assertion that it has special applicability to the situation at hand, I believe it can be dismissed as an insubstantial consideration.

^3. Extract from: Oklahoma Bureau of Investigation, Arrest Statistics for September, October, November, and December 1973. Defendants' Exhibit 1, Jurisdictional Statement A22. Extract from: Oklahoma City Police Department, Arrest Statistics for 1973. Defendants' Exhibit 2, Jurisdictional Statement A23. See ante, at 200 n. 8.

^4. Extract from: Federal Bureau of Investigation, Crime in the United States, 1972. Defendants' Exhibit 6, App. 182-184.

^5. Extract from: Oklahoma Department of Public Safety, Summary of Statewide Collisions for 1972, 1973. Defendants' Exhibits 4 and 5, Jurisdictional Statement A30-A31.

^6. Extract from: Oklahoma Management and Engineering Consulting, Inc., Report to Alcohol Safety Action Program (1973). Defendants' Exhibit 3, Table 1, Jurisdictional Statement A25.

^7. Id., at A27 (Table 3), A29 (Table 5).

^8. Id., at A25 (Table 1). See ante, at 203 n. 16.

        Responding to a reported weapons disturbance in a private residence, Houston police entered petitioner Lawrence's apartment and saw him and another adult man, petitioner Garner, engaging in a private, consensual sexual act. Petitioners were arrested and convicted of deviate sexual intercourse in violation of a Texas statute forbidding two persons of the same sex to engage in certain intimate sexual conduct. In affirming, the State Court of Appeals held, inter alia, that the statute was not unconstitutional under the Due Process Clause of the Fourteenth Amendment. The court considered Bowers v. Hardwick, 478 U. S. 186, controlling on that point.

        Held: The Texas statute making it a crime for two persons of the same sex to engage in certain intimate sexual conduct violates the Due Process Clause. Pp. 564-579.

        (a) Resolution of this case depends on whether petitioners were free as adults to engage in private conduct in the exercise of their liberty under the Due Process Clause. For this inquiry the Court deems it necessary to reconsider its Bowers holding. The Bowers Court's initial substantive statement—"The issue presented is whether the Federal Constitution confers a fundamental right upon homosexuals to engage in sodomy . . .," 478 U. S., at 190—discloses the Court's failure to appreciate the extent of the liberty at stake. To say that the issue in Bowers was simply the right to engage in certain sexual conduct demeans the claim the individual put forward, just as it would demean a married couple were it said that marriage is just about the right to have sexual intercourse. Although the laws involved in Bowers and here purport to do no more than prohibit a particular sexual act, their penalties and purposes have more far-reaching consequences, touching upon the most private human conduct, sexual behavior, and in the most private of places, the home. They seek to control a personal relationship that, whether or not entitled to formal recognition in the law, is within the liberty of persons to choose without being punished as criminals. The liberty protected by the Constitution allows homosexual persons the right to choose to enter upon relationships in the confines of their homes and their own private lives and still retain their dignity as free persons. Pp. 564-567.

[539 U.S. 559]

        (b) Having misapprehended the liberty claim presented to it, the Bowers Court stated that proscriptions against sodomy have ancient roots. 478 U. S., at 192. It should be noted, however, that there is no longstanding history in this country of laws directed at homosexual conduct as a distinct matter. Early American sodomy laws were not directed at homosexuals as such but instead sought to prohibit nonprocreative sexual activity more generally, whether between men and women or men and men. Moreover, early sodomy laws seem not to have been enforced against consenting adults acting in private. Instead, sodomy prosecutions often involved predatory acts against those who could not or did not consent: relations between men and minor girls or boys, between adults involving force, between adults implicating disparity in status, or between men and animals. The longstanding criminal prohibition of homosexual sodomy upon which Bowers placed such reliance is as consistent with a general condemnation of nonprocreative sex as it is with an established tradition of prosecuting acts because of their homosexual character. Far from possessing "ancient roots," ibid., American laws targeting same-sex couples did not develop until the last third of the 20th century. Even now, only nine States have singled out same-sex relations for criminal prosecution. Thus, the historical grounds relied upon in Bowers are more complex than the majority opinion and the concurring opinion by Chief Justice Burger there indicated. They are not without doubt and, at the very least, are overstated. The Bowers Court was, of course, making the broader point that for centuries there have been powerful voices to condemn homosexual conduct as immoral, but this Court's obligation is to define the liberty of all, not to mandate its own moral code, Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 850. The Nation's laws and traditions in the past half century are most relevant here. They show an emerging awareness that liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex. See County of Sacramento v. Lewis, 523 U. S. 833, 857. Pp. 567-573.

        (c) Bowers' deficiencies became even more apparent in the years following its announcement. The 25 States with laws prohibiting the conduct referenced in Bowers are reduced now to 13, of which 4 enforce their laws only against homosexual conduct. In those States, including Texas, that still proscribe sodomy (whether for same-sex or heterosexual conduct), there is a pattern of nonenforcement with respect to consenting adults acting in private. Casey, supra, at 851—which confirmed that the Due Process Clause protects personal decisions relating to marriage, procreation, contraception, family relationships, child rearing, and education—and Romer v. Evans, 517 U. S. 620, 624—which struck down class-based legislation directed at homosexuals—cast Bowers'

[539 U.S. 560]

holding into even more doubt. The stigma the Texas criminal statute imposes, moreover, is not trivial. Although the offense is but a minor misdemeanor, it remains a criminal offense with all that imports for the dignity of the persons charged, including notation of convictions on their records and on job application forms, and registration as sex offenders under state law. Where a case's foundations have sustained serious erosion, criticism from other sources is of greater significance. In the United States, criticism of Bowers has been substantial and continuing, disapproving of its reasoning in all respects, not just as to its historical assumptions. And, to the extent Bowers relied on values shared with a wider civilization, the case's reasoning and holding have been rejected by the European Court of Human Rights, and that other nations have taken action consistent with an affirmation of the protected right of homosexual adults to engage in intimate, consensual conduct. There has been no showing that in this country the governmental interest in circumscribing personal choice is somehow more legitimate or urgent. Stare decisis is not an inexorable command. Payne v. Tennessee, 501 U. S. 808, 828. Bowers' holding has not induced detrimental reliance of the sort that could counsel against overturning it once there are compelling reasons to do so. Casey, supra, at 855-856. Bowers causes uncertainty, for the precedents before and after it contradict its central holding. Pp. 573-577.

        (d) Bowers' rationale does not withstand careful analysis. In his dissenting opinion in Bowers Justice Stevens concluded that (1) the fact that a State's governing majority has traditionally viewed a particular practice as immoral is not a sufficient reason for upholding a law prohibiting the practice, and (2) individual decisions concerning the intimacies of physical relationships, even when not intended to produce offspring, are a form of "liberty" protected by due process. That analysis should have controlled Bowers, and it controls here. Bowers was not correct when it was decided, is not correct today, and is hereby overruled. This case does not involve minors, persons who might be injured or coerced, those who might not easily refuse consent, or public conduct or prostitution. It does involve two adults who, with full and mutual consent, engaged in sexual practices common to a homosexual lifestyle. Petitioners' right to liberty under the Due Process Clause gives them the full right to engage in private conduct without government intervention. Casey, supra, at 847. The Texas statute furthers no legitimate state interest which can justify its intrusion into the individual's personal and private life. Pp. 577-579.

        41 S. W. 3d 349, reversed and remanded.

[539 U.S. 561]

        KENNEDY, J., delivered the opinion of the Court, in which STEVENS, SOUTER, GINSBURG, and BREYER, JJ., joined. O'CONNOR, J., filed an opinion concurring in the judgment, post, p. 579. SCALIA, J., filed a dissenting opinion, in which REHNQUIST, C. J., and THOMAS, J., joined, post, p. 586. THOMAS, J., filed a dissenting opinion, post, p. 605.

        CERTIORARI TO THE COURT OF APPEALS OF TEXAS, FOURTEENTH DISTRICT

        Paul M. Smith argued the cause for petitioners. With him on the briefs were William M. Hohengarten, Daniel Mach, Mitchell Katine, Ruth E. Harlow, Patricia M. Logue, and Susan L. Sommer.

        Charles A. Rosenthal, Jr., argued the cause for respondent. With him on the brief were William J. Delmore III and Scott A. Durfee.^*

[539 U.S. 562]

        JUSTICE KENNEDY delivered the opinion of the Court.

        Liberty protects the person from unwarranted government intrusions into a dwelling or other private places. In our tradition the State is not omnipresent in the home. And there are other spheres of our lives and existence, outside the home, where the State should not be a dominant presence. Freedom extends beyond spatial bounds. Liberty presumes an autonomy of self that includes freedom of thought, belief, expression, and certain intimate conduct. The instant case involves liberty of the person both in its spatial and in its more transcendent dimensions.

        The question before the Court is the validity of a Texas statute making it a crime for two persons of the same sex to engage in certain intimate sexual conduct.

        In Houston, Texas, officers of the Harris County Police Department were dispatched to a private residence in response to a reported weapons disturbance. They entered an apartment where one of the petitioners, John Geddes Lawrence,

[539 U.S. 563]

resided. The right of the police to enter does not seem to have been questioned. The officers observed Lawrence and another man, Tyron Garner, engaging in a sexual act. The two petitioners were arrested, held in custody overnight, and charged and convicted before a Justice of the Peace.

        The complaints described their crime as "deviate sexual intercourse, namely anal sex, with a member of the same sex (man)." App. to Pet. for Cert. 127a, 139a. The applicable state law is Tex. Penal Code Ann. § 21.06(a) (2003). It provides: "A person commits an offense if he engages in deviate sexual intercourse with another individual of the same sex." The statute defines "[d]eviate sexual intercourse" as follows:

        "(A) any contact between any part of the genitals of one person and the mouth or anus of another person; or

        "(B) the penetration of the genitals or the anus of another person with an object." § 21.01(1).

        The petitioners exercised their right to a trial de novo in Harris County Criminal Court. They challenged the statute as a violation of the Equal Protection Clause of the Fourteenth Amendment and of a like provision of the Texas Constitution. Tex. Const., Art. 1, § 3a. Those contentions were rejected. The petitioners, having entered a plea of nolo contendere, were each fined $200 and assessed court costs of $141.25. App. to Pet. for Cert. 107a-110a.

        The Court of Appeals for the Texas Fourteenth District considered the petitioners' federal constitutional arguments under both the Equal Protection and Due Process Clauses of the Fourteenth Amendment. After hearing the case en banc the court, in a divided opinion, rejected the constitutional arguments and affirmed the convictions. 41 S. W. 3d 349 (2001). The majority opinion indicates that the Court of Appeals considered our decision in Bowers v. Hardwick, 478 U. S. 186 (1986), to be controlling on the federal due process aspect of the case. Bowers then being authoritative, this was proper.

[539 U.S. 564]

        We granted certiorari, 537 U. S. 1044 (2002), to consider three questions:

        1. Whether petitioners' criminal convictions under the Texas "Homosexual Conduct" law—which criminalizes sexual intimacy by same-sex couples, but not identical behavior by different-sex couples—violate the Fourteenth Amendment guarantee of equal protection of the laws.

        2. Whether petitioners' criminal convictions for adult consensual sexual intimacy in the home violate their vital interests in liberty and privacy protected by the Due Process Clause of the Fourteenth Amendment.

        3. Whether Bowers v. Hardwick, supra, should be overruled? See Pet. for Cert. i.

        The petitioners were adults at the time of the alleged offense. Their conduct was in private and consensual.

        We conclude the case should be resolved by determining whether the petitioners were free as adults to engage in the private conduct in the exercise of their liberty under the Due Process Clause of the Fourteenth Amendment to the Constitution. For this inquiry we deem it necessary to reconsider the Court's holding in Bowers.

        There are broad statements of the substantive reach of liberty under the Due Process Clause in earlier cases, including Pierce v. Society of Sisters, 268 U. S. 510 (1925), and Meyer v. Nebraska, 262 U. S. 390 (1923); but the most pertinent beginning point is our decision in Griswold v. Connecticut, 381 U. S. 479 (1965).

        In Griswold the Court invalidated a state law prohibiting the use of drugs or devices of contraception and counseling or aiding and abetting the use of contraceptives. The Court described the protected interest as a right to privacy and

[539 U.S. 565]

placed emphasis on the marriage relation and the protected space of the marital bedroom. Id., at 485.

        After Griswold it was established that the right to make certain decisions regarding sexual conduct extends beyond the marital relationship. In Eisenstadt v. Baird, 405 U. S. 438 (1972), the Court invalidated a law prohibiting the distribution of contraceptives to unmarried persons. The case was decided under the Equal Protection Clause, id., at 454; but with respect to unmarried persons, the Court went on to state the fundamental proposition that the law impaired the exercise of their personal rights, ibid. It quoted from the statement of the Court of Appeals finding the law to be in conflict with fundamental human rights, and it followed with this statement of its own:

        "It is true that in Griswold the right of privacy in question inhered in the marital relationship. . . . If the right of privacy means anything, it is the right of the individual, married or single, to be free from unwarranted governmental intrusion into matters so fundamentally affecting a person as the decision whether to bear or beget a child." Id., at 453.

        The opinions in Griswold and Eisenstadt were part of the background for the decision in Roe v. Wade, 410 U. S. 113 (1973). As is well known, the case involved a challenge to the Texas law prohibiting abortions, but the laws of other States were affected as well. Although the Court held the woman's rights were not absolute, her right to elect an abortion did have real and substantial protection as an exercise of her liberty under the Due Process Clause. The Court cited cases that protect spatial freedom and cases that go well beyond it. Roe recognized the right of a woman to make certain fundamental decisions affecting her destiny and confirmed once more that the protection of liberty under the Due Process Clause has a substantive dimension of fundamental significance in defining the rights of the person.

[539 U.S. 566]

        In Carey v. Population Services Int'l, 431 U. S. 678 (1977), the Court confronted a New York law forbidding sale or distribution of contraceptive devices to persons under 16 years of age. Although there was no single opinion for the Court, the law was invalidated. Both Eisenstadt and Carey, as well as the holding and rationale in Roe, confirmed that the reasoning of Griswold could not be confined to the protection of rights of married adults. This was the state of the law with respect to some of the most relevant cases when the Court considered Bowers v. Hardwick.

        The facts in Bowers had some similarities to the instant case. A police officer, whose right to enter seems not to have been in question, observed Hardwick, in his own bedroom, engaging in intimate sexual conduct with another adult male. The conduct was in violation of a Georgia statute making it a criminal offense to engage in sodomy. One difference between the two cases is that the Georgia statute prohibited the conduct whether or not the participants were of the same sex, while the Texas statute, as we have seen, applies only to participants of the same sex. Hardwick was not prosecuted, but he brought an action in federal court to declare the state statute invalid. He alleged he was a practicing homosexual and that the criminal prohibition violated rights guaranteed to him by the Constitution. The Court, in an opinion by Justice White, sustained the Georgia law. Chief Justice Burger and Justice Powell joined the opinion of the Court and filed separate, concurring opinions. Four Justices dissented. 478 U. S., at 199 (opinion of Blackmun, J., joined by Brennan, Marshall, and Stevens, JJ.); id., at 214 (opinion of Stevens, J.,joined by Brennan and Marshall, JJ.).

        The Court began its substantive discussion in Bowers as follows: "The issue presented is whether the Federal Constitution confers a fundamental right upon homosexuals to engage in sodomy and hence invalidates the laws of the many States that still make such conduct illegal and have done so

[539 U.S. 567]

for a very long time." Id., at 190. That statement, we now conclude, discloses the Court's own failure to appreciate the extent of the liberty at stake. To say that the issue in Bowers was simply the right to engage in certain sexual conduct demeans the claim the individual put forward, just as it would demean a married couple were it to be said marriage is simply about the right to have sexual intercourse. The laws involved in Bowers and here are, to be sure, statutes that purport to do no more than prohibit a particular sexual act. Their penalties and purposes, though, have more far-reaching consequences, touching upon the most private human conduct, sexual behavior, and in the most private of places, the home. The statutes do seek to control a personal relationship that, whether or not entitled to formal recognition in the law, is within the liberty of persons to choose without being punished as criminals.

        This, as a general rule, should counsel against attempts by the State, or a court, to define the meaning of the relationship or to set its boundaries absent injury to a person or abuse of an institution the law protects. It suffices for us to acknowledge that adults may choose to enter upon this relationship in the confines of their homes and their own private lives and still retain their dignity as free persons. When sexuality finds overt expression in intimate conduct with another person, the conduct can be but one element in a personal bond that is more enduring. The liberty protected by the Constitution allows homosexual persons the right to make this choice.

        Having misapprehended the claim of liberty there presented to it, and thus stating the claim to be whether there is a fundamental right to engage in consensual sodomy, the Bowers Court said: "Proscriptions against that conduct have ancient roots." Id., at 192. In academic writings, and in many of the scholarly amicus briefs filed to assist the Court in this case, there are fundamental criticisms of the historical premises relied upon by the majority and concurring opinions

[539 U.S. 568]

in Bowers. Brief for Cato Institute as Amicus Curiae 16-17; Brief for American Civil Liberties Union et al. as Amici Curiae 15-21; Brief for Professors of History et al. as Amici Curiae 3-10. We need not enter this debate in the attempt to reach a definitive historical judgment, but the following considerations counsel against adopting the definitive conclusions upon which Bowers placed such reliance.

        At the outset it should be noted that there is no longstanding history in this country of laws directed at homosexual conduct as a distinct matter. Beginning in colonial times there were prohibitions of sodomy derived from the English criminal laws passed in the first instance by the Reformation Parliament of 1533. The English prohibition was understood to include relations between men and women as well as relations between men and men. See, e. g., King v. Wiseman, 92 Eng. Rep. 774, 775 (K. B. 1718) (interpreting "mankind" in Act of 1533 as including women and girls). Nineteenth-century commentators similarly read American sodomy, buggery, and crime-against-nature statutes as criminalizing certain relations between men and women and between men and men. See, e. g., 2 J. Bishop, Criminal Law § 1028 (1858); 2 J. Chitty, Criminal Law 47-50 (5th Am. ed. 1847); R. Desty, A Compendium of American Criminal Law 143 (1882); J. May, The Law of Crimes § 203 (2d ed. 1893). The absence of legal prohibitions focusing on homosexual conduct may be explained in part by noting that according to some scholars the concept of the homosexual as a distinct category of person did not emerge until the late 19th century. See, e. g., J. Katz, The Invention of Heterosexuality 10 (1995); J. D'Emilio & E. Freedman, Intimate Matters: A History of Sexuality in America 121 (2d ed. 1997) ("The modern terms homosexuality and heterosexuality do not apply to an era that had not yet articulated these distinctions"). Thus early American sodomy laws were not directed at homosexuals as such but instead sought to prohibit nonprocreative sexual activity more generally. This does not suggest approval of

[539 U.S. 569]

homosexual conduct. It does tend to show that this particular form of conduct was not thought of as a separate category from like conduct between heterosexual persons.

        Laws prohibiting sodomy do not seem to have been enforced against consenting adults acting in private. A substantial number of sodomy prosecutions and convictions for which there are surviving records were for predatory acts against those who could not or did not consent, as in the case of a minor or the victim of an assault. As to these, one purpose for the prohibitions was to ensure there would be no lack of coverage if a predator committed a sexual assault that did not constitute rape as defined by the criminal law. Thus the model sodomy indictments presented in a 19th-century treatise, see 2 Chitty, supra, at 49, addressed the predatory acts of an adult man against a minor girl or minor boy. Instead of targeting relations between consenting adults in private, 19th-century sodomy prosecutions typically involved relations between men and minor girls or minor boys, relations between adults involving force, relations between adults implicating disparity in status, or relations between men and animals.

        To the extent that there were any prosecutions for the acts in question, 19th-century evidence rules imposed a burden that would make a conviction more difficult to obtain even taking into account the problems always inherent in prosecuting consensual acts committed in private. Under then-prevailing standards, a man could not be convicted of sodomy based upon testimony of a consenting partner, because the partner was considered an accomplice. A partner's testimony, however, was admissible if he or she had not consented to the act or was a minor, and therefore incapable of consent. See, e. g., F. Wharton, Criminal Law 443 (2d ed. 1852); 1 F. Wharton, Criminal Law 512 (8th ed. 1880). The rule may explain in part the infrequency of these prosecutions. In all events that infrequency makes it difficult to say that society approved of a rigorous and systematic

[539 U.S. 570]

punishment of the consensual acts committed in private and by adults. The longstanding criminal prohibition of homosexual sodomy upon which the Bowers decision placed such reliance is as consistent with a general condemnation of nonprocreative sex as it is with an established tradition of prosecuting acts because of their homosexual character.

        The policy of punishing consenting adults for private acts was not much discussed in the early legal literature. We can infer that one reason for this was the very private nature of the conduct. Despite the absence of prosecutions, there may have been periods in which there was public criticism of homosexuals as such and an insistence that the criminal laws be enforced to discourage their practices. But far from possessing "ancient roots," Bowers, 478 U. S., at 192, American laws targeting same-sex couples did not develop until the last third of the 20th century. The reported decisions concerning the prosecution of consensual, homosexual sodomy between adults for the years 1880-1995 are not always clear in the details, but a significant number involved conduct in a public place. See Brief for American Civil Liberties Union et al. as Amici Curiae 14-15, and n. 18.

        It was not until the 1970's that any State singled out same-sex relations for criminal prosecution, and only nine States have done so. See 1977 Ark. Gen. Acts no. 828; 1983 Kan. Sess. Laws p. 652; 1974 Ky. Acts p. 847; 1977 Mo. Laws p. 687; 1973 Mont. Laws p. 1339; 1977 Nev. Stats. p. 1632; 1989 Tenn. Pub. Acts ch. 591; 1973 Tex. Gen. Laws ch. 399; see also Post v. State, 715 P. 2d 1105 (Okla. Crim. App. 1986) (sodomy law invalidated as applied to different-sex couples). Post-Bowers even some of these States did not adhere to the policy of suppressing homosexual conduct. Over the course of the last decades, States with same-sex prohibitions have moved toward abolishing them. See, e. g., Jegley v. Picado, 349 Ark. 600, 80 S. W. 3d 332 (2002); Gryczan v. State, 283 Mont. 433, 942 P. 2d 112 (1997); Campbell v. Sundquist, 926 S. W. 2d 250 (Tenn. App. 1996); Commonwealth v. Wasson,

[539 U.S. 571]

842 S. W. 2d 487 (Ky. 1992); see also 1993 Nev. Stats. p. 518 (repealing Nev. Rev. Stat. § 201.193).

        In summary, the historical grounds relied upon in Bowers are more complex than the majority opinion and the concurring opinion by Chief Justice Burger indicate. Their historical premises are not without doubt and, at the very least, are overstated.

        It must be acknowledged, of course, that the Court in Bowers was making the broader point that for centuries there have been powerful voices to condemn homosexual conduct as immoral. The condemnation has been shaped by religious beliefs, conceptions of right and acceptable behavior, and respect for the traditional family. For many persons these are not trivial concerns but profound and deep convictions accepted as ethical and moral principles to which they aspire and which thus determine the course of their lives. These considerations do not answer the question before us, however. The issue is whether the majority may use the power of the State to enforce these views on the whole society through operation of the criminal law. "Our obligation is to define the liberty of all, not to mandate our own moral code." Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 850 (1992).

        Chief Justice Burger joined the opinion for the Court in Bowers and further explained his views as follows: "Decisions of individuals relating to homosexual conduct have been subject to state intervention throughout the history of Western civilization. Condemnation of those practices is firmly rooted in Judeao-Christian moral and ethical standards." 478 U. S., at 196. As with Justice White's assumptions about history, scholarship casts some doubt on the sweeping nature of the statement by Chief Justice Burger as it pertains to private homosexual conduct between consenting adults. See, e. g., Eskridge, Hardwick and Historiography, 1999 U. Ill. L. Rev. 631, 656. In all events we think that our laws and traditions in the past half century are of

[539 U.S. 572]

most relevance here. These references show an emerging awareness that liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex. "[H]istory and tradition are the starting point but not in all cases the ending point of the substantive due process inquiry." County of Sacramento v. Lewis, 523 U. S. 833, 857 (1998) (Kennedy, J., concurring).

        This emerging recognition should have been apparent when Bowers was decided. In 1955 the American Law Institute promulgated the Model Penal Code and made clear that it did not recommend or provide for "criminal penalties for consensual sexual relations conducted in private." ALI, Model Penal Code § 213.2, Comment 2, p. 372 (1980). It justified its decision on three grounds: (1) The prohibitions undermined respect for the law by penalizing conduct many people engaged in; (2) the statutes regulated private conduct not harmful to others; and (3) the laws were arbitrarily enforced and thus invited the danger of blackmail. ALI, Model Penal Code, Commentary 277-280 (Tent. Draft No. 4, 1955). In 1961 Illinois changed its laws to conform to the Model Penal Code. Other States soon followed. Brief for Cato Institute as Amicus Curiae 15-16.

        In Bowers the Court referred to the fact that before 1961 all 50 States had outlawed sodomy, and that at the time of the Court's decision 24 States and the District of Columbia had sodomy laws. 478 U. S., at 192-193. Justice Powell pointed out that these prohibitions often were being ignored, however. Georgia, for instance, had not sought to enforce its law for decades. Id., at 197-198, n. 2 ("The history of nonenforcement suggests the moribund character today of laws criminalizing this type of private, consensual conduct").

        The sweeping references by Chief Justice Burger to the history of Western civilization and to Judeo-Christian moral and ethical standards did not take account of other authorities pointing in an opposite direction. A committee advising the British Parliament recommended in 1957 repeal of laws

[539 U.S. 573]

punishing homosexual conduct. The Wolfenden Report: Report of the Committee on Homosexual Offenses and Prostitution (1963). Parliament enacted the substance of those recommendations 10 years later. Sexual Offences Act 1967, § 1.

        Of even more importance, almost five years before Bowers was decided the European Court of Human Rights considered a case with parallels to Bowers and to today's case. An adult male resident in Northern Ireland alleged he was a practicing homosexual who desired to engage in consensual homosexual conduct. The laws of Northern Ireland forbade him that right. He alleged that he had been questioned, his home had been searched, and he feared criminal prosecution. The court held that the laws proscribing the conduct were invalid under the European Convention on Human Rights. Dudgeon v. United Kingdom, 45 Eur. Ct. H. R. (1981) ¶ 52. Authoritative in all countries that are members of the Council of Europe (21 nations then, 45 nations now), the decision is at odds with the premise in Bowers that the claim put forward was insubstantial in our Western civilization.

        In our own constitutional system the deficiencies in Bowers became even more apparent in the years following its announcement. The 25 States with laws prohibiting the relevant conduct referenced in the Bowers decision are reduced now to 13, of which 4 enforce their laws only against homosexual conduct. In those States where sodomy is still proscribed, whether for same-sex or heterosexual conduct, there is a pattern of nonenforcement with respect to consenting adults acting in private. The State of Texas admitted in 1994 that as of that date it had not prosecuted anyone under those circumstances. State v. Morales, 869 S. W. 2d 941, 943.

        Two principal cases decided after Bowers cast its holding into even more doubt. In Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833 (1992), the Court reaffirmed the substantive force of the liberty protected by the Due Process Clause. The Casey decision again confirmed

[539 U.S. 574]

that our laws and tradition afford constitutional protection to personal decisions relating to marriage, procreation, contraception, family relationships, child rearing, and education. Id., at 851. In explaining the respect the Constitution demands for the autonomy of the person in making these choices, we stated as follows:

        "These matters, involving the most intimate and personal choices a person may make in a lifetime, choices central to personal dignity and autonomy, 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. Beliefs about these matters could not define the attributes of personhood were they formed under compulsion of the State." Ibid.

        Persons in a homosexual relationship may seek autonomy for these purposes, just as heterosexual persons do. The decision in Bowers would deny them this right.

        The second post-Bowers case of principal relevance is Romer v. Evans, 517 U. S. 620 (1996). There the Court struck down class-based legislation directed at homosexuals as a violation of the Equal Protection Clause. Romer invalidated an amendment to Colorado's Constitution which named as a solitary class persons who were homosexuals, lesbians, or bisexual either by "orientation, conduct, practices or relationships," id., at 624 (internal quotation marks omitted), and deprived them of protection under state antidiscrimination laws. We concluded that the provision was "born of animosity toward the class of persons affected" and further that it had no rational relation to a legitimate governmental purpose. Id., at 634.

        As an alternative argument in this case, counsel for the petitioners and some amici contend that Romer provides the basis for declaring the Texas statute invalid under the Equal Protection Clause. That is a tenable argument, but we conclude

[539 U.S. 575]

the instant case requires us to address whether Bowers itself has continuing validity. Were we to hold the statute invalid under the Equal Protection Clause some might question whether a prohibition would be valid if drawn differently, say, to prohibit the conduct both between same-sex and different-sex participants.

        Equality of treatment and the due process right to demand respect for conduct protected by the substantive guarantee of liberty are linked in important respects, and a decision on the latter point advances both interests. If protected conduct is made criminal and the law which does so remains unexamined for its substantive validity, its stigma might remain even if it were not enforceable as drawn for equal protection reasons. When homosexual conduct is made criminal by the law of the State, that declaration in and of itself is an invitation to subject homosexual persons to discrimination both in the public and in the private spheres. The central holding of Bowers has been brought in question by this case, and it should be addressed. Its continuance as precedent demeans the lives of homosexual persons.

        The stigma this criminal statute imposes, moreover, is not trivial. The offense, to be sure, is but a class C misdemeanor, a minor offense in the Texas legal system. Still, it remains a criminal offense with all that imports for the dignity of the persons charged. The petitioners will bear on their record the history of their criminal convictions. Just this Term we rejected various challenges to state laws requiring the registration of sex offenders. Smith v. Doe, 538 U. S. 84 (2003); Connecticut Dept. of Public Safety v. Doe, 538 U. S. 1 (2003). We are advised that if Texas convicted an adult for private, consensual homosexual conduct under the statute here in question the convicted person would come within the registration laws of at least four States were he or she to be subject to their jurisdiction. Pet. for Cert. 13, and n. 12 (citing Idaho Code §§ 18-8301 to 18-8326 (Cum. Supp. 2002); La. Code Crim. Proc. Ann. §§ 15:540-15:549

[539 U.S. 576]

(West 2003); Miss. Code Ann. §§ 45-33-21 to 45-33-57 (Lexis 2003); S. C. Code Ann. §§ 23-3-400 to 23-3-490 (West 2002)). This underscores the consequential nature of the punishment and the state-sponsored condemnation attendant to the criminal prohibition. Furthermore, the Texas criminal conviction carries with it the other collateral consequences always following a conviction, such as notations on job application forms, to mention but one example.

        The foundations of Bowers have sustained serious erosion from our recent decisions in Casey and Romer. When our precedent has been thus weakened, criticism from other sources is of greater significance. In the United States criticism of Bowers has been substantial and continuing, disapproving of its reasoning in all respects, not just as to its historical assumptions. See, e. g., C. Fried, Order and Law: Arguing the Reagan Revolution—A Firsthand Account 81-84 (1991); R. Posner, Sex and Reason 341-350 (1992). The courts of five different States have declined to follow it in interpreting provisions in their own state constitutions parallel to the Due Process Clause of the Fourteenth Amendment, see Jegley v. Picado, 349 Ark. 600, 80 S. W. 3d 332 (2002); Powell v. State, 270 Ga. 327, 510 S. E. 2d 18, 24 (1998); Gryczan v. State, 283 Mont. 433, 942 P. 2d 112 (1997); Campbell v. Sundquist, 926 S. W. 2d 250 (Tenn. App. 1996); Commonwealth v. Wasson, 842 S. W. 2d 487 (Ky. 1992).

        To the extent Bowers relied on values we share with a wider civilization, it should be noted that the reasoning and holding in Bowers have been rejected elsewhere. The European Court of Human Rights has followed not Bowers but its own decision in Dudgeon v. United Kingdom. See P. G. & J. H. v. United Kingdom, App. No. 00044787/98, ¶ 56 (Eur. Ct. H. R., Sept. 25, 2001); Modinos v. Cyprus, 259 Eur. Ct. H. R. (1993); Norris v. Ireland, 142 Eur. Ct. H. R. (1988). Other nations, too, have taken action consistent with an affirmation of the protected right of homosexual adults to engage in intimate, consensual conduct. See Brief for Mary

[539 U.S. 577]

Robinson et al. as Amici Curiae 11-12. The right the petitioners seek in this case has been accepted as an integral part of human freedom in many other countries. There has been no showing that in this country the governmental interest in circumscribing personal choice is somehow more legitimate or urgent.

        The doctrine of stare decisis is essential to the respect accorded to the judgments of the Court and to the stability of the law. It is not, however, an inexorable command. Payne v. Tennessee, 501 U. S. 808, 828 (1991) ("Stare decisis is not an inexorable command; rather, it `is a principle of policy and not a mechanical formula of adherence to the latest decision' " (quoting Helvering v. Hallock, 309 U. S. 106, 119 (1940))). In Casey we noted that when a court is asked to overrule a precedent recognizing a constitutional liberty interest, individual or societal reliance on the existence of that liberty cautions with particular strength against reversing course. 505 U. S., at 855-856; see also id., at 844 ("Liberty finds no refuge in a jurisprudence of doubt"). The holding in Bowers, however, has not induced detrimental reliance comparable to some instances where recognized individual rights are involved. Indeed, there has been no individual or societal reliance on Bowers of the sort that could counsel against overturning its holding once there are compelling reasons to do so. Bowers itself causes uncertainty, for the precedents before and after its issuance contradict its central holding.

        The rationale of Bowers does not withstand careful analysis. In his dissenting opinion in Bowers Justice Stevens came to these conclusions:

        "Our prior cases make two propositions abundantly clear. First, the fact that the governing majority in a State has traditionally viewed a particular practice as immoral is not a sufficient reason for upholding a law prohibiting the practice; neither history nor tradition could save a law prohibiting miscegenation from constitutional

[539 U.S. 578]

attack. Second, individual decisions by married persons, concerning the intimacies of their physical relationship, even when not intended to produce offspring, are a form of `liberty' protected by the Due Process Clause of the Fourteenth Amendment. Moreover, this protection extends to intimate choices by unmarried as well as married persons." 478 U. S., at 216 (footnotes and citations omitted).

        JUSTICE STEVENS' analysis, in our view, should have been controlling in Bowers and should control here.

        Bowers was not correct when it was decided, and it is not correct today. It ought not to remain binding precedent. Bowers v. Hardwick should be and now is overruled.

        The present case does not involve minors. It does not involve persons who might be injured or coerced or who are situated in relationships where consent might not easily be refused. It does not involve public conduct or prostitution. It does not involve whether the government must give formal recognition to any relationship that homosexual persons seek to enter. The case does involve two adults who, with full and mutual consent from each other, engaged in sexual practices common to a homosexual lifestyle. The petitioners are entitled to respect for their private lives. The State cannot demean their existence or control their destiny by making their private sexual conduct a crime. Their right to liberty under the Due Process Clause gives them the full right to engage in their conduct without intervention of the government. "It is a promise of the Constitution that there is a realm of personal liberty which the government may not enter." Casey, supra, at 847. The Texas statute furthers no legitimate state interest which can justify its intrusion into the personal and private life of the individual.

        Had those who drew and ratified the Due Process Clauses of the Fifth Amendment or the Fourteenth Amendment known the components of liberty in its manifold possibilities, they might have been more specific. They did not presume

[539 U.S. 579]

to have this insight. They knew times can blind us to certain truths and later generations can see that laws once thought necessary and proper in fact serve only to oppress. As the Constitution endures, persons in every generation can invoke its principles in their own search for greater freedom.

        The judgment of the Court of Appeals for the Texas Fourteenth District is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.

        It is so ordered.

---------------

Notes:

* Briefs of amici curiae urging reversal were filed for the Alliance of Baptists et al. by Robert A. Long, Jr., and Thomas L. Cubbage III; for the American Psychological Association et al. by David W. Ogden, Paul R. Q. Wolfson, Richard G. Taranto, Nathalie F. P. Gilfoyle, and Carolyn I. Polowy; for the American Public Health Association et al. by Jeffrey S. Trachtman and Norman C. Simon; for the Cato Institute by Robert A. Levy; for Constitutional Law Professors by Pamela S. Karlan and William B. Rubenstein; for the Human Rights Campaign et al. by Walter Dellinger, Pamela Harris, and Jonathan D. Hacker; for the Log Cabin Republicans et al. by C. Martin Meekins; for the NOW Legal Defense and Education Fund by David C. Codell, Laura W. Brill, and Wendy R. Weiser; for Professors of History by Roy T. Englert, Jr., Alan Untereiner, and Sherri Lynn Wolson; for the Republican Unity Coalition et al. by Erik S. Jaffe; and for Mary Robinson et al. by Harold Hongju Koh and Joseph F. Tringali.

        Briefs of amici curiae urging affirmance were filed for the State of Alabama et al. by William H. Pryor, Jr., Attorney General of Alabama, Nathan A. Forrester, Solicitor General, and George M. Weaver, and by the Attorneys General for their respective States as follows: Henry D. McMaster of South Carolina and Mark L. Shurtleff of Utah; for Agudath Israel of America by David Zwiebel; for the American Center for Law and Justice by Jay Alan Sekulow, Stuart J. Roth, Colby M. May, James M. Henderson, Sr., Joel H. Thornton, and Walter M. Weber; for the American Family Association, Inc., et al. by Stephen M. Crampton, Brian Fahling, and Michael J. DePrimo; for the Center for Arizona Policy et al. by Len L. Munsil; for the Center for Law and Justice International by Thomas Patrick Monaghan and John P. Tuskey; for the Center for Marriage Law by Vincent P. McCarthy and Lynn D. Wardle; for the Center for the Original Intent of the Constitution by Michael P. Farris and Jordan W. Lorence; for Concerned Women for America by Janet M. LaRue; for the Family Research Council, Inc., by Robert P. George; for First Principles, Inc., by Ronald D. Ray; for Liberty Counsel by Mathew D. Staver and Rena M. Lindevaldsen; for the Pro Family Law Center et al. by Richard D. Ackerman and Gary G. Kreep; for Public Advocate of the United States et al. by Herbert W. Titus and William J. Olson; for the Texas Eagle Forum et al. by Teresa Stanton Collett; for Texas Legislator Warren Chisum et al. by Kelly Shackelford and Scott Roberts; for the Texas Physicians Resource Council et al. by Glen Lavy; and for United Families International by Paul Benjamin Linton.

        Briefs of amici curiae were filed for the American Bar Association by Alfred P. Carlton, Jr., Ruth N. Borenstein, and Beth S. Brinkmann; for the American Civil Liberties Union et al. by Laurence H. Tribe, James D. Esseks, Steven R. Shapiro, and Matthew A. Coles; for the Institute for Justice by William H. Mellor, Clint Bolick, Dana Berliner, and Randy E. Barnett; and for the National Lesbian and Gay Law Association et al. by Chai R. Feldblum, J. Paul Oetken, and Scott Ruskay-Kidd.

---------------

        JUSTICE O'CONNOR, concurring in the judgment.

        The Court today overrules Bowers v. Hardwick, 478 U. S. 186 (1986). I joined Bowers, and do not join the Court in overruling it. Nevertheless, I agree with the Court that Texas' statute banning same-sex sodomy is unconstitutional. See Tex. Penal Code Ann. § 21.06 (2003). Rather than relying on the substantive component of the Fourteenth Amendment's Due Process Clause, as the Court does, I base my conclusion on the Fourteenth Amendment's Equal Protection Clause.

        The Equal Protection Clause of the Fourteenth Amendment "is essentially a direction that all persons similarly situated should be treated alike." Cleburne v. Cleburne Living Center, Inc., 473 U. S. 432, 439 (1985); see also Plyler v. Doe, 457 U. S. 202, 216 (1982). Under our rational basis standard of review, "legislation is presumed to be valid and will be sustained if the classification drawn by the statute is rationally related to a legitimate state interest." Cleburne v. Cleburne Living Center, supra, at 440; see also Department of Agriculture v. Moreno, 413 U. S. 528, 534 (1973); Romer v. Evans, 517 U. S. 620, 632-633 (1996); Nordlinger v. Hahn, 505 U. S. 1, 11-12 (1992).

        Laws such as economic or tax legislation that are scrutinized under rational basis review normally pass constitutional muster, since "the Constitution presumes that even improvident decisions will eventually be rectified by the

[539 U.S. 580]

democratic processes." Cleburne v. Cleburne Living Center, supra, at 440; see also Fitzgerald v. Racing Assn. of Central Iowa, ante, p. 103; Williamson v. Lee Optical of Okla., Inc., 348 U. S. 483 (1955). We have consistently held, however, that some objectives, such as "a bare . . . desire to harm a politically unpopular group," are not legitimate state interests. Department of Agriculture v. Moreno, supra, at 534. See also Cleburne v. Cleburne Living Center, supra, at 446-447; Romer v. Evans, supra, at 632. When a law exhibits such a desire to harm a politically unpopular group, we have applied a more searching form of rational basis review to strike down such laws under the Equal Protection Clause.

        We have been most likely to apply rational basis review to hold a law unconstitutional under the Equal Protection Clause where, as here, the challenged legislation inhibits personal relationships. In Department of Agriculture v. Moreno, for example, we held that a law preventing those households containing an individual unrelated to any other member of the household from receiving food stamps violated equal protection because the purpose of the law was to "`discriminate against hippies.'" 413 U. S., at 534. The asserted governmental interest in preventing food stamp fraud was not deemed sufficient to satisfy rational basis review. Id., at 535-538. In Eisenstadt v. Baird, 405 U. S. 438, 447-455 (1972), we refused to sanction a law that discriminated between married and unmarried persons by prohibiting the distribution of contraceptives to single persons. Likewise, in Cleburne v. Cleburne Living Center, supra, we held that it was irrational for a State to require a home for the mentally disabled to obtain a special use permit when other residences—like fraternity houses and apartment buildings—did not have to obtain such a permit. And in Romer v. Evans, we disallowed a state statute that "impos[ed] a broad and undifferentiated disability on a single named group"—specifically, homosexuals. 517 U. S., at 632.

[539 U.S. 581]

        The statute at issue here makes sodomy a crime only if a person "engages in deviate sexual intercourse with another individual of the same sex." Tex. Penal Code Ann. § 21.06(a) (2003). Sodomy between opposite-sex partners, however, is not a crime in Texas. That is, Texas treats the same conduct differently based solely on the participants. Those harmed by this law are people who have a same-sex sexual orientation and thus are more likely to engage in behavior prohibited by § 21.06.

        The Texas statute makes homosexuals unequal in the eyes of the law by making particular conduct—and only that conduct—subject to criminal sanction. It appears that prosecutions under Texas' sodomy law are rare. See State v. Morales, 869 S. W. 2d 941, 943 (Tex. 1994) (noting in 1994 that § 21.06 "has not been, and in all probability will not be, enforced against private consensual conduct between adults"). This case shows, however, that prosecutions under § 21.06 do occur. And while the penalty imposed on petitioners in this case was relatively minor, the consequences of conviction are not. It appears that petitioners' convictions, if upheld, would disqualify them from or restrict their ability to engage in a variety of professions, including medicine, athletic training, and interior design. See, e. g., Tex. Occ. Code Ann. § 164.051(a)(2)(B) (2003 Pamphlet) (physician); § 451.251(a)(1) (athletic trainer); § 1053.252(2) (interior designer). Indeed, were petitioners to move to one of four States, their convictions would require them to register as sex offenders to local law enforcement. See, e. g., Idaho Code § 18-8304 (Cum. Supp. 2002); La. Stat. Ann. § 15:542 (West Cum. Supp. 2003); Miss. Code Ann. § 45-33-25 (West 2003); S. C. Code Ann. § 23-3-430 (West Cum. Supp. 2002); cf. ante, at 575-576.

        And the effect of Texas' sodomy law is not just limited to the threat of prosecution or consequence of conviction. Texas' sodomy law brands all homosexuals as criminals, thereby making it more difficult for homosexuals to be treated in the same manner as everyone else. Indeed, Texas

[539 U.S. 582]

itself has previously acknowledged the collateral effects of the law, stipulating in a prior challenge to this action that the law "legally sanctions discrimination against [homosexuals] in a variety of ways unrelated to the criminal law," including in the areas of "employment, family issues, and housing." State v. Morales, 826 S. W. 2d 201, 203 (Tex. App. 1992).

        Texas attempts to justify its law, and the effects of the law, by arguing that the statute satisfies rational basis review because it furthers the legitimate governmental interest of the promotion of morality. In Bowers, we held that a state law criminalizing sodomy as applied to homosexual couples did not violate substantive due process. We rejected the argument that no rational basis existed to justify the law, pointing to the government's interest in promoting morality. 478 U. S., at 196. The only question in front of the Court in Bowers was whether the substantive component of the Due Process Clause protected a right to engage in homosexual sodomy. Id., at 188, n. 2. Bowers did not hold that moral disapproval of a group is a rational basis under the Equal Protection Clause to criminalize homosexual sodomy when heterosexual sodomy is not punished.

        This case raises a different issue than Bowers: whether, under the Equal Protection Clause, moral disapproval is a legitimate state interest to justify by itself a statute that bans homosexual sodomy, but not heterosexual sodomy. It is not. Moral disapproval of this group, like a bare desire to harm the group, is an interest that is insufficient to satisfy rational basis review under the Equal Protection Clause. See, e. g., Department of Agriculture v. Moreno, 413 U. S., at 534; Romer v. Evans, 517 U. S., at 634-635. Indeed, we have never held that moral disapproval, without any other asserted state interest, is a sufficient rationale under the Equal Protection Clause to justify a law that discriminates among groups of persons.

[539 U.S. 583]

        Moral disapproval of a group cannot be a legitimate governmental interest under the Equal Protection Clause because legal classifications must not be "drawn for the purpose of disadvantaging the group burdened by the law." Id., at 633. Texas' invocation of moral disapproval as a legitimate state interest proves nothing more than Texas' desire to criminalize homosexual sodomy. But the Equal Protection Clause prevents a State from creating "a classification of persons undertaken for its own sake." Id., at 635. And because Texas so rarely enforces its sodomy law as applied to private, consensual acts, the law serves more as a statement of dislike and disapproval against homosexuals than as a tool to stop criminal behavior. The Texas sodomy law "raise[s] the inevitable inference that the disadvantage imposed is born of animosity toward the class of persons affected." Id., at 634.

        Texas argues, however, that the sodomy law does not discriminate against homosexual persons. Instead, the State maintains that the law discriminates only against homosexual conduct. While it is true that the law applies only to conduct, the conduct targeted by this law is conduct that is closely correlated with being homosexual. Under such circumstances, Texas' sodomy law is targeted at more than conduct. It is instead directed toward gay persons as a class. "After all, there can hardly be more palpable discrimination against a class than making the conduct that defines the class criminal." Id., at 641 (SCALIA, J., dissenting) (internal quotation marks omitted). When a State makes homosexual conduct criminal, and not "deviate sexual intercourse" committed by persons of different sexes, "that declaration in and of itself is an invitation to subject homosexual persons to discrimination both in the public and in the private spheres." Ante, at 575.

        Indeed, Texas law confirms that the sodomy statute is directed toward homosexuals as a class. In Texas, calling a person a homosexual is slander per se because the word "homosexual"

[539 U.S. 584]

"impute[s] the commission of a crime." Plumley v. Landmark Chevrolet, Inc., 122 F. 3d 308, 310 (CA5 1997) (applying Texas law); see also Head v. Newton, 596 S. W. 2d 209, 210 (Tex. App. 1980). The State has admitted that because of the sodomy law, being homosexual carries the presumption of being a criminal. See State v. Morales, 826 S. W. 2d, at 202-203 ("[T]he statute brands lesbians and gay men as criminals and thereby legally sanctions discrimination against them in a variety of ways unrelated to the criminal law"). Texas' sodomy law therefore results in discrimination against homosexuals as a class in an array of areas outside the criminal law. See ibid. In Romer v. Evans, we refused to sanction a law that singled out homosexuals "for disfavored legal status." 517 U. S., at 633. The same is true here. The Equal Protection Clause "`neither knows nor tolerates classes among citizens.'" Id., at 623 (quoting Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (Harlan, J., dissenting)).

        A State can of course assign certain consequences to a violation of its criminal law. But the State cannot single out one identifiable class of citizens for punishment that does not apply to everyone else, with moral disapproval as the only asserted state interest for the law. The Texas sodomy statute subjects homosexuals to "a lifelong penalty and stigma. A legislative classification that threatens the creation of an underclass . . . cannot be reconciled with" the Equal Protection Clause. Plyler v. Doe, 457 U. S., at 239 (Powell, J., concurring).

        Whether a sodomy law that is neutral both in effect and application, see Yick Wo v. Hopkins, 118 U. S. 356 (1886), would violate the substantive component of the Due Process Clause is an issue that need not be decided today. I am confident, however, that so long as the Equal Protection Clause requires a sodomy law to apply equally to the private consensual conduct of homosexuals and heterosexuals alike, such a

[539 U.S. 585]

law would not long stand in our democratic society. In the words of Justice Jackson:

        "The framers of the Constitution knew, and we should not forget today, that there is no more effective practical guaranty against arbitrary and unreasonable government than to require that the principles of law which officials would impose upon a minority be imposed generally. Conversely, nothing opens the door to arbitrary action so effectively as to allow those officials to pick and choose only a few to whom they will apply legislation and thus to escape the political retribution that might be visited upon them if larger numbers were affected." Railway Express Agency, Inc. v. New York, 336 U. S. 106, 112-113 (1949) (concurring opinion).

        That this law as applied to private, consensual conduct is unconstitutional under the Equal Protection Clause does not mean that other laws distinguishing between heterosexuals and homosexuals would similarly fail under rational basis review. Texas cannot assert any legitimate state interest here, such as national security or preserving the traditional institution of marriage. Unlike the moral disapproval of same-sex relations—the asserted state interest in this case— other reasons exist to promote the institution of marriage beyond mere moral disapproval of an excluded group.

        A law branding one class of persons as criminal based solely on the State's moral disapproval of that class and the conduct associated with that class runs contrary to the values of the Constitution and the Equal Protection Clause, under any standard of review. I therefore concur in the Court's judgment that Texas' sodomy law banning "deviate sexual intercourse" between consenting adults of the same sex, but not between consenting adults of different sexes, is unconstitutional.

[539 U.S. 586]

        JUSTICE SCALIA, with whom THE CHIEF JUSTICE and JUSTICE THOMAS join, dissenting.

        "Liberty finds no refuge in a jurisprudence of doubt." Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 844 (1992). That was the Court's sententious response, barely more than a decade ago, to those seeking to overrule Roe v. Wade, 410 U. S. 113 (1973). The Court's response today, to those who have engaged in a 17-year crusade to overrule Bowers v. Hardwick, 478 U. S. 186 (1986), is very different. The need for stability and certainty presents no barrier.

        Most of the rest of today's opinion has no relevance to its actual holding—that the Texas statute "furthers no legitimate state interest which can justify" its application to petitioners under rational-basis review. Ante, at 578 (overruling Bowers to the extent it sustained Georgia's antisodomy statute under the rational-basis test). Though there is discussion of "fundamental proposition[s]," ante, at 565, and "fundamental decisions," ibid., nowhere does the Court's opinion declare that homosexual sodomy is a "fundamental right" under the Due Process Clause; nor does it subject the Texas law to the standard of review that would be appropriate (strict scrutiny) if homosexual sodomy were a "fundamental right." Thus, while overruling the outcome of Bowers, the Court leaves strangely untouched its central legal conclusion: "[R]espondent would have us announce . . . a fundamental right to engage in homosexual sodomy. This we are quite unwilling to do." 478 U. S., at 191. Instead the Court simply describes petitioners' conduct as "an exercise of their liberty"—which it undoubtedly is—and proceeds to apply an unheard-of form of rational-basis review that will have far-reaching implications beyond this case. Ante, at 564.

        I begin with the Court's surprising readiness to reconsider a decision rendered a mere 17 years ago in Bowers v. Hardwick.

[539 U.S. 587]

I do not myself believe in rigid adherence to stare decisis in constitutional cases; but I do believe that we should be consistent rather than manipulative in invoking the doctrine. Today's opinions in support of reversal do not bother to distinguish—or indeed, even bother to mention— the paean to stare decisis coauthored by three Members of today's majority in Planned Parenthood v. Casey. There, when stare decisis meant preservation of judicially invented abortion rights, the widespread criticism of Roe was strong reason to reaffirm it:

        "Where, in the performance of its judicial duties, the Court decides a case in such a way as to resolve the sort of intensely divisive controversy reflected in Roe[,] . . . its decision has a dimension that the resolution of the normal case does not carry. . . . [T]o overrule under fire in the absence of the most compelling reason . . . would subvert the Court's legitimacy beyond any serious question." 505 U. S., at 866-867.

        Today, however, the widespread opposition to Bowers, a decision resolving an issue as "intensely divisive" as the issue in Roe, is offered as a reason in favor of overruling it. See ante, at 576-577. Gone, too, is any "enquiry" (of the sort conducted in Casey) into whether the decision sought to be overruled has "proven `unworkable,'" Casey, supra, at 855.

        Today's approach to stare decisis invites us to overrule an erroneously decided precedent (including an "intensely divisive" decision) if: (1) its foundations have been "ero[ded]" by subsequent decisions, ante, at 576; (2) it has been subject to "substantial and continuing" criticism, ibid.; and (3) it has not induced "individual or societal reliance" that counsels against overturning, ante, at 577. The problem is that Roe itself—which today's majority surely has no disposition to overrule—satisfies these conditions to at least the same degree as Bowers.

[539 U.S. 588]

        (1) A preliminary digressive observation with regard to the first factor: The Court's claim that Planned Parenthood v. Casey, supra, "casts some doubt" upon the holding in Bowers (or any other case, for that matter) does not withstand analysis. Ante, at 571. As far as its holding is concerned, Casey provided a less expansive right to abortion than did Roe, which was already on the books when Bowers was decided. And if the Court is referring not to the holding of Casey, but to the dictum of its famed sweet-mystery-of-life passage, ante, at 574 ("`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'"): That "casts some doubt" upon either the totality of our jurisprudence or else (presumably the right answer) nothing at all. I have never heard of a law that attempted to restrict one's "right to define" certain concepts; and if the passage calls into question the government's power to regulate actions based on one's self-defined "concept of existence, etc.," it is the passage that ate the rule of law.

        I do not quarrel with the Court's claim that Romer v. Evans, 517 U. S. 620 (1996), "eroded" the "foundations" of Bowers' rational-basis holding. See Romer, supra, at 640-643 (Scalia, J., dissenting). But Roe and Casey have been equally "eroded" by Washington v. Glucksberg, 521 U. S. 702, 721 (1997), which held that only fundamental rights which are "`deeply rooted in this Nation's history and tradition' " qualify for anything other than rational-basis scrutiny under the doctrine of "substantive due process." Roe and Casey, of course, subjected the restriction of abortion to heightened scrutiny without even attempting to establish that the freedom to abort was rooted in this Nation's tradition.

        (2) Bowers, the Court says, has been subject to "substantial and continuing [criticism], disapproving of its reasoning in all respects, not just as to its historical assumptions." Ante, at 576. Exactly what those nonhistorical criticisms are, and whether the Court even agrees with them, are left

[539 U.S. 589]

unsaid, although the Court does cite two books. See ibid. (citing C. Fried, Order and Law: Arguing the Reagan Revolution —A Firsthand Account 81-84 (1991); R. Posner, Sex and Reason 341-350 (1992)).¹ Of course, Roe too (and by extension Casey) had been (and still is) subject to unrelenting criticism, including criticism from the two commentators cited by the Court today. See Fried, supra, at 75 ("Roe was a prime example of twisted judging"); Posner, supra, at 337 ("[The Court's] opinion in Roe... fails to measure up to professional expectations regarding judicial opinions"); Posner, Judicial Opinion Writing, 62 U. Chi. L. Rev. 1421, 1434 (1995) (describing the opinion in Roe as an "embarrassing performanc[e]").

        (3) That leaves, to distinguish the rock-solid, unamendable disposition of Roe from the readily overrulable Bowers, only the third factor. "[T]here has been," the Court says, "no individual or societal reliance on Bowers of the sort that could counsel against overturning its holding...." Ante, at 577. It seems to me that the "societal reliance" on the principles confirmed in Bowers and discarded today has been overwhelming. Countless judicial decisions and legislative enactments have relied on the ancient proposition that a governing majority's belief that certain sexual behavior is "immoral and unacceptable" constitutes a rational basis for regulation. See, e. g., Williams v. Pryor, 240 F. 3d 944, 949 (CA11 2001) (citing Bowers in upholding Alabama's prohibition on the sale of sex toys on the ground that "[t]he crafting and safeguarding of public morality . . . indisputably is a legitimate government interest under rational basis scrutiny"); Milner v. Apfel, 148 F. 3d 812, 814 (CA7 1998) (citing Bowers for the proposition that "[l]egislatures are permitted to legislate with regard to morality . . . rather than confined

[539 U.S. 590]

to preventing demonstrable harms"); Holmes v. California Army National Guard, 124 F. 3d 1126, 1136 (CA9 1997) (relying on Bowers in upholding the federal statute and regulations banning from military service those who engage in homosexual conduct); Owens v. State, 352 Md. 663, 683, 724 A. 2d 43, 53 (1999) (relying on Bowers in holding that "a person has no constitutional right to engage in sexual intercourse, at least outside of marriage"); Sherman v. Henry, 928 S. W. 2d 464, 469-473 (Tex. 1996) (relying on Bowers in rejecting a claimed constitutional right to commit adultery). We ourselves relied extensively on Bowers when we concluded, in Barnes v. Glen Theatre, Inc., 501 U. S. 560, 569 (1991), that Indiana's public indecency statute furthered "a substantial government interest in protecting order and morality," ibid. (plurality opinion); see also id., at 575 (Scalia, J., concurring in judgment). State laws against bigamy, same-sex marriage, adult incest, prostitution, masturbation, adultery, fornication, bestiality, and obscenity are likewise sustainable only in light of Bowers' validation of laws based on moral choices. Every single one of these laws is called into question by today's decision; the Court makes no effort to cabin the scope of its decision to exclude them from its holding. See ante, at 572 (noting "an emerging awareness that liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex" (emphasis added)). The impossibility of distinguishing homosexuality from other traditional "morals" offenses is precisely why Bowers rejected the rational-basis challenge. "The law," it said, "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." 478 U. S., at 196.²

[539 U.S. 591]

        What a massive disruption of the current social order, therefore, the overruling of Bowers entails. Not so the overruling of Roe, which would simply have restored the regime that existed for centuries before 1973, in which the permissibility of, and restrictions upon, abortion were determined legislatively State by State. Casey, however, chose to base its stare decisis determination on a different "sort" of reliance. "[P]eople," it said, "have organized intimate relationships and made choices that define their views of themselves and their places in society, in reliance on the availability of abortion in the event that contraception should fail." 505 U. S., at 856. This falsely assumes that the consequence of overruling Roe would have been to make abortion unlawful. It would not; it would merely have permitted

[539 U.S. 592]

the States to do so. Many States would unquestionably have declined to prohibit abortion, and others would not have prohibited it within six months (after which the most significant reliance interests would have expired). Even for persons in States other than these, the choice would not have been between abortion and childbirth, but between abortion nearby and abortion in a neighboring State.

        To tell the truth, it does not surprise me, and should surprise no one, that the Court has chosen today to revise the standards of stare decisis set forth in Casey. It has thereby exposed Casey's extraordinary deference to precedent for the result-oriented expedient that it is.

        Having decided that it need not adhere to stare decisis, the Court still must establish that Bowers was wrongly decided and that the Texas statute, as applied to petitioners, is unconstitutional.

        Texas Penal Code Ann. § 21.06(a) (2003) undoubtedly imposes constraints on liberty. So do laws prohibiting prostitution, recreational use of heroin, and, for that matter, working more than 60 hours per week in a bakery. But there is no right to "liberty" under the Due Process Clause, though today's opinion repeatedly makes that claim. Ante, at 567 ("The liberty protected by the Constitution allows homosexual persons the right to make this choice"); ante, at 574 ("`These matters . . . are central to the liberty protected by the Fourteenth Amendment' "); ante, at 578 ("Their right to liberty under the Due Process Clause gives them the full right to engage in their conduct without intervention of the government"). The Fourteenth Amendment expressly allows States to deprive their citizens of "liberty," so long as "due process of law" is provided:

        "No state shall . . . deprive any person of life, liberty, or property, without due process of law." Amdt. 14 (emphasis added).

[539 U.S. 593]

        Our opinions applying the doctrine known as "substantive due process" hold that the Due Process Clause prohibits States from infringing fundamental liberty interests, unless the infringement is narrowly tailored to serve a compelling state interest. Washington v. Glucksberg, 521 U. S., at 721. We have held repeatedly, in cases the Court today does not overrule, that only fundamental rights qualify for this so-called "heightened scrutiny" protection—that is, rights which are "`deeply rooted in this Nation's history and tradition,'" ibid. See Reno v. Flores, 507 U. S. 292, 303 (1993) (fundamental liberty interests must be "so rooted in the traditions and conscience of our people as to be ranked as fundamental" (internal quotation marks and citations omitted)); United States v. Salerno, 481 U. S. 739, 751 (1987) (same). See also Michael H. v. Gerald D., 491 U. S. 110, 122 (1989) ("[W]e have insisted not merely that the interest denominated as a `liberty' be `fundamental' . . . but also that it be an interest traditionally protected by our society"); Moore v. East Cleveland, 431 U. S. 494, 503 (1977) (plurality opinion); Meyer v. Nebraska, 262 U. S. 390, 399 (1923) (Fourteenth Amendment protects "those privileges long recognized at common law as essential to the orderly pursuit of happiness by free men" (emphasis added)).³ All other liberty interests may be abridged or abrogated pursuant to a validly enacted state law if that law is rationally related to a legitimate state interest.

[539 U.S. 594]

        Bowers held, first, that criminal prohibitions of homosexual sodomy are not subject to heightened scrutiny because they do not implicate a "fundamental right" under the Due Process Clause, 478 U. S., at 191-194. Noting that "[p]roscriptions against that conduct have ancient roots," id., at 192, that "[s]odomy was a criminal offense at common law and was forbidden by the laws of the original 13 States when they ratified the Bill of Rights," ibid., and that many States had retained their bans on sodomy, id., at 193, Bowers concluded that a right to engage in homosexual sodomy was not "`deeply rooted in this Nation's history and tradition,'" id., at 192.

        The Court today does not overrule this holding. Not once does it describe homosexual sodomy as a "fundamental right" or a "fundamental liberty interest," nor does it subject the Texas statute to strict scrutiny. Instead, having failed to establish that the right to homosexual sodomy is "`deeply rooted in this Nation's history and tradition,'" the Court concludes that the application of Texas's statute to petitioners' conduct fails the rational-basis test, and overrules Bowers' holding to the contrary, see id., at 196. "The Texas statute furthers no legitimate state interest which can justify its intrusion into the personal and private life of the individual." Ante, at 578.

        I shall address that rational-basis holding presently. First, however, I address some aspersions that the Court casts upon Bowers' conclusion that homosexual sodomy is not a "fundamental right"—even though, as I have said, the Court does not have the boldness to reverse that conclusion.

        The Court's description of "the state of the law" at the time of Bowers only confirms that Bowers was right. Ante, at 566. The Court points to Griswold v. Connecticut, 381 U. S. 479, 481-482 (1965). But that case expressly disclaimed any reliance on the doctrine of "substantive due

[539 U.S. 595]

process," and grounded the so-called "right to privacy" in penumbras of constitutional provisions other than the Due Process Clause. Eisenstadt v. Baird, 405 U. S. 438 (1972), likewise had nothing to do with "substantive due process"; it invalidated a Massachusetts law prohibiting the distribution of contraceptives to unmarried persons solely on the basis of the Equal Protection Clause. Of course Eisenstadt contains well-known dictum relating to the "right to privacy," but this referred to the right recognized in Griswold —a right penumbral to the specific guarantees in the Bill of Rights, and not a "substantive due process" right.

        Roe v. Wade recognized that the right to abort an unborn child was a "fundamental right" protected by the Due Process Clause. 410 U. S., at 155. The Roe Court, however, made no attempt to establish that this right was "`deeply rooted in this Nation's history and tradition' "; instead, it based its conclusion that "the Fourteenth Amendment's concept of personal liberty . . . is broad enough to encompass a woman's decision whether or not to terminate her pregnancy" on its own normative judgment that antiabortion laws were undesirable. See id., at 153. We have since rejected Roe's holding that regulations of abortion must be narrowly tailored to serve a compelling state interest, see Planned Parenthood v. Casey, 505 U. S., at 876 ( joint opinion of O'CONNOR, KENNEDY, and SOUTER, JJ.); id., at 951-953 (Rehnquist, C. J., concurring in judgment in part and dissenting in part)—and thus, by logical implication, Roe's holding that the right to abort an unborn child is a "fundamental right." See 505 U. S., at 843-912 ( joint opinion of O'CONNOR, KENNEDY, and SOUTER, JJ.) (not once describing abortion as a "fundamental right" or a "fundamental liberty interest").

        After discussing the history of antisodomy laws, ante, at 568-571, the Court proclaims that, "it should be noted that there is no longstanding history in this country of laws directed at homosexual conduct as a distinct matter," ante,

[539 U.S. 596]

at 568. This observation in no way casts into doubt the "definitive [historical] conclusio[n]," ibid., on which Bowers relied: that our Nation has a longstanding history of laws prohibiting sodomy in general—regardless of whether it was performed by same-sex or opposite-sex couples:

        "It is obvious to us that neither of these formulations would extend a fundamental right to homosexuals to engage in acts of consensual sodomy. Proscriptions against that conduct have ancient roots. Sodomy was a criminal offense at common law and was forbidden by the laws of the original 13 States when they ratified the Bill of Rights. In 1868, when the Fourteenth Amendment was ratified, all but 5 of the 37 States in the Union had criminal sodomy laws. In fact, until 1961, all 50 States outlawed sodomy, and today, 24 States and the District of Columbia continue to provide criminal penalties for sodomy performed in private and between consenting adults. Against this background, to claim that a right to engage in such conduct is `deeply rooted in this Nation's history and tradition' or `implicit in the concept of ordered liberty' is, at best, facetious." 478 U. S., at 192-194 (citations and footnotes omitted; emphasis added).

        It is (as Bowers recognized) entirely irrelevant whether the laws in our long national tradition criminalizing homosexual sodomy were "directed at homosexual conduct as a distinct matter." Ante, at 568. Whether homosexual sodomy was prohibited by a law targeted at same-sex sexual relations or by a more general law prohibiting both homosexual and heterosexual sodomy, the only relevant point is that it was criminalized—which suffices to establish that homosexual sodomy is not a right "deeply rooted in our Nation's history and tradition." The Court today agrees that homosexual sodomy was criminalized and thus does not dispute the facts on which Bowers actually relied.

[539 U.S. 597]

        Next the Court makes the claim, again unsupported by any citations, that "[l]aws prohibiting sodomy do not seem to have been enforced against consenting adults acting in private." Ante, at 569. The key qualifier here is "acting in private"—since the Court admits that sodomy laws were enforced against consenting adults (although the Court contends that prosecutions were "infrequen[t]," ibid.). I do not know what "acting in private" means; surely consensual sodomy, like heterosexual intercourse, is rarely performed on stage. If all the Court means by "acting in private" is "on private premises, with the doors closed and windows covered," it is entirely unsurprising that evidence of enforcement would be hard to come by. (Imagine the circumstances that would enable a search warrant to be obtained for a residence on the ground that there was probable cause to believe that consensual sodomy was then and there occurring.) Surely that lack of evidence would not sustain the proposition that consensual sodomy on private premises with the doors closed and windows covered was regarded as a "fundamental right," even though all other consensual sodomy was criminalized. There are 203 prosecutions for consensual, adult homosexual sodomy reported in the West Reporting system and official state reporters from the years 1880-1995. See W. Eskridge, Gaylaw: Challenging the Apartheid of the Closet 375 (1999) (hereinafter Gaylaw). There are also records of 20 sodomy prosecutions and 4 executions during the colonial period. J. Katz, Gay/Lesbian Almanac 29, 58, 663 (1983). Bowers' conclusion that homosexual sodomy is not a fundamental right "deeply rooted in this Nation's history and tradition" is utterly unassailable.

        Realizing that fact, the Court instead says: "[W]e think that our laws and traditions in the past half century are of most relevance here. These references show an emerging awareness that liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex." Ante, at 571-572 (emphasis

[539 U.S. 598]

added). Apart from the fact that such an "emerging awareness" does not establish a "fundamental right," the statement is factually false. States continue to prosecute all sorts of crimes by adults "in matters pertaining to sex": prostitution, adult incest, adultery, obscenity, and child pornography. Sodomy laws, too, have been enforced "in the past half century," in which there have been 134 reported cases involving prosecutions for consensual, adult, homosexual sodomy. Gaylaw 375. In relying, for evidence of an "emerging recognition," upon the American Law Institute's 1955 recommendation not to criminalize "`consensual sexual relations conducted in private,'" ante, at 572, the Court ignores the fact that this recommendation was "a point of resistance in most of the states that considered adopting the Model Penal Code." Gaylaw 159.

        In any event, an "emerging awareness" is by definition not "deeply rooted in this Nation's history and tradition[s]," as we have said "fundamental right" status requires. Constitutional entitlements do not spring into existence because some States choose to lessen or eliminate criminal sanctions on certain behavior. Much less do they spring into existence, as the Court seems to believe, because foreign nations decriminalize conduct. The Bowers majority opinion never relied on "values we share with a wider civilization," ante, at 576, but rather rejected the claimed right to sodomy on the ground that such a right was not "`deeply rooted in this Nation's history and tradition,'" 478 U. S., at 193-194 (emphasis added). Bowers' rational-basis holding is likewise devoid of any reliance on the views of a "wider civilization," see id., at 196. The Court's discussion of these foreign views (ignoring, of course, the many countries that have retained criminal prohibitions on sodomy) is therefore meaningless dicta. Dangerous dicta, however, since "this Court... should not impose foreign moods, fads, or fashions on Americans." Foster v. Florida, 537 U. S. 990, n. (2002) (Thomas, J., concurring in denial of certiorari).

[539 U.S. 599]

        I turn now to the ground on which the Court squarely rests its holding: the contention that there is no rational basis for the law here under attack. This proposition is so out of accord with our jurisprudence—indeed, with the jurisprudence of any society we know—that it requires little discussion.

        The Texas statute undeniably seeks to further the belief of its citizens that certain forms of sexual behavior are "immoral and unacceptable," Bowers, supra, at 196—the same interest furthered by criminal laws against fornication, bigamy, adultery, adult incest, bestiality, and obscenity. Bowers held that this was a legitimate state interest. The Court today reaches the opposite conclusion. The Texas statute, it says, "furthers no legitimate state interest which can justify its intrusion into the personal and private life of the individual," ante, at 578 (emphasis added). The Court embraces instead Justice Stevens' declaration in his Bowers dissent, that "`the fact that the governing majority in a State has traditionally viewed a particular practice as immoral is not a sufficient reason for upholding a law prohibiting the practice,'" ante, at 577. This effectively decrees the end of all morals legislation. If, as the Court asserts, the promotion of majoritarian sexual morality is not even a legitimate state interest, none of the above-mentioned laws can survive rational-basis review.

        Finally, I turn to petitioners' equal-protection challenge, which no Member of the Court save Justice O'Connor, ante, at 579 (opinion concurring in judgment), embraces: On its face § 21.06(a) applies equally to all persons. Men and women, heterosexuals and homosexuals, are all subject to its prohibition of deviate sexual intercourse with someone of the same sex. To be sure, § 21.06 does distinguish between the sexes insofar as concerns the partner with whom the sexual

[539 U.S. 600]

acts are performed: men can violate the law only with other men, and women only with other women. But this cannot itself be a denial of equal protection, since it is precisely the same distinction regarding partner that is drawn in state laws prohibiting marriage with someone of the same sex while permitting marriage with someone of the opposite sex.

        The objection is made, however, that the antimiscegenation laws invalidated in Loving v. Virginia, 388 U. S. 1, 8 (1967), similarly were applicable to whites and blacks alike, and only distinguished between the races insofar as the partner was concerned. In Loving, however, we correctly applied heightened scrutiny, rather than the usual rational-basis review, because the Virginia statute was "designed to maintain White Supremacy." Id., at 6, 11. A racially discriminatory purpose is always sufficient to subject a law to strict scrutiny, even a facially neutral law that makes no mention of race. See Washington v. Davis, 426 U. S. 229, 241-242 (1976). No purpose to discriminate against men or women as a class can be gleaned from the Texas law, so rational-basis review applies. That review is readily satisfied here by the same rational basis that satisfied it in Bowers —society's belief that certain forms of sexual behavior are "immoral and unacceptable," 478 U. S., at 196. This is the same justification that supports many other laws regulating sexual behavior that make a distinction based upon the identity of the partner—for example, laws against adultery, fornication, and adult incest, and laws refusing to recognize homosexual marriage.

        JUSTICE O'CONNOR argues that the discrimination in this law which must be justified is not its discrimination with regard to the sex of the partner but its discrimination with regard to the sexual proclivity of the principal actor.

        "While it is true that the law applies only to conduct, the conduct targeted by this law is conduct that is closely correlated with being homosexual. Under such circumstances, Texas' sodomy law is targeted at more than conduct.

[539 U.S. 601]

It is instead directed toward gay persons as a class." Ante, at 583.

        Of course the same could be said of any law. A law against public nudity targets "the conduct that is closely correlated with being a nudist," and hence "is targeted at more than conduct"; it is "directed toward nudists as a class." But be that as it may. Even if the Texas law does deny equal protection to "homosexuals as a class," that denial still does not need to be justified by anything more than a rational basis, which our cases show is satisfied by the enforcement of traditional notions of sexual morality.

        JUSTICE O'CONNOR simply decrees application of "a more searching form of rational basis review" to the Texas statute. Ante, at 580. The cases she cites do not recognize such a standard, and reach their conclusions only after finding, as required by conventional rational-basis analysis, that no conceivable legitimate state interest supports the classification at issue. See Romer v. Evans, 517 U. S., at 635; Cleburne v. Cleburne Living Center, Inc., 473 U. S. 432, 448-450 (1985); Department of Agriculture v. Moreno, 413 U. S. 528, 534-538 (1973). Nor does JUSTICE O'CONNOR explain precisely what her "more searching form" of rational-basis review consists of. It must at least mean, however, that laws exhibiting "a desire to harm a politically unpopular group," ante, at 580, are invalid even though there may be a conceivable rational basis to support them.

        This reasoning leaves on pretty shaky grounds state laws limiting marriage to opposite-sex couples. JUSTICE O'CONNOR seeks to preserve them by the conclusory statement that "preserving the traditional institution of marriage" is a legitimate state interest. Ante, at 585. But "preserving the traditional institution of marriage" is just a kinder way of describing the State's moral disapproval of same-sex couples. Texas's interest in § 21.06 could be recast in similarly euphemistic terms: "preserving the traditional sexual mores of our society." In the jurisprudence JUSTICE O'CONNOR

[539 U.S. 602]

has seemingly created, judges can validate laws by characterizing them as "preserving the traditions of society" (good); or invalidate them by characterizing them as "expressing moral disapproval" (bad).

        Today's opinion is the product of a Court, which is the product of a law-profession culture, that has largely signed on to the so-called homosexual agenda, by which I mean the agenda promoted by some homosexual activists directed at eliminating the moral opprobrium that has traditionally attached to homosexual conduct. I noted in an earlier opinion the fact that the American Association of Law Schools (to which any reputable law school must seek to belong) excludes from membership any school that refuses to ban from its job-interview facilities a law firm (no matter how small) that does not wish to hire as a prospective partner a person who openly engages in homosexual conduct. See Romer, supra, at 653.

        One of the most revealing statements in today's opinion is the Court's grim warning that the criminalization of homosexual conduct is "an invitation to subject homosexual persons to discrimination both in the public and in the private spheres." Ante, at 575. It is clear from this that the Court has taken sides in the culture war, departing from its role of assuring, as neutral observer, that the democratic rules of engagement are observed. Many Americans do not want persons who openly engage in homosexual conduct as partners in their business, as scoutmasters for their children, as teachers in their children's schools, or as boarders in their home. They view this as protecting themselves and their families from a lifestyle that they believe to be immoral and destructive. The Court views it as "discrimination" which it is the function of our judgments to deter. So imbued is the Court with the law profession's anti-anti-homosexual culture, that it is seemingly unaware that the attitudes of that

[539 U.S. 603]

culture are not obviously "mainstream"; that in most States what the Court calls "discrimination" against those who engage in homosexual acts is perfectly legal; that proposals to ban such "discrimination" under Title VII have repeatedly been rejected by Congress, see Employment Non-Discrimination Act of 1994, S. 2238, 103d Cong., 2d Sess. (1994); Civil Rights Amendments, H. R. 5452, 94th Cong., 1st Sess. (1975); that in some cases such "discrimination" is mandated by federal statute, see 10 U. S. C. § 654(b)(1) (mandating discharge from the Armed Forces of any service member who engages in or intends to engage in homosexual acts); and that in some cases such "discrimination" is a constitutional right, see Boy Scouts of America v. Dale, 530 U. S. 640 (2000).

        Let me be clear that I have nothing against homosexuals, or any other group, promoting their agenda through normal democratic means. Social perceptions of sexual and other morality change over time, and every group has the right to persuade its fellow citizens that its view of such matters is the best. That homosexuals have achieved some success in that enterprise is attested to by the fact that Texas is one of the few remaining States that criminalize private, consensual homosexual acts. But persuading one's fellow citizens is one thing, and imposing one's views in absence of democratic majority will is something else. I would no more require a State to criminalize homosexual acts—or, for that matter, display any moral disapprobation of them—than I would forbid it to do so. What Texas has chosen to do is well within the range of traditional democratic action, and its hand should not be stayed through the invention of a brand-new "constitutional right" by a Court that is impatient of democratic change. It is indeed true that "later generations can see that laws once thought necessary and proper in fact serve only to oppress," ante, at 579; and when that happens, later generations can repeal those laws. But it is the premise of our system that those judgments are to be made

[539 U.S. 604]

by the people, and not imposed by a governing caste that knows best.

        One of the benefits of leaving regulation of this matter to the people rather than to the courts is that the people, unlike judges, need not carry things to their logical conclusion. The people may feel that their disapprobation of homosexual conduct is strong enough to disallow homosexual marriage, but not strong enough to criminalize private homosexual acts—and may legislate accordingly. The Court today pretends that it possesses a similar freedom of action, so that we need not fear judicial imposition of homosexual marriage, as has recently occurred in Canada (in a decision that the Canadian Government has chosen not to appeal). See Halpern v. Toronto, 2003 WL 34950 (Ontario Ct. App.); Cohen, Dozens in Canada Follow Gay Couple's Lead, Washington Post, June 12, 2003, p. A25. At the end of its opinion—after having laid waste the foundations of our rational-basis jurisprudence —the Court says that the present case "does not involve whether the government must give formal recognition to any relationship that homosexual persons seek to enter." Ante, at 578. Do not believe it. More illuminating than this bald, unreasoned disclaimer is the progression of thought displayed by an earlier passage in the Court's opinion, which notes the constitutional protections afforded to "personal decisions relating to marriage, procreation, contraception, family relationships, child rearing, and education," and then declares that "[p]ersons in a homosexual relationship may seek autonomy for these purposes, just as heterosexual persons do." Ante, at 574 (emphasis added). Today's opinion dismantles the structure of constitutional law that has permitted a distinction to be made between heterosexual and homosexual unions, insofar as formal recognition in marriage is concerned. If moral disapprobation of homosexual conduct is "no legitimate state interest" for purposes of proscribing that conduct, ante, at 578; and if, as the Court coos (casting aside all pretense of neutrality), "[w]hen

[539 U.S. 605]

sexuality finds overt expression in intimate conduct with another person, the conduct can be but one element in a personal bond that is more enduring," ante, at 567; what justification could there possibly be for denying the benefits of marriage to homosexual couples exercising "[t]he liberty protected by the Constitution," ibid.? Surely not the encouragement of procreation, since the sterile and the elderly are allowed to marry. This case "does not involve" the issue of homosexual marriage only if one entertains the belief that principle and logic have nothing to do with the decisions of this Court. Many will hope that, as the Court comfortingly assures us, this is so.

        The matters appropriate for this Court's resolution are only three: Texas's prohibition of sodomy neither infringes a "fundamental right" (which the Court does not dispute), nor is unsupported by a rational relation to what the Constitution considers a legitimate state interest, nor denies the equal protection of the laws. I dissent.

---------------

Notes:

1. This last-cited critic of Bowers actually writes: "[Bowers] is correct nevertheless that the right to engage in homosexual acts is not deeply rooted in America's history and tradition." Posner, Sex and Reason, at 343.

2. While the Court does not overrule Bowers' holding that homosexual sodomy is not a "fundamental right," it is worth noting that the "societal reliance" upon that aspect of the decision has been substantial as well. See 10 U. S. C. § 654(b)(1) ("A member of the armed forces shall be separated from the armed forces . . . if . . . the member has engaged in . . . a homosexual act or acts"); Marcum v. McWhorter, 308 F. 3d 635, 640-642 (CA6 2002) (relying on Bowers in rejecting a claimed fundamental right to commit adultery); Mullins v. Oregon, 57 F. 3d 789, 793-794 (CA9 1995) (relying on Bowers in rejecting a grandparent's claimed "fundamental liberty interes[t]" in the adoption of her grandchildren); Doe v. Wigginton, 21 F. 3d 733, 739-740 (CA6 1994) (relying on Bowers in rejecting a prisoner's claimed "fundamental right" to on-demand HIV testing); Schowengerdt v. United States, 944 F. 2d 483, 490 (CA9 1991) (relying on Bowers in upholding a bisexual's discharge from the armed services); Charles v. Baesler, 910 F. 2d 1349, 1353 (CA6 1990) (relying on Bowers in rejecting fire department captain's claimed "fundamental" interest in a promotion); Henne v. Wright, 904 F. 2d 1208, 1214-1215 (CA8 1990) (relying on Bowers in rejecting a claim that state law restricting surnames that could be given to children at birth implicates a "fundamental right"); Walls v. Petersburg, 895 F. 2d 188, 193 (CA4 1990) (relying on Bowers in rejecting substantive-due-process challenge to a police department questionnaire that asked prospective employees about homosexual activity); High Tech Gays v. Defense Industrial Security Clearance Office, 895 F. 2d 563, 570-571 (CA9 1988) (relying on Bowers' holding that homosexual activity is not a fundamental right in rejecting—on the basis of the rational-basis standard—an equalprotection challenge to the Defense Department's policy of conducting expanded investigations into backgrounds of gay and lesbian applicants for secret and top-secret security clearances).

3. The Court is quite right that "`[h]istory and tradition are the starting point but not in all cases the ending point of the substantive due process inquiry,'" ante, at 572. An asserted "fundamental liberty interest" must not only be "`deeply rooted in this Nation's history and tradition,'" Washington v. Glucksberg, 521 U. S. 702, 721 (1997), but it must also be "`implicit in the concept of ordered liberty,'" so that "`neither liberty nor justice would exist if [it] were sacrificed,'" ibid.Moreover, liberty interests unsupported by history and tradition, though not deserving of "heightened scrutiny," are still protected from state laws that are not rationally related to any legitimate state interest. Id., at 722. As I proceed to discuss, it is this latter principle that the Court applies in the present case.

---------------

        JUSTICE THOMAS, dissenting.

        I join JUSTICE SCALIA'S dissenting opinion. I write separately to note that the law before the Court today "is . . . uncommonly silly." Griswold v. Connecticut, 381 U. S. 479, 527 (1965) (Stewart, J., dissenting). If I were a member of the Texas Legislature, I would vote to repeal it. Punishing someone for expressing his sexual preference through noncommercial consensual conduct with another adult does not appear to be a worthy way to expend valuable law enforcement resources.

        Notwithstanding this, I recognize that as a Member of this Court I am not empowered to help petitioners and others similarly situated. My duty, rather, is to "decide cases `agreeably to the Constitution and laws of the United States.'" Id., at 530. And, just like Justice Stewart, I "can find [neither in the Bill of Rights nor any other part of the

[539 U.S. 606]

Constitution a] general right of privacy," ibid., or as the Court terms it today, the "liberty of the person both in its spatial and more transcendent dimensions," ante, at 562.

Page 526

          Messrs. Felix Frankfurter, of Cambridge, Mass., and F. H. Stephens, of Washington, D. C., for appellants.

  [Argument of Counsel from pages 526-535 intentionally omitted]

Page 535

          Messrs. Wade H. Ellis and Challen B. Ellis, both of Washington, D. C., for appellees.

  [Argument of Counsel from pages 535-539 intentionally omitted]

Page 539

          Mr. Wm. L. Brewster, of Portland, Or., amicus curiae.

           Mr. Justice SUTHERLAND delivered the opinion of the Court.

          The question presented for determination by these appeals is the constitutionality of the Act of September 19, 1918, providing for the fixing of minimum wages for women and children in the District of Columbia. 40 Stat. 960, c. 174 (Comp. St. Ann. Supp. 1919, §§ 3421 1/2 a-3421 1/2 w).

          The act provides for a board of three members to be constituted, as far as practicable, so as to be equally representative

Page 540

of employers, employees and the public. The board is authorized to have public hearings, at which persons interested in the matter being investigated may appear and testify, to administer oaths, issue subpoenas requiring the attendance of witnesses and production of books, etc., and to make rules and regulations for carrying the act into effect.

          By section 8 the board is authorized——

          '(1) To investigate and ascertain the wages of women and minors in the different occupations in which they are employed in the District of Columbia; (2) to examine, through any member or authorized representative, any book, pay roll or other record of any employer of women or minors that in any way appertains to or has a bearing upon the question of wages of any such women or minors; and (3) to require from such employer full and true statements of the wages paid to all women and minors in his employment.'

          And by section 9——

          'To ascertain and declare, in the manner hereinafter provided, the following things: (a) Standards of minimum wages for women in any occupation within the District of Columbia, and what wages are inadequate to supply the necessary cost of living to any such women workers to maintain them in good health and to protect their morals; and (b) standards of minimum wages for minors in any occupation within the District of Columbia, and what wages are unreasonably low for any such minor workers.'

          The act then provides (section 10) that if the board, after investigation, is of opinion that any substantial number of women workers in any occupation are receiving wages inadequate to supply them with the necessary cost of living, maintain them in health and protect their morals, a conference may be called to consider and inquire into and report on the subject investigated, the conference to be equally representative of employers and employees in

Page 541

such occupation and of the public, and to include one or more members of the board.

          The conference is required to make and transmit to the board a report including, among other things:

          'Recommendations as to standards of minimum wages for women workers in the occupation under inquiry and as to what wages are inadequate to supply the necessary cost of living to women workers in such occupation and to maintain them in health and to protect their morals.'

          The board is authorized (section 12) to consider and review these recommendations and to approve or disapprove any or all of them. If it approve any recommendations it must give public notice of its intention and hold a public hearing at which the persons interested will be heard. After such hearing, the board is authorized to make such order as to it may appear necessary to carry into effect the recommendations, and to require all employers in the occupation affected to comply therewith. It is made unlawful for any such employer to violate in this regard any provision of the order or to employ any woman worker at lower wages than are thereby permitted.

          There is a provision (section 13) under which the board may issue a special license to a woman whose earning capacity 'has been impaired by age or otherwise,' authorizin her employment at less than the minimum wages fixed under the act.

          All questions of fact (section 17) are to be determined by the board, from whose decision there is no appeal; but an appeal is allowed on questions of law.

          Any violation of the act (section 18) by an employer or his agent or by corporate agents is declared to be a misdemeanor, punishable by fine and imprisonment.

          Finally, after some further provisions not necessary to be stated, it is declared (section 23) that the purposes of the act are——

          'To protect the women and minors of the District

Page 542

          from conditions detrimental to their health and morals, resulting from wages which are inadequate to maintain decent standards of living; and the act in each of its provisions and in its entirety shall be interpreted to effectuate these purposes.'

          The appellee in the first case is a corporation maintaining a hospital for children in the District. It employs a large number of women in various capacities, with whom it had agreed upon rates of wages and compensation satisfactory to such employees, but which in some instances were less than the minimum wage fixed by an order of the board made in pursuance of the act. The women with whom appellee had so contracted were all of full age and under no legal disability. The instant suit was brought by the appellee in the Supreme Court of the District to restrain the board from enforcing or attempting to enforce its order on the ground that the same was in contravention of the Constitution, and particularly the due process clause of the Fifth Amendment.

          In the second case the appellee, a woman 21 years of age, was employed by the Congress Hall Hotel Company as an elevator operator, at a salary of $35 per month and two meals a day. She alleges that the work was light and healthful, the hours short, with surroundings clean and moral, and that she was anxious to continue it for the compensation she was receiving, and that she did not earn more. Her services were satisfactory to the Hotel Company, and it would have been glad to retain her, but was obliged to dispense with her services by reason of the order of the board and on account of the penalties prescribed by the act. The wages received by this appellee were the best she was able to obtain for any work she was capable of performing, and the enforcement of the order, she alleges, deprived her of such employment and wages. She further averred that she could not secure any other position at which she could make a living, with

Page 543

as good physical and moral surroundings, and earn as good wages, and that she was desirous of continuing and would continue the employment, but for the order of the board. An injunction was prayed as in the other case.

          The Supreme Court of the District denied the injunction and dismissed the bill in each case. Upon appeal the Court of Appeals by a majority first affirmed, and subsequently, on a rehearing, reversed, the trial court. Upon the first argument a justice of the District Supreme Court was called in to take the place of one of the appellate court justices, who was ill. Application for rehearing was made, and, by the court as thus constituted, was denied. Subsequently, and during the term, a rehearing was granted by an order concurred in by two of the appellate court justices, one being the justice whose place on the prior occasion had been filled by the Supreme Court member. Upon the rehearing thus granted, the Court of Appeals, rejecting the first opinion, held the act in question to be unconstitutional and reversed the decrees of the trial court. 284 Fed. 613. Thereupon the cases were remanded, and the trial court entered decrees in pursuance of the mandate, declaring the act in question to be unconstitutional and granting permanent injunctions. Appeals to the Court of Appeals followed, and the decrees on the trial court were affirmed. It is from these final decrees that the cases come here.

          Upon this state of facts the jurisdiction of the lower court to rant a rehearing, after first denying it, is challenged. We do not deem it necessary to consider the matter farther than to say that we are here dealing with the second appeals, while the proceedings complained of occurred upon the first appeals. That the lower court could properly entertain the second appeals, and decide the cases does not admit of doubt; and this the appellants virtually concede by having themselves invoked the jurisdiction. See Rooker et al. v. Fidelity Trust Company et al., 261 U. S. 114, 43 Sup. Ct. 288, 67 L. Ed. ——, February 19, 1923.

Page 544

          We come then, at once, to the substantive question involved.

          The judicial duty of passing upon the constitutionality of an act of Congress is one of great gravity and delicacy. The statute here in question has successfully borne the scrutiny of the legislative branch of the government, which, by enacting it, has affirmed its validity, and that determination must be given great weight. This court, by an unbroken line of decisions from Chief Justice Marshall to the present day, has steadily adhered to the rule that every possible presumption is in favor of the validity of an act of Congress until overcome beyond rational doubt. But, if by clear and indubitable demonstration a statute be opposed to the Constitution, we have no choice but to say so. The Constitution, by its own terms, is the supreme law of the land, emanating from the people, the repository of ultimate sovereignty under our form of government. A congressional statute, on the other hand, is the act of an agency of this sovereign authority, and if it conflict with the Constitution must fall; for that which is not supreme must yield to that which is. To hold it invalid (if it be invalid) is a plain exercise of the judicial power—that power vested in courts to enable them to administer justice according to law. From the authority to ascertain and determine the law in a given case, there necessarily results, in case of conflict, the duty to declare and enforce the rule of the supreme law and reject that of an inferior act of legislation which, transcending the Constitution, is of no effect and binding on no one. This is not the exercise of a substantive power to review and nullify acts of Congress, for no such substantive power exists. It is simply a necessary concomitant of the power to hear and dispose of a case or controversy properly before the court, to the determination of which must be brought the test and measure of the law.

Page 545

          The statute now under consideration is attacked upon the ground that it authorizes an unconstitutional interference with the freedom of contract included within the guaranties of the due process clause of the Fifth Amendment. That the right to contract about one's affairs is a part of the liberty of the individual protected by this clause is settled by the decisions of this court and is no longer open to question. Allgeyer v. Louisiana, 165 U. S. 578, 591, 17 Sup. Ct. 427, 41 L. Ed. 832; New York Life Ins. Co. v. Dodge, 246 U. S. 357, 373-374, 38 Sup. Ct. 337, 62 L. Ed. 772, Ann. Cas. 1918E, 593; Coppage v. Kansas, 236 U. S. 1, 10, 14, 35 Sup. Ct. 240, 59 L. Ed. 441, L. R. A. 1915C, 960; Adair v. United States, 208 U. S. 161, 28 Sup. Ct. 277, 52 L. Ed. 436, 13 Ann. Cas. 764; Lochner v. New York, 198 U. S. 45, 25 Sup. Ct. 539, 49 L. Ed. 937, 3 Ann. Cas. 1133; Butchers' Union, etc., v. Crescent City, etc., 111 U. S. 746, 4 Sup. Ct. 652, 28 L. Ed. 585; Muller v. Oregon, 208 U. S. 412, 421, 28 Sup. Ct. 324, 52 L. Ed. 551, 13 Ann. Cas. 957. Within this liberty are contracts of employment of labor. In making such contracts, generally speaking, the parties have an equal right to obtain from each other the best terms they can as the result of private bargaining.

          In Adair v. United States, supra, Mr. Justice Harlan (208 U. S. at pages 174, 175, 28 Sup. Ct. 280, 52 L. Ed. 436, 13 Ann. Cas. 764), speaking for the court said:

          'The right of a person to sell his labor upon such terms as he deems proper is, i its essence, the same as the right of the purchaser of labor to prescribe the conditions upon which he will accept such labor from the person offering to sell. * * * In all such particulars the employer and the employe have equality of right, and any legislation that disturbs that equality is an arbitrary interference with the liberty of contract which no government can legally justify in a free land.'

          In Coppage v. Kansas, supra (236 U. S. at page 14, 35 Sup. Ct. 243, 59 L. Ed. 441, L. R. A. 1915C, 960), this court, speaking through Mr. Justice Pitney, said:

          'Included in the right of personal liberty and the right of private property—partaking of the nature of each—is the right to make contracts for the acquisition of property. Chief among such contracts is that of personal employment, by which labor and other services are exchanged for money or other forms of property. If this

Page 546

          right be struck down or arbitrarily interfered with, there is a substantial impairment of liberty in the long-established constitutional sense. The right is as essential to the laborer as to the capitalist, to the poor as to the rich; for the vast majority of persons have no other honest way to begin to acquire property, save by working for money.

          'An interference with this liberty so serious as that now under consideration, and so disturbing of equality of right must be deemed to be arbitrary, unless it be supportable as a reasonable exercise of the police power of the state.'

          There is, of course, no such thing as absolute freedom of contract. It is subject to a great variety of restraints. But freedom of contract is, nevertheless, the general rule and restraint the exception, and the exercise of legislative authority to abridge it can be justified only by the existence of exceptional circumstances. Whether these circumstances exist in the present case constitutes the question to be answered. It will be helpful to this end to review some of the decisions where the interference has been upheld and consider the grounds upon which they rest:

          (1) Those dealing with statutes fixing rates and charges to be exacted by businesses impressed with a public interest. There are many cases, but it is sufficient to cite Munn v. Illinois, 94 U. S. 113, 24 L. Ed. 77. The power here rests upon the ground that, where property is devoted to a public use, the owner thereby in effect grants to the public an interest in the use which may be controlled by the public for the common good to the extent of the interest thus created. It is upon this theory that these statutes have been upheld, and, it may be noted in passing, so upheld, even in respect of their incidental and injurious or destructive effect upon preexisting contracts. See Louisville & Nashville Railway Co. v. Mottley, 219 U. S. 467, 31 Sup. Ct. 265, 55 L. Ed. 297, 34 L. R. A. (N. S.) 671. In the case at bar the statute does not depend upon

Page 547

the existence of a public interest in any business to be affected, and this class of cases may be laid aside as inapplicable.

          (2) Statutes relating to contracts for the performance of public work. Atkin v. Kansas, 191 U. S. 207, 24 Sup. Ct. 124, 48 L. Ed. 148; Heim v. McCall, 239 U. S. 175, 36 Sup. Ct. 78, 60 L. Ed. 206, Ann. Cas. 1917B, 287; Ellis v. United States, 206 U. S. 246, 27 Sup. Ct. 600, 51 L. Ed. 1047, 11 Ann. Cas. 589. These cases sustain such statutes as depending, not upon the right to condition private contracts, but upon the right of the government to prescribe the conditions upon which it will permit work of a public character to be done for it, or, in the case of a state, for its municipalities. We may therefore, in like manner, dismiss these decisions from consideration as inapplicable.

          (3) Statutes prescribing the character, methods, and time for payment of wages. Under this head may be included McLean v. Arkansas, 211 U. S. 539, 29 Sup. Ct. 206, 53 L. Ed. 315, sustaining a state statute requiring coal to be measured for payment of miners' wages before screening; Knox ille Iron Co. v. Harbison, 183 U. S. 13, 22 Sup. Ct. 1, 46 L. Ed. 55, sustaining a Tennessee statute requiring the redemption in cash of store orders issued in payment of wages; Erie Railway Co. v. Williams, 233 U. S. 685, 34 Sup. Ct. 761, 58 L. Ed. 1155, 51 L. R. A. (N. S.) 1097, upholding a statute regulating the time within which wages shall be paid to employees in certain specified industries; and other cases sustaining statutes of like import and effect. In none of the statutes thus sustained was the liberty of employer or employee to fix the amount of wages the one was willing to pay and the other willing to receive interfered with. Their tendency and purpose was to prevent unfair, and perhaps fraudulent, methods in the payment of wages, and in no sense can they be said to be, or to furnish a precedent for, wagefixing statutes.

          (4) Statutes fixing hours of labor. It is upon this class that the greatest emphasis is laid in argument, and therefore, and because such cases approach most nearly the line of principle applicable to the statute here in volved, we shall consider them more at length In some instances

Page 548

the statute limited the hours of labor for men in certain occupations, and in others it was confined in its application to women. No statute has thus far been brought to the attention of this court which by its terms, applied to all occupations. In Holden v. Hardy, 169 U. S. 366, 18 Sup. Ct. 383, 42 L. Ed. 780, the court considered an act of the Utah Legislature, restricting the hours of labor in mines and smelters. This statute was sustained as a legitimate exercise of the police power, on the ground that the Legislature had determined that these particular employments, when too long pursued, were injurious to the health of the employees, and that, as there were reasonable grounds for supporting this determination on the part of the Legislature, its decision in that respect was beyond the reviewing power of the federal courts.

          That this constituted the basis of the decision is emphasized by the subsequent decision in Lochner v. New York, 198 U. S. 45, 25 Sup. Ct. 539, 49 L. Ed. 937, 3 Ann. Cas. 1133, reviewing a state statute which restricted the employment of all persons in bakeries to 10 hours in any one day. The court referred to Holden v. Hardy, supra, and, declaring it to be inapplicable, held the statute unconstitutional as an unreasonable, unnecessary and arbitrary interference with the liberty of contract and therefore void under the Constitution.

          Mr. Justice Peckham, speaking for the court (198 U. S. 56, 25 Sup. Ct. 542, 49 L. Ed. 937, 3 Ann. Cas. 1133), said:

          'It must, of course, be conceded that there is a limit to the valid exercise of the police power by the state. There is no dispute concerning this general proposition. Otherwise the Fourteenth Amendment would have no efficacy and the legislatures of the states would have unbounded power, and it would be enough to say that any piece of legislation was enacted to conserve the morals, the health or the safety of the people; such legislation would be valid, no matter how absolutely without foundation the claim might be. The claim of the police power

Page 549

          would be a mere pretext—become another and delusive name for the supreme sovereignty of the state to be exercised free from constitutional restraint.'

          And again (198 U. S. 57, 58, 25 Sup. Ct. 543, 49 L. Ed. 937, 3 Ann. Cas. 1133):

          'It is a question of which of two powers or rights shall prevail—the power of the state to legislate or the right of the individual to liberty of person and freedom of contract. The mere assertion that the subject relates, though but in a remote degree, to the public health does not necessarily render the enactment valid. The act must have a more direct relation, as a means to an end, and the end itself must be appropriate and legitimate, before an act can be held to be valid which interferes with the general right of an individual to be free in his person and in his power to contract in relation to his own labor.'

          Coming then directly to the statute (198 U. S. 58, 25 Sup. Ct. 543, 49 L. Ed. 937, 3 Ann. Cas. 1133), the court said:

          'We think the limit of the police power has been reached and passed in this case. There is, in our judgment, no reasonable foundation for holding this to be necessary or appropriate as a health law to safeguard the public health or the health of the individuals who are following the trade of a baker. If this statute be valid, and if, therefore, a proper case is made out in which to deny the right of an individual, sui juris, as employer or employe, to make contracts for the labor of the latter under the protection of the provisions of the federal Constitution, there would seem to be no length to which legislation of this nature might not go.'

          And, after pointing out the unreasonable range to which the principle of the statute might be extended the court said (198 U. S. 60, 25 Sup. Ct. 544, 49 L. Ed. 937, 3 Ann. Cas. 1133):

          'It is also urged, pursuing the same line of argument, that it is to the interest of the state that its population should be strong and robust, and therefore any legislation which may be said to tend to make people healthy must

Page 550

          be valid as health laws, enacted under the police power. If this be a valid argument and a justification for this kind of legislation, it follows that the protection of the federal Constitution from undue interference with liberty of person and freedom of contract is visionary, wherever the law is sought to be justified as a valid exercise of the police power. Scarcely any law but might find shelter under such assumptions, and conduct, properly so called, as well as contract, would come under the restrictive sway of the Legislature.'

          And further (198 U. S. 61, 25 Sup. Ct. 545, 49 L. Ed. 937, 3 Ann. Cas. 1133):

          'Statutes of the nature of that under review, limiting the hours in which grown and intelligent men may labor to earn their living, are mere meddlesome interferences with the rights of the individual and they are not saved from condemnation by the claim that they are passed in the exercise of the police power and upon the subject of the health of the individual whose rights are interfered with, unless there be some fair ground, reasonable in and of itself, to say that there is material danger to the public health or to the health of the employes, if the hours of labor are not curtailed.'

          Subsequent cases in this court have been distinguished from that decision, but the principles therein stated have never been disapproved.

          In Bunting v. Oregon, 243 U. S. 426, 37 Sup. Ct. 435, 61 L. Ed. 830, Ann. Cas. 1918A, 1043, a state statute forbidding the employment of any person in any mill, factory, or manufacturing establishment more than ten hours in any one day, and providing payment for overtime not exceeding three hours in any one day at the rate of time and a half of the regular wage, was sustained on the ground that, since the state Legislature and state Supreme Court had found such a law necessary for the preservation of the health of employees in these industries, this court would accept their judgment, in the absence of facts to support the contrary conclusion. The law was attacked

Page 551

on the ground that it constituted an attempt to fix wages, but that contention was rejected and the law sustained as a reasonable regulation of hours of service.

          Wilson v. New, 243 U. S. 332, 37 Sup. Ct. 298, 61 L. Ed. 755, L. R. A. 1917E, 938, Ann. Cas. 1918A, 1024, involved the validity of the so-called Adamson Law (Comp. St. §§ 8680a-8680d), which established an 8-hour day for employees of interstate carriers for which it fixed a scale of minimum wages with proportionate increases for overtime, to be enforced, however, only for a limited period. The act was sustained primarily upon the ground that it was a regulation of a business charged with a public interest. The court, speaking through the Chief Justice, pointed out that regarding 'the private right and private interest as contradistinguished from the public interest the power exists between the parties, the employers and employees, to agree as to a standard of wages free from legislative interference,' but that this did not affect the power to deal with the matter with a view to protect the public right, and then said (243 U. S. 353, 37 Sup. Ct. 304, 61 L. Ed. 755, L. R. A. 1917E, 938, Ann. Cas. 1918A, 1024):

          'And this emphasizes that there is no question here of purely private right since the law is concerned only with those who are engaged in a business charged with a public interest where the subject dealt with as to all the parties is one involved in that business and which we have seen comes under the control of the right to regulate to the extent that the power to do so is appropriate or relevant to the business regulated.'

          Moreover, in sustaining the wage feature, of the law, emphasis was put upon the fact (243 U. S. 345, 37 Sup. Ct. 301, 61 L. Ed. 755, L. R. A. 1917E, 938, Ann. Cas. 1918A, 1024) that it was in this respect temporary 'leaving the employers and employees free as to the subject of wages to govern their relations by their own agreements after the specified time.' The act was not only temporary in this respect, but it was passed to meet a sudden and great emergency. This feature of the law was sustained principally because the parties, for the time being, could not or would not agree. Here they are forbidden to agree.

Page 552

          The same principle was applied in the Rent Cases (Block v. Hirsh, 256 U. S. 135, 41 Sup. Ct. 458, 65 L. Ed. 865, 16 A. L. R. 165, and Marcus Brown Holding Co. v. Feldman, 256 U. S. 170, 41 Sup. Ct. 465, 65 L. Ed. 877), where this court sustained the legislative power to fix rents as between landlord and tenant upon the ground that the operation of the statutes was temporary to tide over an emergency and that the circumstances were such as to clothe 'the letting of buildings * * * with a public interest so great as to justify regulation by law.' The court said (256 U. S. 157, 41 Sup. Ct. 460, 65 L. Ed. 865, 16 A. L. R. 165):

          'The regulation is put and justified only as a temporary measure [citing Wilson v. New, supra]. A limit in time, to tide over a passing trouble, well may justify a law that could not be upheld as a permanent change.'

          In a subsequent case, Pennsylvania Coal Co. v. Mahon, 260 U. S. 393, 43 Sup. Ct. 158, 67 L. Ed. ——, decided December 11, 1922, this court, after saying, 'We are in danger of forgetting that a strong public desire to improve the public condition is not enough to warrant achieving the desire by a shorter cut than the constitutional way of paying for the change,' pointed out that the Rent Cases dealt with laws intended to meet a temporary emergency and 'went to the verge of the law.'

          In addition to the cases cited above, there are decisions of this court dealing with laws especially relating to hours of labor for women. Muller v. Oregon, 208 U. S. 412, 28 Sup. Ct. 324, 52 L. Ed. 551, 13 Ann. Cas. 957; Riley v. Massachusetts, 232 U. S. 671, 34 Sup. Ct. 469, 58 L. Ed. 788; Miller v. Wilson, 236 U. S. 373, 35 Sup. Ct. 342, 59 L. Ed. 628, L. R. A. 1915F, 829; Bosley v. McLaughlin, 236 U. S. 385, 35 Sup. Ct. 345, 59 L. Ed. 632.

          In the Muller Case the validity of an Oregon statute, forbidding the employment of any female in certain industries more than 10 hours during any one day was upheld. The decision proceeded upon the theory that the difference between the sexes may justify a different rule respecting hours of labor in the case of women than in the case of men. It is pointed out that these consist in differences of physical structure, especially in respect

Page 553

of the maternal functions, and also in the fact that historically woman has always been dependent upon man, who has established his control by superior physical strength. The Cases of Riley, Miller, and Bosley follow in this respect the Muller Case. But the ancient inequality of the sexes, otherwise than physical, as suggested in the Muller Case (208 U. S. 421, 28 Sup. Ct. 327, 52 L. Ed. 551, 13 Ann. Cas. 957) has continued 'with diminishing intensity.' In view of the great—not to say revolutionary—changes which have taken place since that utterance, in the contractual, political, and civil status of women, culminating in the Nineteenth Amendment, it is not unreasonable to say that these differences have now come almost, if not quite, to the vanishing point. In this aspect of the matter, while the physical differences must be recognized in appropriate cases, and legislation fixing hours or conditions of work may properly take them into account, we cannot accept the doctrine that women of mature age, sui juris, require or may be subjected to restrictions upon their liberty of contract which could not lawfully be imposed in the case of men under similar circumstances. To do so would be to ignore all the implications to be drawn from the present day trend of legislation, as well as that of common thought and usage, by which woman is accorded emancipation from the old doctrine that she must be given special protection or be subjected to special restraint in her contractual and civil relationships. In passing, it may be noted that the instant statute applies in the case of a woman employer contracting with a woman employee as it does when the former is a man.

          The essential characteristics of the statute now under consideration, which differentiate it from the laws fixing hours of labor, will be made to appear as we proceed. It is sufficient now to point out that the latter, as well as the statutes mentioned under paragraph (3), deal with incidents of the employment having no necessary effect upon

Page 554

the heart of the contract; that is, the amount of wages to be paid and received. A law forbidding work to continue beyond a given number of hours leaves the parties free to contract about wages and thereby equalize whatever additional burdens may be imposed upon the employer as a result of the restrictions as to hours, by an adjustment in respect of the amount of wages. Enough has been said to show that the authority to fix hours of labor cannot be exercised except in respect of those occupations where work of long continued duration is detrimental to health. This court has been careful in every case where the question has been raised, to place its decision upon this limited authority of the Legislature to regulate hours of labor and to disclaim any purpose to uphold the legislation as fixing wages, thus recognizing an essential difference between the two. It seems plain that these decisions afford no real support for any form of law establishing minimum wages.

          If now, in the light furnished by the foregoing exceptions to the general rule forbidding legislative interference with freedom of contract, we examine and analyze the statute in question, we shall see that it differs from them in every material respect. It is not a law dealing with any business charged with a public interest or with public work, or to meet and tide over a temporary emergency. It has nothing to do with the character, methods or periods of wage payments. It does not prescribe hours of labor or conditions under which labor is to be done. It is not for the protection of persons under legal disability or for the prevention of fraud. It is simply and exclusively a price-fixing law, confined to adult women (for we are not now considering the provisions relating to minors), who are legally as capable of contracting for themselves as men. It forbids two parties having lawful capacity—under penalties as to the employer—to freely contract with one another in respect of the price for

Page 555

which one shall render service to the other in a purely private employment where both are willing, perhaps anxious, to agree, even though the consequence may be to oblige one to surrender a desirable engagement and the other to dispense with the services of a desirable employee.¹ The price fixed by the board need have no relation to he capacity or earning power of the employee, the number of hours which may happen to constitute the day's work, the character of the place where the work is to be done, or the circumstances or surroundings of the employment, and, while it has no other basis to support its validity than the assumed necessities of the employee, it takes no account of any independent resources she may have. It is based wholly on the opinions of the members of the board and their advisers—perhaps an average of their opinions, if they do not precisely agree—as to what will be necessary to provide a living for a woman, keep her in health and preserve her morals. It applies to any and every occupation in the District, without regard to its nature or the character of the work.

          The standard furnished by the statute for the guidance of the board is so vague as to be impossible of practical application with any reasonable degree of accuracy. What is sufficient to supply the necessary cost of living for a woman worker and maintain her in good health and protect her morals is obviously not a precise or unvarying sum—not even approximately so. The amount will depend upon a variety of circumstances: The individual temperament, habits of thrift, care, ability to buy necessaries intelligently, and whether the woman live alone or with her family. To those who practice economy, a given sum will afford comfort, while to those or contrary habit the same sum will be wholly inadequate. The co-operative economies of the family group are not taken into account,

Page 556

though they constitute an important consideration in estimating the cost of living, for it is obvious that the individual expense will be less in the case of a member of a family than in the case of one living alone. The relation between earnings and morals is not capable of standardization. It cannot be shown that well-paid women safeguard their morals more carefully than those who are poorly paid. Morality rests upon other considerations than wages, and there is, certainly, no such prevalent connection between the two as to justify a broad attempt to adjust the latter with reference to the former. As a means of safeguarding morals the attempted classification, in our opinion, is without reasonable basis. No distinction can be made between women who work for others and those who do not; nor is there ground for distinction between women and men, for, certainly, if women require a minimum wage to preserve their morals men require it to preserve their honesty. For these reasons, and others which might be stated, the inquiry in respect of the necessary cost of living and of the income necessary to preserve health and morals presents an individual and not a composite question, and must be answered for each individual considered by herself and not by a general formula prescribed by a statutory bureau.

          This uncertainty of the statutory standard is demonstrated by a consideration of certain orders of the board already made. These orders fix the sum to be paid to a woman employed in a place where food is served or in a mercantile establishment, at $16.50 per week; in a printing establishment, at $15.50 per week; and in a laundry, at $15 per week, with a provision reducing this to $9 in the case of a beginner. If a woman employed to serve food requires a minimum of $16.50 per week, it is hard to understand how the same woman working in a printing establishment or in a laundry is to get on with an income lessened by from $1 to $7.50 per week. The board probably

Page 557

found it impossible to follow the indefinite standard of the statute, and brought other and different factors into the problem; and this goes far in the direction of demonstrating the fatal uncertainty of the act, an infirmity which, in our opinion, plainly exists.

          The law takes account of the necessities of only one party to the contract. It ignores the necessitie of the employer by compelling him to pay not less than a certain sum, not only whether the employee is capable of earning it, but irrespective of the ability of his business to sustain the burden, generously leaving him, of course, the privilege of abandoning his business as an alternative for going on at a loss. Within the limits of the minimum sum, he is precluded, under penalty of fine and imprisonment, from adjusting compensation to the differing merits of his employees. It compels him to pay at least the sum fixed in any event, because the employee needs it, but requires no service of equivalent value from the employee. It therefore undertakes to solve but one-half of the problem. The other half is the establishment of a corresponding standard of efficiency, and this forms no part of the policy of the legislation, although in practice the former half without the latter must lead to ultimate failure, in accordance with the inexorable law that no one can continue indefinitely to take out more than he puts in without ultimately exhausting the supply. The law is not confined to the great and powerful employers but embraces those whose bargaining power may be as weak as that of the employee. It takes no account of periods of stress and business depression, of crippling losses, which may leave the employer himself without adequate means of livelihood. To the extent that the sum fixed exceeds the fair value of the services rendered, it amounts to a compulsory exaction from the employer for the support of a partially indigent person, for whose condition there

Page 558

rests upon him no peculiar responsibility, and therefore, in effect, arbitrarily shifts to his shoulders a burden which, if it belongs to anybody, belongs to society as a whole.

          The feature of this statute, which perhaps more than any other, puts upon it the stamp of invalidity, is that it exacts from the employer an arbitrary payment for a purpose and upon a basis having no causal connection with his business, or the contract or the work the employee engages to do. The declared basis, as already pointed out, is not the value of the service rendered, but the extraneous circumstance that the employee needs to get a prescribed sum of money to insure her subsistence, health, and morals. The ethical right of every worker, man or woman, to a living wage may be conceded. One of the declared and important purposes of trade organizations is to secure it. And with that principle and with every legitimate effort to realize it in fact, no one can quarrel; but the fallacy of the proposed method of attaining it is that it assumes that every employer is bound at all events to furnish it. The moral requirement implicit in every contract of employment, viz. that the amount to be paid and the service to be rendered shall bear to each other some relation of just equivalence, is completely ignored. The necessities of the employee are alone considered, and these arise outside of the employment, are the same when there is no employment, and as great in one occupation as in another. Certainly the employer, by paying a fair equivalent for the service rendered, though not sufficient to support the employee, has neither caused nor contributed to her poverty. On the contrary, to the extent of what he pays, he has relieved it. In principle, there can be no difference between the case of selling labor and the case of selling goods. If one goes to the butcher, the baker, or grocer to buy food, he is morally entitled to obtain the worth of his money; but he is not entitled to more. If what he gets is worth what he pays, he is not justified in demanding

Page 559

more, simply because he needs more; and the shopkeeper, having dealt fairly and honestly in that transaction, is not concerned in any peculiar sense with the question of his customer's necessities. Should a statute undertake to vest in a commission power to determine the quantity of food necessary for individual support, and require the shopkeeper, if he sell to the individual at all, to furnish tha quantify at not more than a fixed maximum, it would undoubtedly fall before the constitutional test. The fallacy of any argument in support of the validity of such a statute would be quickly exposed. The argument in support of that now being considered is equally fallacious, though the weakness of it may not be so plain. A statute requiring an employer to pay in money, to pay at prescribed and regular intervals, to pay the value of the services rendered, even to pay with fair relation to the extent of the benefit obtained from the service, would be understandable. But a statute which prescribes payment without regard to any of these things, and solely with relation to circumstances apart from the contract of employment, the business affected by it, and the work done under it, is so clearly the product of a naked, arbitrary exercise of power that it cannot be allowed to stand under the Constitution of the United States.

          We are asked, upon the one land, to consider the fact that several states have adopted similar statutes, and we are invited, upon the other hand, to give weight to the fact that three times as many states, presumably as well informed and as anxious to promote the health and morals of their people, have refrained from enacting such legislation, We have also been furnished with a large number of printed opinions approving the policy of the minimum wage, and our own reading has disclosed a large number to the contrary. These are all proper enough for the consideration of the lawmaking bodies, since their tendency is to establish the desirability or undesirability of the

Page 560

legislation; but they reflect no legitimate light upon the question of its validity, and that is what we are called upon to decide. The elucidation of that question cannot be aided by counting heads.

          It is said that great benefits have resulted from the operation of such statutes, not alone in the District of Columbia but in the several states where they have been in force. A mass of reports, opinions of special observers and students of the subject, and the like, has been brought before us in support of this statement, all of which we have found interesting, but only mildly persuasive. That the earnings of women are now greater than they were formerly, and that conditions affecting women have become better in other respects, may be conceded; but convincing indications of the logical relation of these desirable changes to the law in question are significantly lacking. They may be, and quite probably are, due to other causes. We cannot close our eyes to the notorious fact that earnings everywhere in all occupations have greatly increased—not alone in states where the minimum wage law obtains but in the country generally—quite as much or more among men as among women, and in occupations outside the reach of the law as in those governed by it. No real test of the economic value of the law can be had during periods of maximum employment, when general causes keep wages up to or above the minimum; that will come in periods of depression and struggle for employment, when the efficient will be employed at the minimum rate, while the less capable may not be employed at all.

          Finally, it may be said that if, in the interest of the public welfare, the police power may be invoked to justify the fixing of a minimum wage, it may, when the public welfare is thought to require it, be invoked to justify a maximum wage. The power to fix high wages connotes, by like course of reasoning, the power to fix low wages. If, in the face of the guaranties of the Fifth

Page 561

Amendment, this form of legislation shall be legally justified, the field for the operation of the police power will have been widened to a great and dangerous degree. If, for example, in the opinion of future lawmakers, wages in the building trades shall become so high as to preclude people of ordinary means from building and owning homes, an authority which sustains the minimum wage will be invoked to support a maximum wage or building laborers and artisans, and the same argument which has been here urged to strip the employer of his constitutional liberty of contract in one direction will be utilized to strip the employee of his constitutional liberty of contract in the opposite direction. A wrong decision does not end with itself; it is a precedent, and, with the swing of sentiment, its bad influence may run from one extremity of the arc to the other.

          It has been said that legislation of the of contract in the opposite direction. A of social justice, for whose ends freedom of contract may lawfully be subjected to restraint. The liberty of the individual to do as he pleases, even in innocent matters, is not absolute. It must frequently yield to the common good, and the line beyond which the power of interference may not be pressed is neither definite nor unalterable, but may be made to move, within limits not well defined, with changing need and circumstance. Any attempt to fix a rigid boundary would be unwise as well as futile. But, nevertheless, there are limits to the power, and, when these have been passed, it becomes the plain duty of the courts in the proper exercise of their authority to so declare. To sustain the individual freedom of action contemplated by the Constitution is not to strike down the common good, but to exalt it; for surely the good of society as a whole cannot be better served than by the preservation against arbitrary restraint of the liberties of its constituent members.

Page 562

          It follows, from what has been said, that the act in question passes the limit prescribed by the Constitution, and accordingly the decrees of the court below are

          Affirmed.

          Mr. Justice BRANDEIS took no part in the consideration or decision of these cases.

           Mr. CHIEF JUSTICE TAFT, dissenting.

          I regret much to differ from the court in these cases.

          The boundary of the police power beyond which its exercise becomes an invasion of the guaranty of liberty under the Fifth and Fourteenth Amendments to the Constitutions is not easy to mark. Our court has been laboriously engaged in pricking out a line in successive cases. We must be careful, it seems to me, to follow that line as well as we can, and not to depart from it by suggesting a distinction that is formal rather than real.

          Legislatures in limiting freedom of contract between employee and employer by a minimum wage proceed on the assumption that employees, in the class receiving least pay, are not upon a full level of equality of choice with their employer and in their necessitous circumstances are prone to accept pretty much anything that is offered. They are peculiarly subject to the overreaching of the harsh and greedy employer. The evils of the sweating system and of the long hours and low wages which are characteristic of it are well known. Now, I agree that it is a disputable question in the field of political economy how far a statutory requirement of maximum hours or minimum wages may be a useful remedy for these evils, and whether it may not make the case of the oppressed employee worse than it was before. But it is not the function of this court to hold congressional acts invalid simply because they are passed to carry out economic views which the court believes to be unwise or unsound.

Page 563

          Legislatures which adopt a requirement of maximum hours or minimum wages may be presumed to believe that when sweating employers are prevented from paying unduly low wages by positive law they will continue their business, abating that part of their profits, which were wrung from the necessities of their employees, and will concede the better terms required by the law, and that while in individual cases, hardship may result, the restriction will inure to the benefit of the general class of employees in whose interest the law is passed, and so to that of the community at large.

          The right of the Legislature under the Fifth and Fourteenth Amendments to limit the hours of employment on the score of the health of he employee, it seems to me, has been firmly established. As to that, one would think, the line had been pricked out so that it has become a well formulated rule. In Holden v. Hardy, 169 U. S. 366, 18 Sup. Ct. 383, 42 L. Ed. 780, it was applied to miners and rested on the unfavorable environment of employment in mining and smelting. In Lochner v. New York, 198 U. S. 45, 25 Sup. Ct. 539, 49 L. Ed. 937, 3 Ann. Cas. 1133, it was held that restricting those employed in bakeries to 10 hours a day was an arbitrary and invalid interference with the liberty of contract secured by the Fourteenth Amendment. Then followed a number of cases beginning with Muller v. Oregon, 208 U. S. 412, 28 Sup. Ct. 324, 52 L. Ed. 551, 13 Ann. Cas. 957, sustaining the validity of a limit on maximum hours of labor for women to which I shall hereafter allude, and following these cases came Bunting v. Oregon, 243 U. S. 426, 37 Sup. Ct. 435, 61 L. Ed. 830, Ann. Cas. 1918A, 1043. In that case, this court sustained a law limiting the hours of labor of any person, whether man or woman, working in any mill, factory, or manufacturing establishment to 10 hours a day with a proviso as to further hours to which I shall hereafter advert. The law covered the whole field of industrial employment and certainly covered the case of persons employed in bakeries. Yet the opinion in the Bunting Case does not mention the Lochner Case. No one can

Page 564

suggest any constitutional distinction between employment in a bakery and one in any other kind of a manufacturing establishment which should make a limit of hours in the one invalid, and the same limit in the other permissible. It is impossible for me to reconcile the Bunting Case and the Lochner Case, and I have always supposed that the Lochner Case was thus overruled sub silentio. Yet the opinion of the court herein in support of its conclusion quotes from the opinion in the Lochner Case as one which has been sometimes distinguished but never overruled. Certainly there was no attempt to distinguish it in the Bunting Case.

          However, the opinion herein does not overrule the Bunting Case in express terms, and therefore. I assume that the conclusion in this case rests on the distinction between a minimum of wages and a maximum of hours in the limiting of liberty to contract. I regret to be at variance with the court as to the substance of this distinction. In absolute freedom of contract the one term is as important as the other, for both enter equally into the consideration given and received, a restriction as to one is not any greater in essence than the other, and is of the same kind. One is the multiplier and the other the multiplicand.

          If it be said that long hours of labor have a more direct effect upon the health of the employee than the low wage, there is very respectable authority from close observers, disclosed in the record and in the literature on the subject quoted at length in the briefs that they are equally harmful in this regard. Congress took this view and we cannot say it was not warranted in so doing.

          With deference to the very able opinion of the court and my brethren who concur in it, it appears to me to exaggerate the importance of the wage term of the contract of employment as more inviolate than its other terms. Its conclusion seems influenced by the fear that the

Page 565

concession of the power to impose a minimum wage must carry with it a concession of the power to fix a maximum wage. This, I submit, is a non sequitur. A line of distinction like the one under discussion in this case is, as the opinion elsewhere admits, a matter of degree and practical experience and not of pure logic. Certainly the wide difference between prescribing a minimum wage and a maximum wage could as a matter of degree and experience be easily affirmed.

          Moreover, there are decisions by this court which have sustained legislative limitations in respect to the wage term in contracts of employment. In McLean v. Arkansas, 211 U. S. 539, 29 Sup. Ct. 206, 53 L. d. 315, it was held within legislative power to make it unlawful to estimate the graduated pay of miners by weight after screening the coal. In Knoxville Iron Co. v. Harbison, 183 U. S. 13, 22 Sup. Ct. 1, 46 L. Ed. 55, it was held that stores orders issued for wages must be redeemable in cash. In Patterson v. Bark Eudora, 190 U. S. 169, 23 Sup. Ct. 821, 47 L. Ed. 1002, a law forbidding the payment of wages in advance was held valid. A like case is Strathearn S. S. Co. v. Dillon, 252 U. S. 348, 40 Sup. Ct. 350, 64 L. Ed. 607. While these did not impose a minimum on wages, they did take away from the employee the freedom to agree as to how they should be fixed, in what medium they should be paid, and when they should be paid, all features that might affect the amount or the mode of enjoyment of them. The first two really rested on the advantage the employer had in dealing with the employee. The third was deemed a proper curtailment of a sailor's right of contract in his own interest because of his proneness to squander his wages in port before sailing. In Bunting v. Oregon, supra, employees in a mill, factory, or manufacturing establishment were required if they worked over 10 hours a day to accept for the 3 additional hours permitted not less than 50 per cent, more than their usual wage. This was sustained as a mild penalty imposed on the employer to enforce the limitation as to hours; but it necessarily

Page 566

curtailed the employee's freedom to contract to work for the wages he saw fit to accept during those 3 hours. I do not feel, therefore, that either on the basis of reason, experience, or authority, the boundary of the police power should be drawn to include maximum hours and exclude a minimum wage.

          Without, however, expressing an opinion that a minimum wage limitation can be enacted for adult men, it is enough to say that the case before us involves only the application of the minimum wage to women. If I am right in thinking that the Legislature can find as much support in experience for the view that a sweating wage has as great and as direct a tendency to bring about an injury to the health and morals of workers, as for the view that long hours injure their health, then I respectfully submit that Muller v. Oregon, 208 U. S. 412, 28 Sup. Ct. 324, 52 L. Ed. 551, 13 Ann. Cas. 957, controls this case. The law which was there sustained forbade the employment of any female in any mechanical establishment or factory or laundry for more than 10 hours. This covered a pretty wide field in women's work, and it would not seem that any sound distinction between that case, and this can be built up on the fact that the law before us applies to all occupations of women with power in the board to make certain exceptions. Mr. Justice Brewer, who spoke for the court in Muller v. Oregon, based its conclusion on the natural limit to women's physical strength and the likelihood that long hours would therefore injure her health, and we have had since a series of cases which may be said to have established a rule of decision. Riley v. Massachusetts, 232 U. S. 671, 34 Sup. Ct. 469, 58 L. Ed. 788; Miller v. Wilson, 236 U. S. 373, 35 Sup. Ct. 342, 59 L. Ed. 628, L. R. A. 1915F, 829; Bosley v. McLaughlin, 236 U. S. 385, 35 Sup. Ct. 345, 59 L. Ed. 632. The cases covered restrictions in wide and varying fields of employment and in the later cases it will be found that the objection to the particular law was based, not on the ground that it had general application, but because it left out some employments.

Page 567

          I am not sure from a reading of the opinion whether the court thinks the authority of Muller v. Oregon is shaken by the adoption of the Nineteenth Amendment. The Nineteenth Amendment did not change the physical strength or limitations of women upon which the decision in Muller v. Oregon rests. The amendment did give women political power and makes more certain that legislative provisions for their protection will be in accord with their interests as they see them. But I do not think we are warra ted in varying constitutional construction based on physical differences between men and women, because of the amendment.

          But for my inability to agree with some general observations in the forcible opinion of Mr. Justice HOLMES, who follows me, I should be silent and merely record my concurrence in what he says. It is perhaps wiser for me, however, in a case of this importance separately to give my reasons for dissenting.

          I am authorized to say that Mr. Justice SANFORD concurs in this opinion.

           Mr. Justice HOLMES, dissenting.

          The question in this case is the broad one, Whether Congress can establish minimum rates of wages for women in the District of Columbia with due provision for special circumstances, or whether we must say that Congress had no power to meddle with the matter at all. To me, notwithstanding the deference due to the prevailing judgment of the Court, the power of Congress seems absolutely free from doubt. The end, to remove conditions leading to ill health, immorality and the deterioration of the race, no one would deny to be within the scope of constitutional legislation. The means are means that have the approval of Congress, of many States, and of those governments from which we have learned our greatest

Page 568

lessons. When so many intelligent persons, who have studied the matter more than any of us can, have thought that the means are effective and are worth the price it seems to me impossible to deny that the belief reasonably may be held by reasonable men. If the law encountered no other objection than that the means bore no relation to the end or that they cost too much I do not suppose that anyone would venture to say that it was bad. I agree, of course, that a law answering the foregoing requirements might be invalidated by specific provisions of the Constitution. For instance it might take private property without just compensation. But in the present instance the only objection that can be urged is found within the vague contours of the Fifth Amendment, prohibiting the depriving any person of liberty or property without due process of law. To that I turn.

          The earlier decisions upon the same words in the Fourteenth Amendment began within our memory and went no farther than an unpretentious assertion of the liberty to follow the ordinary callings. Later that innocuous generality was expanded into the dogma, Liberty of Contract. Contract is not specially mentioned in the text that we have to construe. It is merely an example of doing what you want to do, embodied in the word liberty. But pretty much all law consists in forbidding men to do some things that they want to do, and contract is no more exempt from law than other acts. Without enumerating all the restrictive laws that have been upheld I will mention a few that seem to me to have interfered with liberty of contract quite as seriously and directly as the one before us. Usury laws prohibit contracts by which a man receives more than so much interest for the money that he lends. Statutes of frauds restrict many contracts to certain forms. Some Sunday laws prohibit practically all contracts during one-seventh of our whole life. Insurance rates may be regulated. German Alliance Ins. Co.

Page 569

v. Lewis, 233 U. S. 389, 34 Sup. Ct. 612, 58 L. Ed. 1011, L. R. A. 1915C, 1189. (I concurred in that decision without regard to the public interest with which insurance was said to be clothed. It seemed to me that the principle was general.) Contracts may be forced upon the companies. National Union Fire Ins. Co. v. Wanberg, 260 U. S. 71, 43 Sup. Ct. 32, 67 L. Ed. ——, November 13, 1922. Employers of miners may be required to pay for coal by weight before screening. McLean v. Arkansas, 211 U. S. 539, 29 Sup. Ct. 206, 53 L. Ed. 315. Employers generally may be required to redeem in cash store orders accepted by their employees in payment. Knoxville Iron Co. v. Harbison, 183 U. S. 13, 22 Sup. Ct. 1, 46 L. Ed. 55. Payment of sailors in advance may be forbidden. Patterson v. The Eudora, 90 U. S. 169, 23 Sup. Ct. 821, 47 L. Ed. 1002. The size of a loaf of bread may be established. Schmidinger v. Chicago, 226 U. S. 578, 33 Sup. Ct. 182, 57 L. Ed. 364, Ann. Cas. 1914B, 284. The responsibility of employers to their employees may be profoundly modified. New York Central R. R. Co. v. White, 243 U. S. 188, 37 Sup. Ct. 247, 61 L. Ed. 667, L. R. A. 1917D, 1, Ann. Cas. 1917D, 629; Arizona Employers' Liability Cases, 250 U. S. 400, 39 Sup. Ct. 553, 63 L. Ed. 1058, 6 A. L. R. 1537. Finally women's hours of labor may be fixed, Muller v. Oregon, 208 U. S. 412, 28 Sup. Ct. 324, 52 L. Ed. 551, 13 Ann. Cas. 957; Riley v. Massachusetts, 232 U. S. 671, 679, 34 Sup. Ct. 469, 58 L. Ed. 788; Hawley v. Walker, 232 U. S. 718, 34 Sup. Ct. 479, 58 L. Ed. 813; Miller v. Wilson, 236 U. S. 373, 35 Sup. Ct. 342, 59 L. Ed. 628, L. R. A. 1915F, 829; Bosley v. McLaughlin, 236 U. S. 385, 35 Sup. Ct. 345, 59 L. Ed. 632; and the principle was extended to men with the allowance of a limited overtime to be paid for 'at the rate of time and one-half of the regular wage,' in Bunting v. Oregon, 243 U. S. 426, 37 Sup. Ct. 435, 61 L. Ed. 830, Ann. Cas. 1918A, 1043.

          I confess that I do not understand the principle on which the power to fix a minimum for the wages of women can be denied by those who admit the power to fix a maximum for their hours of work. I fully assent to the proposition that here as elsewhere the distinctions of the law are distinctions of degree, but I perceive no difference in the kind or degree of interference with liberty, the only matter with which we have any concern, between the one case and the other. The bargain is equally affected whichever half you regulate. Muller v. Oregon, I take it, is as good law today as it was in 1908. It will

Page 570

need more than the Nineteenth Amendment to convince me that there are no differences between men and women, or that legislation cannot take those differences into account. I should not hesitate to take them into account if I thought it necessary to sustain this Act. Quong Wing v. Kirkendall, 223 U. S. 59, 63, 32 Sup. Ct. 192, 56 L. Ed. 350. But after Bunting v. Oregon, 243 U. S. 426, 37 Sup. Ct. 435, 61 L. Ed. 830, Ann. Cas. 1918A, 1043, I had supposed that it was not necessary, and that Lochner v. New York, 198 U. S. 45, 25 Sup. Ct. 539, 49 L. Ed. 937, 3 Ann. Cas. 1133, would be allowed a deserved repose.

          This statute does not compel anybody to pay anything. It simply forbids employment at rates below those fixed as the minimum requirement of health and right living. It is safe to assume that women will not be employed at even the lowest wages allowed unless they earn them, or unless the employer's business can sustain the burden. In short the law in its character and operation is like hundreds of so-called police laws that have been upheld. I see no greater objection to using a Board to apply the standard fixed by the Act than there is to the other commissions with which we have become familiar or than there is to the requirement of a license in other cases. The fact that the statute warrants classification, which like all classifications may bear hard upon some individuals, or in exceptional cases, notwithstanding the power given to the Board to issue a special license, is no greater infirmity than is incident to all law. But the ground on which the law is held to fail is fundamental and therefore it is unnecessary to consider matters of detail.

          The criterion of constitutionality is not whether we believe the law to be for the public good. We certainly cannot be prepared to deny that a reasonable man reasonably might have that belief in view of the legislation of Great Britain, Victoria and a number of the States of this Union. The belief is fortified by a very remarkable collection of documents submitted on behalf of the appellants, material here, I conceive, only as showing that the

Page 571

belief reasonably may be held. In Australia the power to fix a minimum for wages in the case of industrial disputes extendi g beyond the limits of any one State was given to a Court, and its President wrote a most interesting account of its operation. 29 Harv. Law Rev. 13. If a legislature should adopt what he thinks the doctrine of modern economists of all schools, that 'freedom of contract is a misnomer as applied to a contract between an employer and an ordinary individual employee,' Ibid. 25, I could not pronounce an opinion with which I agree impossible to be entertained by reasonable men. If the same legislature should accept his further opinion that industrial peace was best attained by the device of a Court having the above powers, I should not feel myself able to contradict it, or to deny that the end justified restrictive legislation quite as adequately as beliefs concerning Sunday or exploded theories about usury. I should have my doubts, as I have them about this statute—but they would be whether the bill that has to be paid for every gain, although hidden as interstitial detriments, was not greater than the gain was worth: a matter that it is not for me to decide.

          I am of opinion that the statute is valid and that the decree should be reversed.

^{1 This is the exact situation in the Lyons Case, as is shown by the statement in the first part of this opinion.}

APPEAL FROM THE SUPREME COURT OF IDAHO.

Allen R. Derr argued the cause for appellant. With him on the briefs were Melvin L. Wulf, Ruth Bader Ginsburg, Pauli Murray, and Dorothy Kenyon.

Charles S. Stout argued the cause for appellee. With him on the brief was Myron E. Anderson.

Briefs of amici curiae urging reversal were filed by J. Lee Rankin and Norman Redlich for the City of New York; by Martha W. Griffiths, Phineas Indritz, Leo Kanowitz, Marguerite Rawalt, Sylvia Roberts, and Faith Seidenberg for American Veterans Committee, Inc., et al.; and by Birch Bayh for the National Federation of Business and Professional Women’s Clubs, Inc.

MR. CHIEF JUSTICE BURGER delivered the opinion of the Court.

Richard Lynn Reed, a minor, died intestate in Ada County, Idaho, on March 29, 1967. His adoptive parents, who had separated sometime prior to his death, are the parties to this appeal. Approximately seven months after Richard’s death, his mother, appellant Sally Reed, filed a petition in the Probate Court of Ada County, 72*72 seeking appointment as administratrix of her son’s estate.[1] Prior to the date set for a hearing on the mother’s petition, appellee Cecil Reed, the father of the decedent, filed a competing petition seeking to have himself appointed administrator of the son’s estate. The probate court held a joint hearing on the two petitions and thereafter ordered that letters of administration be issued to appellee Cecil Reed upon his taking the oath and filing the bond required by law. The court treated §§ 15-312 and 15-314 of the Idaho Code as the controlling statutes and read those sections as compelling a preference for Cecil Reed because he was a male.

Section 15-312² designates the persons who are entitled to administer the estate of one who dies intestate. In making these designations, that section lists 11 classes of persons who are so entitled and provides, in substance, 73*73 that the order in which those classes are listed in the section shall be determinative of the relative rights of competing applicants for letters of administration. One of the 11 classes so enumerated is ”[t]he father or mother” of the person dying intestate. Under this section, then, appellant and appellee, being members of the same entitlement class, would seem to have been equally entitled to administer their son’s estate. Section 15-314 provides, however, that

”[o]f several persons claiming and equally entitled [under § 15-312] to administer, males must be preferred to females, and relatives of the whole to those of the half blood.”

In issuing its order, the probate court implicitly recognized the equality of entitlement of the two applicants under § 15-312 and noted that neither of the applicants was under any legal disability; the court ruled, however, that appellee, being a male, was to be preferred to the female appellant “by reason of Section 15-314 of the Idaho Code.” In stating this conclusion, the probate judge gave no indication that he had attempted to determine the relative capabilities of the competing applicants to perform the functions incident to the administration of an estate. It seems clear the probate judge considered himself bound by statute to give preference to the male candidate over the female, each being otherwise “equally entitled.”

Sally Reed appealed from the probate court order, and her appeal was treated by the District Court of the Fourth Judicial District of Idaho as a constitutional attack on § 15-314. In dealing with the attack, that court held that the challenged section violated the Equal Protection Clause of the Fourteenth Amendment³ and was, therefore, 74*74 void; the matter was ordered “returned to the Probate Court for its determination of which of the two parties” was better qualified to administer the estate.

This order was never carried out, however, for Cecil Reed took a further appeal to the Idaho Supreme Court, which reversed the District Court and reinstated the original order naming the father administrator of the estate. In reaching this result, the Idaho Supreme Court first dealt with the governing statutory law and held that under § 15-312 “a father and mother are `equally entitled’ to letters of administration,” but the preference given to males by § 15-314 is “mandatory” and leaves no room for the exercise of a probate court’s discretion in the appointment of administrators. Having thus definitively and authoritatively interpreted the statutory provisions involved, the Idaho Supreme Court then proceeded to examine, and reject, Sally Reed’s contention that § 15-314 violates the Equal Protection Clause by giving a mandatory preference to males over females, without regard to their individual qualifications as potential estate administrators. 93 Idaho 511, 465 P. 2d 635.

Sally Reed thereupon appealed for review by this Court pursuant to 28 U. S. C. § 1257 (2), and we noted probable jurisdiction. 401 U. S. 934. Having examined the record and considered the briefs and oral arguments of the parties, we have concluded that the arbitrary preference established in favor of males by § 15-314 of the Idaho Code cannot stand in the face of the Fourteenth Amendment’s command that no State deny the equal protection of the laws to any person within its jurisdiction.[4]

75*75 Idaho does not, of course, deny letters of administration to women altogether. Indeed, under § 15-312, a woman whose spouse dies intestate has a preference over a son, father, brother, or any other male relative of the decedent. Moreover, we can judicially notice that in this country, presumably due to the greater longevity of women, a large proportion of estates, both intestate and under wills of decedents, are administered by surviving widows.

Section 15-314 is restricted in its operation to those situations where competing applications for letters of administration have been filed by both male and female members of the same entitlement class established by § 15-312. In such situations, § 15-314 provides that different treatment be accorded to the applicants on the basis of their sex; it thus establishes a classification subject to scrutiny under the Equal Protection Clause.

In applying that clause, this Court has consistently recognized that the Fourteenth Amendment does not deny to States the power to treat different classes of persons in different ways. Barbier v. Connolly, 113 U. S. 27 (1885); Lindsley v. Natural Carbonic Gas Co., 220 U. S. 61 (1911); Railway Express Agency v. New York, 336 U. S. 106 (1949); McDonald v. Board of Election Commissioners, 394 U. S. 802 (1969). The Equal Protection Clause of that amendment does, however, deny to States the power to legislate that different treatment be accorded to persons placed by a statute into 76*76 different classes on the basis of criteria wholly unrelated to the objective of that statute. A classification “must be reasonable, not arbitrary, and must rest upon some ground of difference having a fair and substantial relation to the object of the legislation, so that all persons similarly circumstanced shall be treated alike.” Royster Guano Co. v. Virginia, 253 U. S. 412, 415 (1920). The question presented by this case, then, is whether a difference in the sex of competing applicants for letters of administration bears a rational relationship to a state objective that is sought to be advanced by the operation of §§ 15-312 and 15-314.

In upholding the latter section, the Idaho Supreme Court concluded that its objective was to eliminate one area of controversy when two or more persons, equally entitled under § 15-312, seek letters of administration and thereby present the probate court “with the issue of which one should be named.” The court also concluded that where such persons are not of the same sex, the elimination of females from consideration “is neither an illogical nor arbitrary method devised by the legislature to resolve an issue that would otherwise require a hearing as to the relative merits . . . of the two or more petitioning relatives . . . .” 93 Idaho, at 514, 465 P. 2d, at 638.

Clearly the objective of reducing the workload on probate courts by eliminating one class of contests is not without some legitimacy. The crucial question, however, is whether § 15-314 advances that objective in a manner consistent with the command of the Equal Protection Clause. We hold that it does not. To give a mandatory preference to members of either sex over members of the other, merely to accomplish the elimination of hearings on the merits, is to make the very kind of arbitrary legislative choice forbidden by the Equal Protection Clause of the Fourteenth Amendment; and whatever may be 77*77 said as to the positive values of avoiding intrafamily controversy, the choice in this context may not lawfully be mandated solely on the basis of sex.

We note finally that if § 15-314 is viewed merely as a modifying appendage to § 15-312 and as aimed at the same objective, its constitutionality is not thereby saved. The objective of § 15-312 clearly is to establish degrees of entitlement of various classes of persons in accordance with their varying degrees and kinds of relationship to the intestate. Regardless of their sex, persons within any one of the enumerated classes of that section are similarly situated with respect to that objective. By providing dissimilar treatment for men and women who are thus similarly situated, the challenged section violates the Equal Protection Clause. Royster Guano Co. v. Virginia, supra.

The judgment of the Idaho Supreme Court is reversed and the case remanded for further proceedings not inconsistent with this opinion.

Reversed and remanded.

[1] In her petition, Sally Reed alleged that her son’s estate, consisting of a few items of personal property and a small savings account, had an aggregate value of less than $1,000.

[2] Section 15-312 provides as follows:

“Administration of the estate of a person dying intestate must be granted to some one or more of the persons hereinafter mentioned, and they are respectively entitled thereto in the following order:

“1. The surviving husband or wife or some competent person whom he or she may request to have appointed.

“2. The children.

“3. The father or mother.

“4. The brothers.

“5. The sisters.

“6. The grandchildren.

“7. The next of kin entitled to share in the distribution of the estate.

“8. Any of the kindred.

“9. The public administrator.

“10. The creditors of such person at the time of death.

“11. Any person legally competent.

“If the decedent was a member of a partnership at the time of his decease, the surviving partner must in no case be appointed administrator of his estate.”

[3] The court also held that the statute violated Art. I, § 1, of the Idaho Constitution.

[4] We note that § 15-312, set out in n. 2, supra, appears to give a superior entitlement to brothers of an intestate (class 4) than is given to sisters (class 5). The parties now before the Court are not affected by the operation of § 15-312 in this respect, however, and appellant has made no challenge to that section.

We further note that on March 12, 1971, the Idaho Legislature adopted the Uniform Probate Code, effective July 1, 1972. Idaho Laws 1971, c. 111, p. 233. On that date, §§ 15-312 and 15-314 of the present code will, then, be effectively repealed, and there is in the new legislation no mandatory preference for males over females as administrators of estates.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT

187*187 Michael E. Hobbs, Senior Assistant Attorney General of Georgia, argued the cause for petitioner. With him on the briefs were Michael J. Bowers, Attorney General, pro se, Marion O. Gordon, First Assistant Attorney General, and Daryl A. Robinson, Senior Assistant Attorney General.

Laurence H. Tribe argued the cause for respondent Hardwick. With him on the brief were Kathleen M. Sullivan and Kathleen L. Wilde.[*]

Briefs of amici curiae urging affirmance were filed for the State of New York et al. by Robert Abrams, Attorney General of New York, Robert Hermann, Solicitor General, Lawrence S. Kahn, Howard L. Zwickel, Charles R. Fraser, and Sanford M. Cohen, Assistant Attorneys General, and John Van de Kamp, Attorney General of California; for the American Jewish Congress by Daniel D. Levenson, David Cohen, and Frederick Mandel; for the American Psychological Association et al. by Margaret Farrell Ewing, Donald N. Bersoff, Anne Simon, Nadine Taub, and Herbert Semmel; for the Association of the Bar of the City of New York by Steven A. Rosen; for the National Organization for Women by John S. L. Katz; and for the Presbyterian Church (U. S. A.) et al. by Jeffrey O. Bramlett.

Briefs of amici curiae were filed for the Lesbian Rights Project et al. by Mary C. Dunlap; and for the National Gay Rights Advocates et al. by Edward P Errante, Leonard Graff, and Jay Kohorn.

JUSTICE WHITE delivered the opinion of the Court.

In August 1982, respondent Hardwick (hereafter respondent) was charged with violating the Georgia statute criminalizing 188*188 sodomy¹ by committing that act with another adult male in the bedroom of respondent’s home. After a preliminary hearing, the District Attorney decided not to present the matter to the grand jury unless further evidence developed.

Respondent then brought suit in the Federal District Court, challenging the constitutionality of the statute insofar as it criminalized consensual sodomy.[2] He asserted that he was a practicing homosexual, that the Georgia sodomy statute, as administered by the defendants, placed him in imminent danger of arrest, and that the statute for several reasons violates the Federal Constitution. The District Court granted the defendants’ motion to dismiss for failure to state a claim, relying on Doe v. Commonwealth’s Attorney for the City of Richmond, 403 F. Supp. 1199 (ED Va. 1975), which this Court summarily affirmed, 425 U. S. 901 (1976).

189*189 A divided panel of the Court of Appeals for the Eleventh Circuit reversed. 760 F. 2d 1202 (1985). The court first held that, because Doe was distinguishable and in any event had been undermined by later decisions, our summary affirmance in that case did not require affirmance of the District Court. Relying on our decisions in Griswold v. Connecticut, 381 U. S. 479 (1965); Eisenstadt v. Baird, 405 U. S. 438 (1972); Stanley v. Georgia, 394 U. S. 557 (1969); and Roe v. Wade, 410 U. S. 113 (1973), the court went on to hold that the Georgia statute violated respondent’s fundamental rights because his homosexual activity is a private and intimate association that is beyond the reach of state regulation by reason of the Ninth Amendment and the Due Process Clause of the Fourteenth Amendment. The case was remanded for trial, at which, to prevail, the State would have to prove that the statute is supported by a compelling interest and is the most narrowly drawn means of achieving that end.

Because other Courts of Appeals have arrived at judgments contrary to that of the Eleventh Circuit in this case,[3] we granted the Attorney General’s petition for certiorari questioning the holding that the sodomy statute violates the fundamental rights of homosexuals. We agree with petitioner that the Court of Appeals erred, and hence reverse its judgment.[4]

190*190 This case does not require a judgment on whether laws against sodomy between consenting adults in general, or between homosexuals in particular, are wise or desirable. It raises no question about the right or propriety of state legislative decisions to repeal their laws that criminalize homosexual sodomy, or of state-court decisions invalidating those laws on state constitutional grounds. The issue presented is whether the Federal Constitution confers a fundamental right upon homosexuals to engage in sodomy and hence invalidates the laws of the many States that still make such conduct illegal and have done so for a very long time. The case also calls for some judgment about the limits of the Court’s role in carrying out its constitutional mandate.

We first register our disagreement with the Court of Appeals and with respondent that the Court’s prior cases have construed the Constitution to confer a right of privacy that extends to homosexual sodomy and for all intents and purposes have decided this case. The reach of this line of cases was sketched in Carey v. Population Services International, 431 U. S. 678, 685 (1977). Pierce v. Society of Sisters, 268 U. S. 510 (1925), and Meyer v. Nebraska, 262 U. S. 390 (1923), were described as dealing with child rearing and education; Prince v. Massachusetts, 321 U. S. 158 (1944), with family relationships; Skinner v. Oklahoma ex rel. Williamson, 316 U. S. 535 (1942), with procreation; Loving v. Virginia, 388 U. S. 1 (1967), with marriage; Griswold v. Connecticut, supra, and Eisenstadt v. Baird, supra, with contraception; and Roe v. Wade, 410 U. S. 113 (1973), with abortion. The latter three cases were interpreted as construing the Due Process Clause of the Fourteenth Amendment to confer a fundamental individual right to decide whether or not to beget or bear a child. Carey v. Population Services International, supra, at 688-689.

Accepting the decisions in these cases and the above description of them, we think it evident that none of the rights announced in those cases bears any resemblance to the 191*191 claimed constitutional right of homosexuals to engage in acts of sodomy that is asserted in this case. No connection between family, marriage, or procreation on the one hand and homosexual activity on the other has been demonstrated, either by the Court of Appeals or by respondent. Moreover, any claim that these cases nevertheless stand for the proposition that any kind of private sexual conduct between consenting adults is constitutionally insulated from state proscription is unsupportable. Indeed, the Court’s opinion in Carey twice asserted that the privacy right, which the Griswold line of cases found to be one of the protections provided by the Due Process Clause, did not reach so far. 431 U. S., at 688, n. 5, 694, n. 17.

Precedent aside, however, respondent would have us announce, as the Court of Appeals did, a fundamental right to engage in homosexual sodomy. This we are quite unwilling to do. It is true that despite the language of the Due Process Clauses of the Fifth and Fourteenth Amendments, which appears to focus only on the processes by which life, liberty, or property is taken, the cases are legion in which those Clauses have been interpreted to have substantive content, subsuming rights that to a great extent are immune from federal or state regulation or proscription. Among such cases are those recognizing rights that have little or no textual support in the constitutional language. Meyer, Prince, and Pierce fall in this category, as do the privacy cases from Griswold to Carey.

Striving to assure itself and the public that announcing rights not readily identifiable in the Constitution’s text involves much more than the imposition of the Justices’ own choice of values on the States and the Federal Government, the Court has sought to identify the nature of the rights qualifying for heightened judicial protection. In Palko v. Connecticut, 302 U. S. 319, 325, 326 (1937), it was said that this category includes those fundamental liberties that are “implicit in the concept of ordered liberty,” such that “neither 192*192 liberty nor justice would exist if [they] were sacrificed.” A different description of fundamental liberties appeared in Moore v. East Cleveland, 431 U. S. 494, 503 (1977) (opinion of POWELL, J.), where they are characterized as those liberties that are “deeply rooted in this Nation’s history and tradition.” Id., at 503 (POWELL, J.). See also Griswold v. Connecticut, 381 U. S., at 506.

It is obvious to us that neither of these formulations would extend a fundamental right to homosexuals to engage in acts of consensual sodomy. Proscriptions against that conduct have ancient roots. See generally Survey on the Constitutional Right to Privacy in the Context of Homosexual Activity, 40 U. Miami L. Rev. 521, 525 (1986). Sodomy was a criminal offense at common law and was forbidden by the laws of the original 13 States when they ratified the Bill of Rights.[5] In 1868, when the Fourteenth Amendment was 193*193 ratified, all but 5 of the 37 States in the Union had criminal sodomy laws.[6] In fact, until 1961,[7] all 50 States outlawed sodomy, and today, 24 States and the District of Columbia 194*194 continue to provide criminal penalties for sodomy performed in private and between consenting adults. See Survey, U. Miami L. Rev., supra, at 524, n. 9. Against this background, to claim that a right to engage in such conduct is “deeply rooted in this Nation’s history and tradition” or “implicit in the concept of ordered liberty” is, at best, facetious.

Nor are we inclined to take a more expansive view of our authority to discover new fundamental rights imbedded in the Due Process Clause. The Court is most vulnerable and comes nearest to illegitimacy when it deals with judge-made constitutional law having little or no cognizable roots in the language or design of the Constitution. That this is so was painfully demonstrated by the face-off between the Executive and the Court in the 1930’s, which resulted in the repudiation 195*195 of much of the substantive gloss that the Court had placed on the Due Process Clauses of the Fifth and Fourteenth Amendments. There should be, therefore, great resistance to expand the substantive reach of those Clauses, particularly if it requires redefining the category of rights deemed to be fundamental. Otherwise, the Judiciary necessarily takes to itself further authority to govern the country without express constitutional authority. The claimed right pressed on us today falls for short of overcoming this resistance.

Respondent, however, asserts that the result should be different where the homosexual conduct occurs in the privacy of the home. He relies on Stanley v. Georgia, 394 U. S. 557 (1969), where the Court held that the First Amendment prevents conviction for possessing and reading obscene material in the privacy of one’s home: “If the First Amendment means anything, it means that a State has no business telling a man, sitting alone in his house, what books he may read or what films he may watch.” Id., at 565.

Stanley did protect conduct that would not have been protected outside the home, and it partially prevented the enforcement of state obscenity laws; but the decision was firmly grounded in the First Amendment. The right pressed upon us here has no similar support in the text of the Constitution, and it does not qualify for recognition under the prevailing principles for construing the Fourteenth Amendment. Its limits are also difficult to discern. Plainly enough, otherwise illegal conduct is not always immunized whenever it occurs in the home. Victimless crimes, such as the possession and use of illegal drugs, do not escape the law where they are committed at home. Stanley itself recognized that its holding offered no protection for the possession in the home of drugs, firearms, or stolen goods. Id., at 568, n. 11. And if respondent’s submission is limited to the voluntary sexual conduct between consenting adults, it would be difficult, except by fiat, to limit the claimed right to homosexual conduct 196*196 while leaving exposed to prosecution adultery, incest, and other sexual crimes even though they are committed in the home. We are unwilling to start down that road.

Even if the conduct at issue here is not a fundamental right, respondent asserts that there must be a rational basis for the law and that there is none in this case other than the presumed belief of a majority of the electorate in Georgia that homosexual sodomy is immoral and unacceptable. This is said to be an inadequate rationale to support the law. 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. Even respondent makes no such claim, but insists that majority sentiments about the morality of homosexuality should be declared inadequate. We do not agree, and are unpersuaded that the sodomy laws of some 25 States should be invalidated on this basis.[8]

Accordingly, the judgment of the Court of Appeals is

Reversed.

CHIEF JUSTICE BURGER, concurring.

I join the Court’s opinion, but I write separately to underscore my view that in constitutional terms there is no such thing as a fundamental right to commit homosexual sodomy.

As the Court notes, ante, at 192, the proscriptions against sodomy have very “ancient roots.” Decisions of individuals relating to homosexual conduct have been subject to state intervention throughout the history of Western civilization. Condemnation of those practices is firmly rooted in Judeao-Christian moral and ethical standards. Homosexual sodomy was a capital crime under Roman law. See Code Theod. 9.7.6; Code Just. 9.9.31. See also D. Bailey, Homosexuality 197*197 and the Western Christian Tradition 70-81 (1975). During the English Reformation when powers of the ecclesiastical courts were transferred to the King’s Courts, the first English statute criminalizing sodomy was passed. 25 Hen. VIII, ch. 6. Blackstone described “the infamous crime against nature” as an offense of “deeper malignity” than rape, a heinous act “the very mention of which is a disgrace to human nature,” and “a crime not fit to be named.” 4 W. Blackstone, Commentaries *215. The common law of England, including its prohibition of sodomy, became the received law of Georgia and the other Colonies. In 1816 the Georgia Legislature passed the statute at issue here, and that statute has been continuously in force in one form or another since that time. To hold that the act of homosexual sodomy is somehow protected as a fundamental right would be to cast aside millennia of moral teaching.

This is essentially not a question of personal “preferences” but rather of the legislative authority of the State. I find nothing in the Constitution depriving a State of the power to enact the statute challenged here.

JUSTICE POWELL, concurring.

I join the opinion of the Court. I agree with the Court that there is no fundamental right — i. e., no substantive right under the Due Process Clause — such as that claimed by respondent Hardwick, and found to exist by the Court of Appeals. This is not to suggest, however, that respondent may not be protected by the Eighth Amendment of the Constitution. The Georgia statute at issue in this case, Ga. Code Ann. § 16-6-2 (1984), authorizes a court to imprison a person for up to 20 years for a single private, consensual act of sodomy. In my view, a prison sentence for such conduct — certainly a sentence of long duration — would create a serious Eighth Amendment issue. Under the Georgia statute a single act of sodomy, even in the private setting of a home, is a 198*198 felony comparable in terms of the possible sentence imposed to serious felonies such as aggravated battery, § 16-5-24, first-degree arson, § 16-7-60, and robbery, § 16-8-40.[1]

In this case, however, respondent has not been tried, much less convicted and sentenced.[2] Moreover, respondent has not raised the Eighth Amendment issue below. For these reasons this constitutional argument is not before us.

199*199 JUSTICE BLACKMUN, with whom JUSTICE BRENNAN, JUSTICE MARSHALL, and JUSTICE STEVENS join, dissenting.

This case is no more about “a fundamental right to engage in homosexual sodomy,” as the Court purports to declare, ante, at 191, than Stanley v. Georgia, 394 U. S. 557 (1969), was about a fundamental right to watch obscene movies, or Katz v. United States, 389 U. S. 347 (1967), was about a fundamental right to place interstate bets from a telephone booth. Rather, this case is about “the most comprehensive of rights and the right most valued by civilized men,” namely, “the right to be let alone.” Olmstead v. United States, 277 U. S. 438, 478 (1928) (Brandeis, J., dissenting).

The statute at issue, Ga. Code Ann. § 16-6-2 (1984), denies individuals the right to decide for themselves whether to engage in particular forms of private, consensual sexual activity. The Court concludes that § 16-6-2 is valid essentially because “the laws of . . . many States . . . still make such conduct illegal and have done so for a very long time.” Ante, at 190. But the fact that the moral judgments expressed by statutes like § 16-6-2 may be ” `natural and familiar . . . ought not to conclude our judgment upon the question whether statutes embodying them conflict with the Constitution of the United States.’ ” Roe v. Wade, 410 U. S. 113, 117 (1973), quoting Lochner v. New York, 198 U. S. 45, 76 (1905) (Holmes, J., dissenting). Like Justice Holmes, I believe that ”[i]t is revolting to have no better reason for a rule of law than that so it was laid down in the time of Henry IV. It is still more revolting if the grounds upon which it was laid down have vanished long since, and the rule simply persists from blind imitation of the past.” Holmes, The Path of the Law, 10 Harv. L. Rev. 457, 469 (1897). I believe we must analyze respondent Hardwick’s claim in the light of the values that underlie the constitutional right to privacy. If that right means anything, it means that, before Georgia can prosecute its citizens for making choices about the most intimate 200*200 aspects of their lives, it must do more than assert that the choice they have made is an ” `abominable crime not fit to be named among Christians.’ ” Herring v. State, 119 Ga. 709, 721, 46 S. E. 876, 882 (1904).

I

In its haste to reverse the Court of Appeals and hold that the Constitution does not “confe® a fundamental right upon homosexuals to engage in sodomy,” ante, at 190, the Court relegates the actual statute being challenged to a footnote and ignores the procedural posture of the case before it. A fair reading of the statute and of the complaint clearly reveals that the majority has distorted the question this case presents.

First, the Court’s almost obsessive focus on homosexual activity is particularly hard to justify in light of the broad language Georgia has used. Unlike the Court, the Georgia Legislature has not proceeded on the assumption that homosexuals are so different from other citizens that their lives may be controlled in a way that would not be tolerated if it limited the choices of those other citizens. Cf. ante, at 188, n. 2. Rather, Georgia has provided that ”[a] person commits the offense of sodomy when he performs or submits to any sexual act involving the sex organs of one person and the mouth or anus of another.” Ga. Code Ann. § 16-6-2(a) (1984). The sex or status of the persons who engage in the act is irrelevant as a matter of state law. In fact, to the extent I can discern a legislative purpose for Georgia’s 1968 enactment of § 16-6-2, that purpose seems to have been to broaden the coverage of the law to reach heterosexual as well as homosexual activity.[1] I therefore see no basis for the 201*201 Court’s decision to treat this case as an “as applied” challenge to § 16-6-2, see ante, at 188, n. 2, or for Georgia’s attempt, both in its brief and at oral argument, to defend § 16-6-2 solely on the grounds that it prohibits homosexual activity. Michael Hardwick’s standing may rest in significant part on Georgia’s apparent willingness to enforce against homosexuals a law it seems not to have any desire to enforce against heterosexuals. See Tr. of Oral Arg. 4-5; cf. 760 F. 2d 1202, 1205-1206 (CA11 1985). But his claim that § 16-6-2 involves an unconstitutional intrusion into his privacy and his right of intimate association does not depend in any way on his sexual orientation.

Second, I disagree with the Court’s refusal to consider whether § 16-6-2 runs afoul of the Eighth or Ninth Amendments or the Equal Protection Clause of the Fourteenth Amendment. Ante, at 196, n. 8. Respondent’s complaint expressly invoked the Ninth Amendment, see App. 6, and he relied heavily before this Court on Griswold v. Connecticut, 381 U. S. 479, 484 (1965), which identifies that Amendment as one of the specific constitutional provisions giving “life and substance” to our understanding of privacy. See Brief for Respondent Hardwick 10-12; Tr. of Oral Arg. 33. More importantly, the procedural posture of the case requires that we affirm the Court of Appeals’ judgment if there is any ground on which respondent may be entitled to relief. This case is before us on petitioner’s motion to dismiss for failure to state a claim, Fed. Rule Civ. Proc. 12(b)(6). See App. 17. It is a well-settled principle of law that “a complaint should not be dismissed merely because a plaintiff’s allegations do not support the particular legal theory he advances, for the court is under a duty to examine the complaint to determine if the allegations provide for relief on any possible theory.” 202*202 Bramlet v. Wilson, 495 F. 2d 714, 716 (CA8 1974); see Parr v. Great Lakes Express Co., 484 F. 2d 767, 773 (CA7 1973); Due v. Tallahassee Theaters, Inc., 333 F. 2d 630, 631 (CA5 1964); United States v. Howell, 318 F. 2d 162, 166 (CA9 1963); 5 C. Wright & A. Miller, Federal Practice and Procedure § 1357, pp. 601-602 (1969); see also Conley v. Gibson, 355 U. S. 41, 45-46 (1957). Thus, even if respondent did not advance claims based on the Eighth or Ninth Amendments, or on the Equal Protection Clause, his complaint should not be dismissed if any of those provisions could entitle him to relief. I need not reach either the Eighth Amendment or the Equal Protection Clause issues because I believe that Hardwick has stated a cognizable claim that § 16-6-2 interferes with constitutionally protected interests in privacy and freedom of intimate association. But neither the Eighth Amendment nor the Equal Protection Clause is so clearly irrelevant that a claim resting on either provision should be peremptorily dismissed.[2] The Court’s cramped reading of the 203*203 issue before it makes for a short opinion, but it does little to make for a persuasive one.

II

“Our cases long have recognized that the Constitution embodies a promise that a certain private sphere of individual liberty will be kept largely beyond the reach of government.” Thornburgh v. American College of Obstetricians & Gynecologists, 476 U. S. 747, 772 (1986). In construing the right to privacy, the Court has proceeded along two somewhat distinct, 204*204 albeit complementary, lines. First, it has recognized a privacy interest with reference to certain decisions that are properly for the individual to make. E. g., Roe v. Wade, 410 U. S. 113 (1973); Pierce v. Society of Sisters, 268 U. S. 510 (1925). Second, it has recognized a privacy interest with reference to certain places without regard for the particular activities in which the individuals who occupy them are engaged. E. g., United States v. Karo, 468 U. S. 705 (1984); Payton v. New York, 445 U. S. 573 (1980); Rios v. United States, 364 U. S. 253 (1960). The case before us implicates both the decisional and the spatial aspects of the right to privacy.

A

The Court concludes today that none of our prior cases dealing with various decisions that individuals are entitled to make free of governmental interference “bears any resemblance to the claimed constitutional right of homosexuals to engage in acts of sodomy that is asserted in this case.” Ante, at 190-191. While it is true that these cases may be characterized by their connection to protection of the family, see Roberts v. United States Jaycees, 468 U. S. 609, 619 (1984), the Court’s conclusion that they extend no further than this boundary ignores the warning in Moore v. East Cleveland, 431 U. S. 494, 501 (1977) (plurality opinion), against “clos[ing] our eyes to the basic reasons why certain rights associated with the family have been accorded shelter under the Fourteenth Amendment’s Due Process Clause.” We protect those rights not because they contribute, in some direct and material way, to the general public welfare, but because they form so central a part of an individual’s life. ”[T]he concept of privacy embodies the `moral fact that a person belongs to himself and not others nor to society as a whole.’ ” Thornburgh v. American College of Obstetricians & Gynecologists, 476 U. S., at 777, n. 5 (STEVENS, J., concurring), quoting Fried, Correspondence, 6 Phil. & Pub. Affairs 288-289 (1977). And so we protect the decision whether to 205*205 marry precisely because marriage “is an association that promotes a way of life, not causes; a harmony in living, not political faiths; a bilateral loyalty, not commercial or social projects.” Griswold v. Connecticut, 381 U. S., at 486. We protect the decision whether to have a child because parenthood alters so dramatically an individual’s self-definition, not because of demographic considerations or the Bible’s command to be fruitful and multiply. Cf. Thornburgh v. American College of Obstetricians & Gynecologists, supra, at 777, n. 6 (STEVENS, J., concurring). And we protect the family because it contributes so powerfully to the happiness of individuals, not because of a preference for stereotypical households. Cf. Moore v. East Cleveland, 431 U. S., at 500-506 (plurality opinion). The Court recognized in Roberts, 468 U. S., at 619, that the “ability independently to define one’s identity that is central to any concept of liberty” cannot truly be exercised in a vacuum; we all depend on the “emotional enrichment from close ties with others.” Ibid.

Only the most willful blindness could obscure the fact that sexual intimacy is “a sensitive, key relationship of human existence, central to family life, community welfare, and the development of human personality,” Paris Adult Theatre I v. Slaton, 413 U. S. 49, 63 (1973); see also Carey v. Population Services International, 431 U. S. 678, 685 (1977). The fact that individuals define themselves in a significant way through their intimate sexual relationships with others suggests, in a Nation as diverse as ours, that there may be many “right” ways of conducting those relationships, and that much of the richness of a relationship will come from the freedom an individual has to choose the form and nature of these intensely personal bonds. See Karst, The Freedom of Intimate Association, 89 Yale L. J. 624, 637 (1980); cf. Eisenstadt v. Baird, 405 U. S. 438, 453 (1972); Roe v. Wade, 410 U. S., at 153.

In a variety of circumstances we have recognized that a necessary corollary of giving individuals freedom to choose 206*206 how to conduct their lives is acceptance of the fact that different individuals will make different choices. For example, in holding that the clearly important state interest in public education should give way to a competing claim by the Amish to the effect that extended formal schooling threatened their way of life, the Court declared: “There can be no assumption that today’s majority is `right’ and the Amish and others like them are `wrong.’ A way of life that is odd or even erratic but interferes with no rights or interests of others is not to be condemned because it is different.” Wisconsin v. Yoder, 406 U. S. 205, 223-224 (1972). The Court claims that its decision today merely refuses to recognize a fundamental right to engage in homosexual sodomy; what the Court really has refused to recognize is the fundamental interest all individuals have in controlling the nature of their intimate associations with others.

B

The behavior for which Hardwick faces prosecution occurred in his own home, a place to which the Fourth Amendment attaches special significance. The Court’s treatment of this aspect of the case is symptomatic of its overall refusal to consider the broad principles that have informed our treatment of privacy in specific cases. Just as the right to privacy is more than the mere aggregation of a number of entitlements to engage in specific behavior, so too, protecting the physical integrity of the home is more than merely a means of protecting specific activities that often take place there. Even when our understanding of the contours of the right to privacy depends on “reference to a `place,’ ” Katz v. United States, 389 U. S., at 361 (Harlan, J., concurring), “the essence of a Fourth Amendment violation is `not the breaking of [a person’s] doors, and the rummaging of his drawers,’ but rather is `the invasion of his indefensible right of personal security, personal liberty and private property.’ ” California v. Ciraolo, 476 U. S. 207, 226 (1986) (POWELL, J., dissenting), 207*207 quoting Boyd v. United States, 116 U. S. 616, 630 (1886).

The Court’s interpretation of the pivotal case of Stanley v. Georgia, 394 U. S. 557 (1969), is entirely unconvincing. Stanley held that Georgia’s undoubted power to punish the public distribution of constitutionally unprotected, obscene material did not permit the State to punish the private possession of such material. According to the majority here, Stanley relied entirely on the First Amendment, and thus, it is claimed, sheds no light on cases not involving printed materials. Ante, at 195. But that is not what Stanley said. Rather, the Stanley Court anchored its holding in the Fourth Amendment’s special protection for the individual in his home:

” `The makers of our Constitution undertook to secure conditions favorable to the pursuit of happiness. They recognized the significance of man’s spiritual nature, of his feelings and of his intellect. They knew that only a part of the pain, pleasure and satisfactions of life are to be found in material things. They sought to protect Americans in their beliefs, their thoughts, their emotions and their sensations.’ ..... “These are the rights that appellant is asserting in the case before us. He is asserting the right to read or observe what he pleases — the right to satisfy his intellectual and emotional needs in the privacy of his own home.” 394 U. S., at 564-565, quoting Olmstead v. United States, 277 U. S., at 478 (Brandeis, J., dissenting).

The central place that Stanley gives Justice Brandeis’ dissent in Olmstead, a case raising no First Amendment claim, shows that Stanley rested as much on the Court’s understanding of the Fourth Amendment as it did on the First. Indeed, in Paris Adult Theatre I v. Slaton, 413 U. S. 49 (1973), the Court suggested that reliance on the Fourth 208*208 Amendment not only supported the Court’s outcome in Stanley but actually was necessary to it: “If obscene material unprotected by the First Amendment in itself carried with it a `penumbra’ of constitutionally protected privacy, this Court would not have found it necessary to decide Stanley on the narrow basis of the `privacy of the home,’ which was hardly more than a reaffirmation that `a man’s home is his castle.’ ” 413 U. S., at 66. “The right of the people to be secure in their . . . houses,” expressly guaranteed by the Fourth Amendment, is perhaps the most “textual” of the various constitutional provisions that inform our understanding of the right to privacy, and thus I cannot agree with the Court’s statement that ”[t]he right pressed upon us here has no . . . support in the text of the Constitution,” ante, at 195. Indeed, the right of an individual to conduct intimate relationships in the intimacy of his or her own home seems to me to be the heart of the Constitution’s protection of privacy.

III

The Court’s failure to comprehend the magnitude of the liberty interests at stake in this case leads it to slight the question whether petitioner, on behalf of the State, has justified Georgia’s infringement on these interests. I believe that neither of the two general justifications for § 16-6-2 that petitioner has advanced warrants dismissing respondent’s challenge for failure to state a claim.

First, petitioner asserts that the acts made criminal by the statute may have serious adverse consequences for “the general public health and welfare,” such as spreading communicable diseases or fostering other criminal activity. Brief for Petitioner 37. Inasmuch as this case was dismissed by the District Court on the pleading, it is not surprising that the record before us is barren of any evidence to support petitioner’s claim.[3] In light of the state of the record, I see 209*209 no justification for the Court’s attempt to equate the private, consensual sexual activity at issue here with the “possession in the home of drugs, firearms, or stolen goods,” ante, at 195, to which Stanley refused to extend its protection. 394 U. S., at 568, n. 11. None of the behavior so mentioned in Stanley can properly be viewed as ”[v]ictimless,” ante, at 195: drugs and weapons are inherently dangerous, see, e. g., McLaughlin v. United States, 476 U. S. 16 (1986), and for property to be “stolen,” someone must have been wrongfully deprived of it. Nothing in the record before the Court provides any justification for finding the activity forbidden by § 16-6-2 to be physically dangerous, either to the persons engaged in it or to others.[4]

210*210 The core of petitioner’s defense of § 16-6-2, however, is that respondent and others who engage in the conduct prohibited by § 16-6-2 interfere with Georgia’s exercise of the ” `right of the Nation and of the States to maintain a decent society,’ ” Paris Adult Theater I v. Slaton, 413 U. S., at 59-60, quoting Jacobellis v. Ohio, 378 U. S. 184, 199 (1964) (Warren, C. J., dissenting). Essentially, petitioner argues, and the Court agrees, that the fact that the acts described in § 16-6-2 “for hundreds of years, if not thousands, have been uniformly condemned as immoral” is a sufficient reason to permit a State to ban them today. Brief for Petitioner 19; see ante, at 190, 192-194, 196.

I cannot agree that either the length of time a majority has held its convictions or the passions with which it defends them can withdraw legislation from this Court’s security. See, e. g., Roe v. Wade, 410 U. S. 113 (1973); Loving v. Virginia, 388 U. S. 1 (1967); Brown v. Board of Education, 347 U. S. 483 (1954).[5] As Justice Jackson wrote so eloquently 211*211 for the Court in West Virginia Board of Education v. Barnette, 319 U. S. 624, 641-642 (1943), “we apply the limitations of the Constitution with no fear that freedom to be intellectually and spiritually diverse or even contrary will disintegrate the social organization. . . . [F]reedom to differ is not limited to things that do not matter much. That would be a mere shadow of freedom. The test of its substance is the right to differ as to things that touch the heart of the existing order.” See also Karst, 89 Yale L. J., at 627. It is precisely because the issue raised by this case touches the heart of what makes individuals what they are that we should be especially sensitive to the rights of those whose choices upset the majority.

The assertion that “traditional Judeo-Christian values proscribe” the conduct involved, Brief for Petitioner 20, cannot provide an adequate justification for § 16-6-2. That certain, but by no means all, religious groups condemn the behavior at issue gives the State no license to impose their judgments on the entire citizenry. The legitimacy of secular legislation depends instead on whether the State can advance some justification for its law beyond its conformity to religious doctrine. See, e. g., McGowan v. Maryland, 366 U. S. 420, 429-453 (1961); Stone v. Graham, 449 U. S. 39 (1980). Thus, far from buttressing his case, petitioner’s invocation of Leviticus, Romans, St. Thomas Aquinas, and sodomy’s heretical status during the Middle Ages undermines his suggestion that § 16-6-2 represents a legitimate use of secular coercive power.[6] A State can no more punish private behavior because 212*212 of religious intolerance than it can punish such behavior because of racial animus. “The Constitution cannot control such prejudices, but neither can it tolerate them. Private biases may be outside the reach of the law, but the law cannot, directly or indirectly, give them effect.” Palmore v. Sidoti, 466 U. S. 429, 433 (1984). No matter how uncomfortable a certain group may make the majority of this Court, we have held that ”[m]ere public intolerance or animosity cannot constitutionally justify the deprivation of a person’s physical liberty.” O’Connor v. Donaldson, 422 U. S. 563, 575 (1975). See also Cleburne v. Cleburne Living Center, Inc., 473 U. S. 432 (1985); United States Dept. of Agriculture v. Moreno, 413 U. S. 528, 534 (1973).

Nor can § 16-6-2 be justified as a “morally neutral” exercise of Georgia’s power to “protect the public environment,” Paris Adult Theatre I, 413 U. S., at 68-69. Certainly, some private behavior can affect the fabric of society as a whole. Reasonable people may differ about whether particular sexual acts are moral or immoral, but “we have ample evidence for believing that people will not abandon morality, will not think any better of murder, cruelty and dishonesty, merely because some private sexual practice which they abominate is not punished by the law.” H. L. A. Hart, Immorality and Treason, reprinted in The Law as Literature 220, 225 (L. Blom-Cooper ed. 1961). Petitioner and the Court fail to see the difference between laws that protect public sensibilities and those that enforce private morality. Statutes banning 213*213 public sexual activity are entirely consistent with protecting the individual’s liberty interest in decisions concerning sexual relations: the same recognition that those decisions are intensely private which justifies protecting them from governmental interference can justify protecting individuals from unwilling exposure to the sexual activities of others. But the mere fact that intimate behavior may be punished when it takes place in public cannot dictate how States can regulate intimate behavior that occurs in intimate places. See Paris Adult Theatre I, 413 U. S., at 66, n. 13 (“marital intercourse on a street corner or a theater stage” can be forbidden despite the constitutional protection identified in Griswold v. Connecticut, 381 U. S. 479 (1965)).[7]

This case involves no real interference with the rights of others, for the mere knowledge that other individuals do not adhere to one’s value system cannot be a legally cognizable interest, cf. Diamond v. Charles, 476 U. S. 54, 65-66 (1986), let alone an interest that can justify invading the houses, hearts, and minds of citizens who choose to live their lives differently.

IV

It took but three years for the Court to see the error in its analysis in Minersville School District v. Gobitis, 310 U. S. 214*214 586 (1940), and to recognize that the threat to national cohesion posed by a refusal to salute the flag was vastly outweighed by the threat to those same values posed by compelling such a salute. See West Virginia Board of Education v. Barnette, 319 U. S. 624 (1943). I can only hope that here, too, the Court soon will reconsider its analysis and conclude that depriving individuals of the right to choose for themselves how to conduct their intimate relationships poses a far greater threat to the values most deeply rooted in our Nation’s history than tolerance of nonconformity could ever do. Because I think the Court today betrays those values, I dissent.

JUSTICE STEVENS, with whom JUSTICE BRENNAN and JUSTICE MARSHALL join, dissenting.

Like the statute that is challenged in this case,[1] the rationale of the Court’s opinion applies equally to the prohibited conduct regardless of whether the parties who engage in it are married or unmarried, or are of the same or different sexes.[2] Sodomy was condemned as an odious and sinful type of behavior during the formative period of the common law.[3] 215*215 That condemnation was equally damning for heterosexual and homosexual sodomy.[4] Moreover, it provided no special exemption for married couples.[5] The license to cohabit and to produce legitimate offspring simply did not include any permission to engage in sexual conduct that was considered a “crime against nature.”

The history of the Georgia statute before us clearly reveals this traditional prohibition of heterosexual, as well as homosexual, sodomy.[6] Indeed, at one point in the 20th century, Georgia’s law was construed to permit certain sexual conduct between homosexual women even though such conduct was prohibited between heterosexuals.[7] The history of the statutes cited by the majority as proof for the proposition that sodomy is not constitutionally protected, ante, at 192-194, 216*216 and nn. 5 and 6, similarly reveals a prohibition on heterosexual, as well as homosexual, sodomy.[8]

Because the Georgia statute expresses the traditional view that sodomy is an immoral kind of conduct regardless of the identity of the persons who engage in it, I believe that a proper analysis of its constitutionality requires consideration of two questions: First, may a State totally prohibit the described conduct by means of a neutral law applying without exception to all persons subject to its jurisdiction? If not, may the State save the statute by announcing that it will only enforce the law against homosexuals? The two questions merit separate discussion.

I

Our prior cases make two propositions abundantly clear. First, the fact that the governing majority in a State has traditionally viewed a particular practice as immoral is not a sufficient reason for upholding a law prohibiting the practice; neither history nor tradition could save a law prohibiting miscegenation from constitutional attack.[9] Second, individual decisions by married persons, concerning the intimacies of their physical relationship, even when not intended to produce offspring, are a form of “liberty” protected by the Due Process Clause of the Fourteenth Amendment. Griswold v. Connecticut, 381 U. S. 479 (1965). Moreover, this protection extends to intimate choices by unmarried as well as married persons. Carey v. Population Services International, 431 U. S. 678 (1977); Eisenstadt v. Baird, 405 U. S. 438 (1972).

217*217 In consideration of claims of this kind, the Court has emphasized the individual interest in privacy, but its decisions have actually been animated by an even more fundamental concern. As I wrote some years ago:

“These cases do not deal with the individual’s interest in protection from unwarranted public attention, comment, or exploitation. They deal, rather, with the individual’s right to make certain unusually important decisions that will affect his own, or his family’s destiny. The Court has referred to such decisions as implicating `basic values,’ as being `fundamental,’ and as being dignified by history and tradition. The character of the Court’s language in these cases brings to mind the origins of the American heritage of freedom — the abiding interest in individual liberty that makes certain state intrusions on the citizen’s right to decide how he will live his own life intolerable. Guided by history, our tradition of respect for the dignity of individual choice in matters of conscience and the restraints implicit in the federal system, federal judges have accepted the responsibility for recognition and protection of these rights in appropriate cases.” Fitzgerald v. Porter Memorial Hospital, 523 F. 2d 716, 719-720 (CA7 1975) (footnotes omitted), cert. denied, 425 U. S. 916 (1976).

Society has every right to encourage its individual members to follow particular traditions in expressing affection for one another and in gratifying their personal desires. It, of course, may prohibit an individual from imposing his will on another to satisfy his own selfish interests. It also may prevent an individual from interfering with, or violating, a legally sanctioned and protected relationship, such as marriage. And it may explain the relative advantages and disadvantages of different forms of intimate expression. But when individual married couples are isolated from observation by others, the way in which they voluntarily choose to conduct their intimate relations is a matter for them — not the 218*218 State — to decide.[10] The essential “liberty” that animated the development of the law in cases like Griswold, Eisenstadt, and Carey surely embraces the right to engage in nonreproductive, sexual conduct that others may consider offensive or immoral.

Paradoxical as it may seem, our prior cases thus establish that a State may not prohibit sodomy within “the sacred precincts of marital bedrooms,” Griswold, 381 U. S., at 485, or, indeed, between unmarried heterosexual adults. Eisenstadt, 405 U. S., at 453. In all events, it is perfectly clear that the State of Georgia may not totally prohibit the conduct proscribed by § 16-6-2 of the Georgia Criminal Code.

II

If the Georgia statute cannot be enforced as it is written — if the conduct it seeks to prohibit is a protected form of liberty for the vast majority of Georgia’s citizens — the State must assume the burden of justifying a selective application of its law. Either the persons to whom Georgia seeks to apply its statute do not have the same interest in “liberty” that others have, or there must be a reason why the State may be permitted to apply a generally applicable law to certain persons that it does not apply to others.

The first possibility is plainly unacceptable. Although the meaning of the principle that “all men are created equal” is not always clear, it surely must mean that every free citizen has the same interest in “liberty” that the members of the majority share. From the standpoint of the individual, the homosexual and the heterosexual have the same interest in deciding how he will live his own life, and, more narrowly, how he will conduct himself in his personal and voluntary 219*219 associations with his companions. State intrusion into the private conduct of either is equally burdensome.

The second possibility is similarly unacceptable. A policy of selective application must be supported by a neutral and legitimate interest — something more substantial than a habitual dislike for, or ignorance about, the disfavored group. Neither the State nor the Court has identified any such interest in this case. The Court has posited as a justification for the Georgia statute “the presumed belief of a majority of the electorate in Georgia that homosexual sodomy is immoral and unacceptable.” Ante, at 196. But the Georgia electorate has expressed no such belief — instead, its representatives enacted a law that presumably reflects the belief that all sodomy is immoral and unacceptable. Unless the Court is prepared to conclude that such a law is constitutional, it may not rely on the work product of the Georgia Legislature to support its holding. For the Georgia statute does not single out homosexuals as a separate class meriting special disfavored treatment.

Nor, indeed, does not Georgia prosecutor even believe that all homosexuals who violate this statute should be punished. This conclusion is evident from the fact that the respondent in this very case has formally acknowledged in his complaint and in court that he has engaged, and intends to continue to engage, in the prohibited conduct, yet the State has elected not to process criminal charges against him. As JUSTICE POWELL points out, moreover, Georgia’s prohibition on private, consensual sodomy has not been enforced for decades.[11] The record of nonenforcement, in this case and in the last several decades, belies the Attorney General’s representations 220*220 about the importance of the State’s selective application of its generally applicable law.[12]

Both the Georgia statute and the Georgia prosecutor thus completely fail to provide the Court with any support for the conclusion that homosexual sodomy, simpliciter, is considered unacceptable conduct in that State, and that the burden of justifying a selective application of the generally applicable law has been met.

III

The Court orders the dismissal of respondent’s complaint even though the State’s statute prohibits all sodomy; even though that prohibition is concededly unconstitutional with respect to heterosexuals; and even though the State’s post hoc explanations for selective application are belied by the State’s own actions. At the very least, I think it clear at this early stage of the litigation that respondent has alleged a constitutional claim sufficient to withstand a motion to dismiss.[13]

I respectfully dissent.

[*] Briefs of amici curiae urging reversal were filed for the Catholic League for Religious and Civil Rights by Steven Frederick McDowell; for the Rutherford Institute et al. by W. Charles Bundren, Guy O. Farley, Jr., George M. Weaver, William B. Hollberg, Wendell R. Bird, John W Whitehead, Thomas O. Kotouc, and Alfred Lindh; and for David Robinson, Jr., pro se.

[1] Georgia Code Ann. § 16-6-2 (1984) provides, in pertinent part, as follows:

”(a) A person commits the offense of sodomy when he performs or submits to any sexual act involving the sex organs of one person and the mouth or anus of another. . . .

”(b) A person convicted of the offense of sodomy shall be punished by imprisonment for not less than one nor more than 20 years. . . .”

[2] John and Mary Doe were also plaintiffs in the action. They alleged that they wished to engage in sexual activity proscribed by § 16-6-2 in the privacy of their home, App. 3, and that they had been “chilled and deterred” from engaging in such activity by both the existence of the statute and Hardwick’s arrest. Id., at 5. The District Court held, however, that because they had neither sustained, nor were in immediate danger of sustaining, any direct injury from the enforcement of the statute, they did not have proper standing to maintain the action. Id., at 18. The Court of Appeals affirmed the District Court’s judgment dismissing the Does’ claim for lack of standing, 760 F. 2d 1202, 1206-1207 (CA11 1985), and the Does do not challenge that holding in this Court.

The only claim properly before the Court, therefore, is Hardwick’s challenge to the Georgia statute as applied to consensual homosexual sodomy. We express no opinion on the constitutionality of the Georgia statute as applied to other acts of sodomy.

[3] See Baker v. Wade, 769 F. 2d 289, rehearing denied, 774 F. 2d 1285 (CA5 1985) (en banc); Dronenburg v. Zech, 239 U. S. App. D. C. 229, 741 F. 2d 1388, rehearing denied, 241 U. S. App. D. C. 262, 746 F. 2d 1579 (1984).

[4] Petitioner also submits that the Court of Appeals erred in holding that the District Court was not obligated to follow our summary affirmance in Doe. We need not resolve this dispute, for we prefer to give plenary consideration to the merits of this case rather than rely on our earlier action in Doe. See Usery v. Turner Elkhorn Mining Co., 428 U. S. 1, 14 (1976); Massachusetts Board of Retirement v. Murgia, 427 U. S. 307, 309, n. 1 (1976); Edelman v. Jordan, 415 U. S. 651, 671 (1974). Cf. Hicks v. Miranda, 422 U. S. 332, 344 (1975).

[5] Criminal sodomy laws in effect in 1791:

Connecticut: 1 Public Statute Laws of the State of Connecticut, 1808, Title LXVI, ch. 1, § 2 (rev. 1672).

Delaware: 1 Laws of the State of Delaware, 1797, ch. 22, § 5 (passed 1719).

Georgia had no criminal sodomy statute until 1816, but sodomy was a crime at common law, and the General Assembly adopted the common law of England as the law of Georgia in 1784. The First Laws of the State of Georgia, pt. 1, p. 290 (1981).

Maryland had no criminal sodomy statute in 1791. Maryland’s Declaration of Rights, passed in 1776, however, stated that “the inhabitants of Maryland are entitled to the common law of England,” and sodomy was a crime at common law. 4 W. Swindler, Sources and Documents of United States Constitutions 372 (1975).

Massachusetts: Acts and Laws passed by the General Court of Massachusetts, ch. 14, Act of Mar. 3, 1785.

New Hampshire passed its first sodomy statute in 1718. Acts and Laws of New Hampshire 1680-1726, p. 141 (1978).

Sodomy was a crime at common law in New Jersey at the time of the ratification of the Bill of Rights. The State enacted its first criminal sodomy law five years later. Acts of the Twentieth General Assembly, Mar. 18, 1796, ch. DC, § 7.

New York: Laws of New York, ch. 21 (passed 1787).

At the time of ratification of the Bill of Rights, North Carolina had adopted the English statute of Henry VIII outlawing sodomy. See Collection of the Statutes of the Parliament of England in Force in the State of North-Carolina, ch. 17, p. 314 (Martin ed. 1792).

Pennsylvania: Laws of the Fourteenth General Assembly of the Commonwealth of Pennsylvania, ch. CLIV, § 2 (passed 1790).

Rhode Island passed its first sodomy law in 1662. The Earliest Acts and Laws of the Colony of Rhode Island and Providence Plantations 1647-1719, p. 142 (1977).

South Carolina: Public Laws of the State of South Carolina, p. 49 (1790).

At the time of the ratification of the Bill of Rights, Virginia had no specific statute outlawing sodomy, but had adopted the English common law. 9 Hening’s Laws of Virginia, ch. 5, § 6, p. 127 (1821) (passed 1776).

[6] Criminal sodomy statutes in effect in 1868:

Alabama: Ala. Rev. Code § 3604 (1867).

Arizona (Terr.): Howell Code, ch. 10, § 48 (1865).

Arkansas: Ark. Stat., ch. 51, Art. IV, § 5 (1858).

California: 1 Cal. Gen. Laws, ¶ 1450, § 48 (1865).

Colorado (Terr.): Colo. Rev. Stat., ch. 22, §§ 45, 46 (1868).

Connecticut: Conn. Gen. Stat., Tit. 122, ch. 7, § 124 (1866).

Delaware: Del. Rev. Stat., ch. 131, § 7 (1893).

Florida: Fla. Rev. Stat., div. 5, § 2614 (passed 1868) (1892).

Georgia: Ga. Code §§ 4286, 4287, 4290 (1867).

Kingdom of Hawaii: Haw. Penal Code, ch. 13, § 11 (1869).

Illinois: Ill. Rev. Stat., div. 5, §§ 49, 50 (1845).

Kansas (Terr.): Kan. Stat., ch. 53, § 7 (1855).

Kentucky: 1 Ky. Rev. Stat., ch. 28, Art. IV, § 11 (1860).

Louisiana: La. Rev. Stat., Crimes and Offences, § 5 (1856).

Maine: Me. Rev. Stat., Tit. XII, ch. 160, § 4 (1840).

Maryland: 1 Md. Code, Art. 30, § 201 (1860).

Massachusetts: Mass. Gen. Stat., ch. 165, § 18 (1860).

Michigan: Mich. Rev. Stat., Tit. 30, ch. 158, § 16 (1846).

Minnesota: Minn. Stat., ch. 96, § 13 (1859).

Mississippi: Miss. Rev. Code, ch. 64, § LII, Art. 238 (1857).

Missouri: 1 Mo. Rev. Stat., ch. 50, Art. VIII, § 7 (1856).

Montana (Terr.): Mont. Acts, Resolutions, Memorials, Criminal Practice Acts, ch. IV, § 44 (1866).

Nebraska (Terr.): Neb. Rev. Stat., Crim. Code, ch. 4, § 47 (1866).

Nevada (Terr.): Nev. Comp. Laws, 1861-1900, Crimes and Punishments, § 45.

New Hampshire: N. H. Laws, Act. of June 19, 1812, § 5 (1815).

New Jersey: N. J. Rev. Stat., Tit. 8, ch. 1, § 9 (1847).

New York: 3 N. Y. Rev. Stat., pt. 4, ch. 1, Tit. 5, § 20 (5th ed. 1859).

North Carolina: N. C. Rev. Code, ch. 34, § 6 (1855).

Oregon: Laws of Ore., Crimes — Against Morality, etc., ch. 7, § 655 (1874).

Pennsylvania: Act of Mar. 31, 1860, § 32, Pub. L. 392, in 1 Digest of Statute Law of Pa. 1700-1903, p. 1011 (Purdon 1905).

Rhode Island: R. I. Gen. Stat., ch. 232, § 12 (1872).

South Carolina: Act of 1712, in 2 Stat. at Large of S. C. 1682-1716, p. 493 (1837).

Tennessee: Tenn. Code, ch. 8, Art. 1, § 4843 (1858).

Texas: Tex. Rev. Stat., Tit. 10, ch. 5, Art. 342 (1887) (passed 1860).

Vermont: Acts and Laws of the State of Vt. (1779).

Virginia: Va. Code, ch. 149, § 12 (1868).

West Virginia: W. Va. Code, ch. 149, § 12 (1868).

Wisconsin (Terr.): Wis. Stat. § 14, p. 367 (1839).

[7] In 1961, Illinois adopted the American Law Institute’s Model Penal Code, which decriminalized adult, consensual, private, sexual conduct. Criminal Code of 1961, §§ 11-2, 11-3, 1961 Ill. Laws, pp. 1985, 2006 (codified as amended at Ill. Rev. Stat., ch. 38, ¶¶ 11-2, 11-3 (1983) (repealed 1984)). See American Law Institute, Model Penal Code § 213.2 (Proposed Official Draft 1962).

[8] Respondent does not defend the judgment below based on the Ninth Amendment, the Equal Protection Clause, or the Eighth Amendment.

[1] Among those States that continue to make sodomy a crime, Georgia authorizes one of the longest possible sentences. See Ala. Code § 13A-6-65(a)(3) (1982) (1-year maximum); Ariz. Rev. Stat. Ann. §§ 13-1411, 13-1412 (West Supp. 1985) (30 days); Ark. Stat. Ann. § 41-1813 (1977) (1-year maximum); D. C. Code § 22-3502 (1981) (10-year maximum); Fla. Stat. § 800.02 (1985) (60-day maximum); Ga. Code Ann. § 16-6-2 (1984) (1 to 20 years); Idaho Code § 18-6605 (1979) (5-year minimum); Kan. Stat. Ann. § 21-3505 (Supp. 1985) (6-month maximum); Ky. Rev. Stat. § 510.100 (1985) (90 days to 12 months); La. Rev. Stat. Ann. § 14:89 (West 1986) (5-year maximum); Md. Ann. Code, Art. 27, §§ 553-554 (1982) (10-year maximum); Mich. Comp. Laws § 750.158 (1968) (15-year maximum); Minn. Stat. § 609.293 (1984) (1-year maximum); Miss. Code Ann. § 97-29-59 (1973) (10-year maximum); Mo. Rev. Stat. § 566.090 (Supp. 1984) (1-year maximum); Mont. Code Ann. § 45-5-505 (1985) (10-year maximum); Nev. Rev. Stat. § 201.190 (1985) (6-year maximum); N. C. Gen. Stat. § 14-177 (1981) (10-year maximum); Okla. Stat., Tit. 21, § 886 (1981) (10-year maximum); R. I. Gen. Laws § 11-10-1 (1981) (7 to 20 years); S. C. Code § 16-15-120 (1985) (5-year maximum); Tenn. Code Ann. § 39-2-612 (1982) (5 to 15 years); Tex. Penal Code Ann. § 21.06 (1974) ($200 maximum fine); Utah Code Ann. § 76-5-403 (1978) (6-month maximum); Va. Code § 18.2-361 (1982) (5-year maximum).

[2] It was conceded at oral argument that, prior to the complaint against respondent Hardwick, there had been no reported decision involving prosecution for private homosexual sodomy under this statute for several decades. See Thompson v. Aldredge, 187 Ga. 467, 200 S. E. 799 (1939). Moreover, the State has declined to present the criminal charge against Hardwick to a grand jury, and this is a suit for declaratory judgment brought by respondents challenging the validity of the statute. The history of nonenforcement suggests the moribund character today of laws criminalizing this type of private, consensual conduct. Some 26 States have repealed similar statutes. But the constitutional validity of the Georgia statute was put in issue by respondents, and for the reasons stated by the Court, I cannot say that conduct condemned for hundreds of years has now become a fundamental right.

[1] Until 1968, Georgia defined sodomy as “the carnal knowledge and connection against the order of nature, by man with man, or in the same unnatural manner with woman.” Ga. Crim. Code § 26-5901 (1933). In Thompson v. Aldredge, 187 Ga. 467, 200 S. E. 799 (1939), the Georgia Supreme Court held that § 26-5901 did not prohibit lesbian activity. And in Riley v. Garrett, 219 Ga. 345, 133 S. E. 2d 367 (1963), the Georgia Supreme Court held that § 26-5901 did not prohibit heterosexual cunnilingus. Georgia passed the act-specific statute currently in force “perhaps in response to the restrictive court decisions such as Riley,” Note, The Crimes Against Nature, 16 J. Pub. L. 159, 167, n. 47 (1967).

[2] In Robinson v. California, 370 U. S. 660 (1962), the Court held that the Eighth Amendment barred convicting a defendant due to his “status” as a narcotics addict, since that condition was “apparently an illness which may be contracted innocently or involuntarily.” Id., at 667. In Powell v. Texas, 392 U. S. 514 (1968), where the Court refused to extend Robinson to punishment of public drunkenness by a chronic alcoholic, one of the factors relied on by JUSTICE MARSHALL, in writing the plurality opinion, was that Texas had not “attempted to regulate appellant’s behavior in the privacy of his own home.” Id., at 532. JUSTICE WHITE wrote separately:

“Analysis of this difficult case is not advanced by preoccupation with the label `condition.’ In Robinson the Court dealt with `a statute which makes the “status” of narcotic addiction a criminal offense . . . .’ 370 U. S., at 666. By precluding criminal conviction for such a `status’ the Court was dealing with a condition brought about by acts remote in time from the application of the criminal sanctions contemplated, a condition which was relatively permanent in duration, and a condition of great magnitude and significance in terms of human behavior and values. . . . If it were necessary to distinguish between `acts’ and `conditions’ for purposes of the Eighth Amendment, I would adhere to the concept of `condition’ implicit in the opinion in Robinson . . . . The proper subject of inquiry is whether volitional acts brought about the `condition’ and whether those acts are sufficiently proximate to the `condition’ for it to be permissible to impose penal sanctions on the `condition.’ ” Id., at 550-551, n. 2.

Despite historical views of homosexuality, it is no longer viewed by mental health professionals as a “disease” or disorder. See Brief for American Psychological Association and American Public Health Association as Amici Curiae 8-11. But, obviously, neither is it simply a matter of deliberate personal election. Homosexual orientation may well form part of the very fiber of an individual’s personality. Consequently, under JUSTICE WHITE’s analysis in Powell, the Eighth Amendment may pose a constitutional barrier to sending an individual to prison for acting on that attraction regardless of the circumstances. An individual’s ability to make constitutionally protected “decisions concerning sexual relations,” Carey v. Population Services International, 431 U. S. 678, 711 (1977) (POWELL, J., concurring in part and concurring in judgment), is rendered empty indeed if he or she is given no real choice but a life without any physical intimacy.

With respect to the Equal Protection Clause’s applicability to § 16-6-2, I note that Georgia’s exclusive stress before this Court on its interest in prosecuting homosexual activity despite the gender-neutral terms of the statute may raise serious questions of discriminatory enforcement, questions that cannot be disposed of before this Court on a motion to dismiss. See Yick Wo v. Hopkins, 118 U. S. 356, 373-374 (1886). The legislature having decided that the sex of the participants is irrelevant to the legality of the acts, I do not see why the State can defend § 16-6-2 on the ground that individuals singled out for prosecution are of the same sex as their partners. Thus, under the circumstances of this case, a claim under the Equal Protection Clause may well be available without having to reach the more controversial question whether homosexuals are a suspect class. See, e. g., Rowland v. Mad River Local School District, 470 U. S. 1009 (1985) (BRENNAN, J., dissenting from denial of certiorari); Note, The Constitutional Status of Sexual Orientation: Homosexuality as a Suspect Classification, 98 Harv. L. Rev. 1285 (1985).

[3] Even if a court faced with a challenge to § 16-6-2 were to apply simple rational-basis scrutiny to the statute, Georgia would be required to show an actual connection between the forbidden acts and the ill effects it seeks to prevent. The connection between the acts prohibited by § 16-6-2 and the harms identified by petitioner in his brief before this Court is a subject of hot dispute, hardly amenable to dismissal under Federal Rule of Civil Procedure 12(b)(6). Compare, e. g., Brief for Petitioner 36-37 and Brief for David Robinson, Jr., as Amicus Curiae 23-28, on the one hand, with People v. Onofre, 51 N. Y. 2d 476, 489, 415 N. E. 2d 936, 941 (1980); Brief for the Attorney General of the State of New York, joined by the Attorney General of the State of California, as Amici Curiae 11-14; and Brief for the American Psychological Association and American Public Health Association as Amici Curiae 19-27, on the other.

[4] Although I do not think it necessary to decide today issues that are not even remotely before us, it does seem to me that a court could find simple, analytically sound distinctions between certain private, consensual sexual conduct, on the one hand, and adultery and incest (the only two vaguely specific “sexual crimes” to which the majority points, ante, at 196), on the other. For example, marriage, in addition to its spiritual aspects, is a civil contract that entitles the contracting parties to a variety of governmentally provided benefits. A State might define the contractual commitment necessary to become eligible for these benefits to include a commitment of fidelity and then punish individuals for breaching that contract. Moreover, a State might conclude that adultery is likely to injure third persons, in particular, spouses and children of persons who engage in extramarital affairs. With respect to incest, a court might well agree with respondent that the nature of familial relationships renders true consent to incestuous activity sufficiently problematical that a blanket prohibition of such activity is warranted. See Tr. of Oral Arg. 21-22. Notably, the Court makes no effort to explain why it has chosen to group private, consensual homosexual activity with adultery and incest rather than with private, consensual heterosexual activity by unmarried persons or, indeed, with oral or anal sex within marriage.

[5] The parallel between Loving and this case is almost uncanny. There, too, the State relied on a religious justification for its law. Compare 388 U. S., at 3 (quoting trial court’s statement that “Almighty God created the races white, black, yellow, malay and red, and he placed them on separate continents. . . . The fact that he separated the races shows that he did not intend for the races to mix”), with Brief for Petitioner 20-21 (relying on the Old and New Testaments and the writings of St. Thomas Aquinas to show that “traditional Judeo-Christian values proscribe such conduct”). There, too, defenders of the challenged statute relied heavily on the fact that when the Fourteenth Amendment was ratified, most of the States had similar prohibitions. Compare Brief for Appellee in Loving v. Virginia, O. T. 1966, No. 395, pp. 28-29, with ante, at 192-194, and n. 6. There, too, at the time the case came before the Court, many of the States still had criminal statutes concerning the conduct at issue. Compare 388 U. S., at 6, n. 5 (noting that 16 States still outlawed interracial marriage), with ante, at 193-194 (noting that 24 States and the District of Columbia have sodomy statutes). Yet the Court held, not only that the invidious racism of Virginia’s law violated the Equal Protection Clause, see 388 U. S., at 7-12, but also that the law deprived the Lovings of due process by denying them the “freedom of choice to marry” that had “long been recognized as one of the vital personal rights essential to the orderly pursuit of happiness by free men.” Id., at 12.

[6] The theological nature of the origin of Anglo-American antisodomy statutes is patent. It was not until 1533 that sodomy was made a secular offense in England. 25 Hen. VIII, ch. 6. Until that time, the offense was, in Sir James Stephen’s words, “merely ecclesiastical.” 2J. Stephen, A History of the Criminal Law of England 429-430 (1883). Pollock and Maitland similarly observed that ”[t]he crime against nature . . . was so closely connected with heresy that the vulgar had but one name for both.” 2 F. Pollock & F. Maitland, The History of English Law 554 (1895). The transfer of jurisdiction over prosecutions for sodomy to the secular courts seems primarily due to the alteration of ecclesiastical jurisdiction attendant on England’s break with the Roman Catholic Church, rather than to any new understanding of the sovereign’s interest in preventing or punishing the behavior involved. Cf. 6 E. Coke, Institutes, ch. 10 (4th ed. 1797).

[7] At oral argument a suggestion appeared that, while the Fourth Amendment’s special protection of the home might prevent the State from enforcing § 16-6-2 against individuals who engage in consensual sexual activity there, that protection would not make the statute invalid. See Tr. of Oral Arg. 10-11. The suggestion misses the point entirely. If the law is not invalid, then the police can invade the home to enforce it, provided, of course, that they obtain a determination of probable cause from a neutral magistrate. One of the reasons for the Court’s holding in Griswold v. Connecticut, 381 U. S. 479 (1965), was precisely the possibility, and repugnance, of permitting searches to obtain evidence regarding the use of contraceptives. Id., at 485-486. Permitting the kinds of searches that might be necessary to obtain evidence of the sexual activity banned by § 16-6-2 seems no less intrusive, or repugnant. Cf. Winston v. Lee, 470 U. S. 753 (1985); Mary Beth G. v. City of Chicago, 723 F. 2d 1263, 1274 (CA7 1983).

[1] See Ga. Code Ann. § 16-6-2(a) (1984) (“A person commits the offense of sodomy when he performs or submits to any sexual act involving the sex organs of one person and the mouth or anus of another”).

[2] The Court states that the “issue presented is whether the Federal Constitution confers a fundamental right upon homosexuals to engage in sodomy and hence invalidates the laws of the many States that still make such conduct illegal and have done so for a very long time.” Ante, at 190. In reality, however, it is the indiscriminate prohibition of sodomy, heterosexual as well as homosexual, that has been present “for a very long time.” See nn. 3, 4, and 5, infra. Moreover, the reasoning the Court employs would provide the same support for the statute as it is written as it does for the statute as it is narrowly construed by the Court.

[3] See, e. g., 1 W. Hawkins, Pleas of the Crown 9 (6th ed. 1787) (“All unnatural carnal copulations, whether with man or beast, seem to come under the notion of sodomy, which was felony by the antient common law, and punished, according to some authors, with burning; according to others, with burying alive”); 4 W. Blackstone, Commentaries *215 (discussing “the infamous crime against nature, committed either with man or beast; a crime which ought to be strictly and impartially proved, and then as strictly and impartially punished”).

[4] See 1 E. East, Pleas of the Crown 480 (1803) (“This offence, concerning which the least notice is the best, consists in a carnal knowledge committed against the order of nature by man with man, or in the same unnatural manner with woman, or by man or woman in any manner with beast”); J. Hawley & M. McGregor, The Criminal Law 287 (3d ed. 1899) (“Sodomy is the carnal knowledge against the order of nature by two persons with each other, or of a human being with a beast. . . . The offense may be committed between a man and a woman, or between two male persons, or between a man or a woman and a beast”).

[5] See J. May, The Law of Crimes § 203 (2d ed. 1893) (“Sodomy, otherwise called buggery, bestiality, and the crime against nature, is the unnatural copulation of two persons with each other, or of a human being with a beast. . . . It may be committed by a man with a man, by a man with a beast, or by a woman with a beast, or by a man with a woman — his wife, in which case, if she consent, she is an accomplice”).

[6] The predecessor of the current Georgia statute provided: “Sodomy is the carnal knowledge and connection against the order of nature, by man with man, or in the same unnatural manner with woman.” Ga. Code, Tit. 1, Pt. 4, § 4251 (1861). This prohibition of heterosexual sodomy was not purely hortatory. See, e. g., Comer v. State, 21 Ga. App. 306, 94 S. E. 314 (1917) (affirming prosecution for consensual heterosexual sodomy).

[7] See Thompson v. Aldredge, 187 Ga. 467, 200 S. E. 799 (1939).

[8] A review of the statutes cited by the majority discloses that, in 1791, in 1868, and today, the vast majority of sodomy statutes do not differentiate between homosexual and heterosexual sodomy.

[9] See Loving v. Virginia, 388 U. S. 1 (1967). Interestingly, miscegenation was once treated as a crime similar to sodomy. See Hawley & McGregor, The Criminal Law, at 287 (discussing crime of sodomy); id., at 288 (discussing crime of miscegenation).

[10] Indeed, the Georgia Attorney General concedes that Georgia’s statute would be unconstitutional if applied to a married couple. See Tr. of Oral Arg. 8 (stating that application of the statute to a married couple “would be unconstitutional” because of the “right of marital privacy as identified by the Court in Griswold”). Significantly, Georgia passed the current statute three years after the Court’s decision in Griswold.

[11] Ante, at 198, n. 2 (POWELL, J., concurring). See also Tr. of Oral Arg. 4-5 (argument of Georgia Attorney General) (noting, in response to question about prosecution “where the activity took place in a private residence,” the “last case I can recall was back in the 1930’s or 40’s”).

[12] It is, of course, possible to argue that a statute has a purely symbolic role. Cf. Carey v. Population Services International, 431 U. S. 678, 715, n. 3 (1977) (STEVENS, J., concurring in part and concurring in judgment) (“The fact that the State admittedly has never brought a prosecution under the statute . . . is consistent with appellants’ position that the purpose of the statute is merely symbolic”). Since the Georgia Attorney General does not even defend the statute as written, however, see n. 10, supra, the State cannot possibly rest on the notion that the statute may be defended for its symbolic message.

[13] Indeed, at this stage, it appears that the statute indiscriminately authorizes a policy of selective prosecution that is neither limited to the class of homosexual persons nor embraces all persons in that class, but rather applies to those who may be arbitrarily selected by the prosecutor for reasons that are not revealed either in the record of this case or in the text of the statute. If that is true, although the text of the statute is clear enough, its true meaning may be “so intolerably vague that evenhanded enforcement of the law is a virtual impossibility.” Marks v. United States, 430 U. S. 188, 198 (1977) (STEVENS, J., concurring in part and dissenting in part).

          After various Colorado municipalities passed ordinances banning discrimination based on sexual orientation in housing, employment, education, public accommodations, health and welfare services, and other transactions and activities, Colorado voters adopted by statewide referendum "Amendment 2" to the State Constitution, which precludes all legislative, executive, or judicial action at any level of state or local government designed to protect the status of persons based on their ``homosexual, lesbian or bisexual orientation, conduct, practices or relationships.'' Respondents, who include aggrieved homosexuals and municipalities, commenced this litigation in state court against petitioner state parties to declare Amendment 2 invalid and enjoin its enforcement. The trial court's grant of a preliminary injunction was sustained by the Colorado Supreme Court, which held that Amendment 2 was subject to strict scrutiny under the Equal Protection Clause of the Fourteenth Amendment because it infringed the fundamental right of gays and lesbians to participate in the political process. On remand, the trial court found that the Amendment failed to satisfy strict scrutiny. It enjoined Amendment 2's enforcement, and the State Supreme Court affirmed.

          Held: Amendment 2 violates the Equal Protection Clause. Pp. 4-14.

            (a) The State's principal argument that Amendment 2 puts gays and lesbians in the same position as all other persons by denying them special rights is rejected as implausible. The extent of the change in legal status effected by this law is evident from the authoritative construction of Colorado's Supreme Court-which establishes that the amendment's immediate effect is to repeal all existing statutes, regulations, ordinances, and policies of state and local entities barring discrimination based on sexual orientation, and that its ultimate effect is to prohibit any governmental entity from adopting similar, or more protective, measures in the future absent state constitutional amendment-and from a review of the terms, structure, and operation of the ordinances that would be repealsed and prohibited by Amendment 2. Even if, as the State contends, homosexuals can find protection in laws and policies of general application, Amendment 2 goes well beyond merely depriving them of special rights. It imposes a broad disability upon those persons alone, forbidding them, but no others, to seek specific legal protection from injuries caused by discrimination in a wide range of public and private transactions. Pp. 4-9.

            (b) In order to reconcile the Fourteenth Amendment's promise that no person shall be denied equal protection with the practical reality that most legislation classifies for one purpose or another, the Court has stated that it will uphold a law that neither burdens a fundamental right nor targets a suspect class so long as the legislative classification bears a rational relation to some independent and legitimate legislative end. See, e.g., Heller v. Doe, 509 U. S. 312, 319-320. Amendment 2 fails, indeed defies, even this conventional inquiry. First, the amendment is at once too narrow and too broad, identifying persons by a single trait and then denying them the possibility of protection across the board. This disqualification of a class of persons from the right to obtain specific protection from the law is unprecedented and is itself a denial of equal protection in the most literal sense. Second, the sheer breadth of Amendment 2, which makes a general announcement that gays and lesbians shall not have any particular protections from the law, is so far removed from the reasons offered for it, i.e., respect for other citizens' freedom of association, particularly landlords or employers who have personal or religious objections to homosexuality, and the State's interest in conserving resources to fight discrimination against other groups, that the amendment cannot be explained by reference to those reasons; the Amendment raises the inevitable inference that it is born of animosity toward the class that it affects. Amendment 2 cannot be said to be directed to an identifiable legitimate purpose or discrete objective. It is a status-based classification of persons undertaken for its own sake, something the Equal Protection Clause does not permit. Pp. 9-14.

882 P. 2d 1335, affirmed.

          Kennedy, J., delivered the opinion of the Court, in which Stevens, O'Connor, Souter, Ginsburg, and Breyer, JJ., joined. Scalia, J., filed a dissenting opinion, in which Rehnquist, C. J., and Thomas, J., joined.

          Justice Kennedy delivered the opinion of the Court.

          One century ago, the first Justice Harlan admonished this Court that the Constitution "neither knows nor tolerates classes among citizens." Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (dissenting opinion). Unheeded then, those words now are understood to state a commitment to the law's neutrality where the rights of persons are at stake. The Equal Protection Clause enforces this principle and today requires us to hold invalid a provision of Colorado's Constitution.

          The enactment challenged in this case is an amendment to the Constitution of the State of Colorado, adopted in a 1992 statewide referendum. The parties and the state courts refer to it as "Amendment 2," its designation when submitted to the voters. The impetus for the amendment and the contentious campaign that preceded its adoption came in large part from ordinances that had been passed in various Colorado municipalities. For example, the cities of Aspen and Boulder and the City and County of Denver each had enacted ordinances which banned discrimination in many transactions and activities, including housing, employment, education, public accommodations, and health and welfare services. Denver Rev. Municipal Code, Art. IV Section(s) 28-91 to 28-116 (1991); Aspen Municipal Code Section(s) 13-98 (1977); Boulder Rev. Code Section(s) 12-1-1 to 12-1-11 (1987). What gave rise to the statewide controversy was the protection the ordinances afforded to persons discriminated against by reason of their sexual orientation. See Boulder Rev. Code Section(s) 12-1-1 (defining "sexual orientation" as "the choice of sexual partners, i.e., bisexual, homosexual or heterosexual"); Denver Rev. Municipal Code, Art. IV Section(s) 28-92 (defining "sexual orientation" as "[t]he status of an individual as to his or her heterosexuality, homosexuality or bisexuality"). Amendment 2 repeals these ordinances to the extent they prohibit discrimination on the basis of "homosexual, lesbian or bisexual orientation, conduct, practices or relationships." Colo. Const., Art. II, Section(s) 30b.

          Yet Amendment 2, in explicit terms, does more than repeal or rescind these provisions. It prohibits all legislative, executive or judicial action at any level of state or local government designed to protect the named class, a class we shall refer to as homosexual persons or gays and lesbians. The amendment reads:

                    "No Protected Status Based on Homosexual, Lesbian, or Bisexual Orientation. 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 Section of the Constitution shall be in all respects self-executing." Ibid.

          Soon after Amendment 2 was adopted, this litigation to declare its invalidity and enjoin its enforcement was commenced in the District Court for the City and County of Denver. Among the plaintiffs (respondents here) were homosexual persons, some of them government employees. They alleged that enforcement of Amendment 2 would subject them to immediate and substantial risk of discrimination on the basis of their sexual orientation. Other plaintiffs (also respondents here) included the three municipalities whose ordinances we have cited and certain other governmental entities which had acted earlier to protect homosexuals from discrimination but would be prevented by Amendment 2 from continuing to do so. Although Governor Romer had been on record opposing the adoption of Amendment 2, he was named in his official capacity as a defendant, together with the Colorado Attorney General and the State of Colorado.

          The trial court granted a preliminary injunction to stay enforcement of Amendment 2, and an appeal was taken to the Supreme Court of Colorado. Sustaining the interim injunction and remanding the case for further proceedings, the State Supreme Court held that Amendment 2 was subject to strict scrutiny under the Fourteenth Amendment because it infringed the fundamental right of gays and lesbians to participate in the political process. Evans v. Romer, 854 P. 2d 1270 (Colo. 1993) (Evans I). To reach this conclusion, the state court relied on our voting rights cases, e.g., Reynolds v. Sims, 377 U. S. 533 (1964); Carrington v. Rash, 380 U. S. 89 (1965); Harper v. Virginia Bd. of Elections, 383 U. S. 663 (1966); Williams v. Rhodes, 393 U. S. 23 (1968), and on our precedents involving discriminatory restructuring of governmental decisionmaking, see, e.g., Hunter v. Erickson, 393 U. S. 385 (1969); Reitman v. Mulkey, 387 U. S. 369 (1967); Washington v. Seattle School Dist. No. 1, 458 U. S. 457 (1982); Gordon v. Lance, 403 U. S. 1 (1971). On remand, the State advanced various arguments in an effort to show that Amendment 2 was narrowly tailored to serve compelling interests, but the trial court found none sufficient. It enjoined enforcement of Amendment 2, and the Supreme Court of Colorado, in a second opinion, affirmed the ruling. Evans v. Romer, 882 P. 2d 1335 (Colo. 1994) (Evans II). We granted certiorari and now affirm the judgment, but on a rationale different from that adopted by the State Supreme Court.

          The State's principal argument in defense of Amendment 2 is that it puts gays and lesbians in the same position as all other persons. So, the State says, the measure does no more than deny homosexuals special rights. This reading of the amendment's language is implausible. We rely not upon our own interpretation of the amendment but upon the authoritative construction of Colorado's Supreme Court. The state court, deeming it unnecessary to determine the full extent of the amendment's reach, found it invalid even on a modest reading of its implications. The critical discussion of the amendment, set out in Evans I, is as follows:

                    "The immediate objective of Amendment 2 is, at a minimum, to repeal existing statutes, regulations, ordinances, and policies of state and local entities that barred discrimination based on sexual orientation. See Aspen, Colo., Mun. Code Section(s) 13-98 (1977) (prohibiting discrimination in employment, housing and public accommodations on the basis of sexual orientation); Boulder, Colo., Rev. Code Section(s) 12-1-2 to -4 (1987) (same); Denver, Colo., Rev. Mun. Code art. IV, Section(s) 28-91 to -116 (1991) (same); Executive Order No. D0035 (December 10, 1990) (prohibiting employment discrimination for `all state employees, classified and exempt' on the basis of sexual orientation); Colorado Insurance Code, Section(s) 10-3-1104, 4A C. R. S. (1992 Supp.) (forbidding health insurance providers from determining insurability and premiums based on an applicant's, a beneficiary's, or an insured's sexual orientation); and various provisions prohibiting discrimination based on sexual orientation at state colleges.26

                    "26 Metropolitan State College of Denver prohibits college sponsored social clubs from discriminating in membership on the basis of sexual orientation and Colorado State University has an antidiscrimination policy which encompasses sexual orientation.

                    "The `ultimate effect' of Amendment 2 is to prohibit any governmental entity from adopting similar, or more protective statutes, regulations, ordinances, or policies in the future unless the state constitution is first amended to permit such measures." 854 P. 2d, at 1284-1285, and n. 26.

          Sweeping and comprehensive is the change in legal status effected by this law. So much is evident from the ordinances that the Colorado Supreme Court declared would be void by operation of Amendment 2. Homosexuals, by state decree, are put in a solitary class with respect to transactions and relations in both the private and governmental spheres. The amendment withdraws from homosexuals, but no others, specific legal protection from the injuries caused by discrimination, and it forbids reinstatement of these laws and policies.

          The change that Amendment 2 works in the legal status of gays and lesbians in the private sphere is far-reaching, both on its own terms and when considered in light of the structure and operation of modern anti-discrimination laws. That structure is well illustrated by contemporary statutes and ordinances prohibiting discrimination by providers of public accommodations. "At common law, innkeepers, smiths, and others who `made profession of a public employment,' were prohibited from refusing, without good reason, to serve a customer." Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U. S. ___, ___ (1995) (slip op., at 13). The duty was a general one and did not specify protection for particular groups. The common law rules, however, proved insufficient in many instances, and it was settled early that the Fourteenth Amendment did not give Congress a general power to prohibit discrimination in public accommodations, Civil Rights Cases, 109 U. S. 3, 25 (1883). In consequence, most States have chosen to counter discrimination by enacting detailed statutory schemes. See, e.g., S. D. Codified Laws Section(s) 20-13-10, 20-13-22, 20-13-23 (1995); Iowa Code Section(s) 216.6-216.8 (1994); Okla. Stat., Tit. 25, Section(s) 1302, 1402 (1987); 43 Pa. Cons. Stat. Section(s) 953, 955 (Supp. 1995); N. J. Stat. Ann. Section(s) 10:5-3, 10:5-4 (West Supp. 1995); N. H. Rev. Stat. Ann. Section(s) 354-A:7, 354-A:10, 354-A:17 (1995); Minn. Stat. Section(s) 363.03 (1991 and Supp. 1995).

          Colorado's state and municipal laws typify this emerging tradition of statutory protection and follow a consistent pattern. The laws first enumerate the persons or entities subject to a duty not to discriminate. The list goes well beyond the entities covered by the common law. The Boulder ordinance, for example, has a comprehensive definition of entities deemed places of "public accommodation." They include "any place of business engaged in any sales to the general public and any place that offers services, facilities, privileges, or advantages to the general public or that receives financial support through solicitation of the general public or through governmental subsidy of any kind." Boulder Rev. Code Section(s) 12-1-1(j) (1987). The Denver ordinance is of similar breadth, applying, for example, to hotels, restaurants, hospitals, dental clinics, theaters, banks, common carriers, travel and insurance agencies, and "shops and stores dealing with goods or services of any kind," Denver Rev. Municipal Code, Art. IV, Section(s) 28-92.

          These statutes and ordinances also depart from the common law by enumerating the groups or persons within their ambit of protection. Enumeration is the essential device used to make the duty not to discriminate concrete and to provide guidance for those who must comply. In following this approach, Colorado's state and local governments have not limited anti-discrimination laws to groups that have so far been given the protection of heightened equal protection scrutiny under our cases. See, e.g., J. E. B. v. Alabama ex rel. T. B., 511 U. S. __, __ (1994) (slip op., at 8) (sex); Lalli v. Lalli, 439 U. S. 259, 265 (1978) (illegitimacy); McLaughlin v. Florida, 379 U. S. 184, 191-192 (1964) (race); Oyama v. California, 332 U. S. 633 (1948) (ancestry). Rather, they set forth an extensive catalogue of traits which cannot be the basis for discrimination, including age, military status, marital status, pregnancy, parenthood, custody of a minor child, political affiliation, physical or mental disability of an individual or of his or her associates--and, in recent times, sexual orientation. Aspen Municipal Code Section(s) 13-98(a)(1) (1977); Boulder Rev. Code Section(s) 12-1-1 to 12-1-4 (1987); Denver Rev. Municipal Code, Art. IV, Section(s) 28-92 to 28-119 (1991); Colo. Rev. Stat. Section(s) 24-34-401 to 24-34-707 (1988 and Supp. 1995).

          Amendment 2 bars homosexuals from securing protection against the injuries that these public-accommodations laws address. That in itself is a severe consequence, but there is more. Amendment 2, in addition, nullifies specific legal protections for this targeted class in all transactions in housing, sale of real estate, insurance, health and welfare services, private education, and employment. See, e.g., Aspen Municipal Code Section(s) 13-98(b), (c) (1977); Boulder Rev. Code Section(s) 12-1-2, 12-1-3 (1987); Denver Rev. Municipal Code, Art. IV Section(s) 28-93 to 28-95, Section(s) 28-97 (1991).

          Not confined to the private sphere, Amendment 2 also operates to repeal and forbid all laws or policies providing specific protection for gays or lesbians from discrimination by every level of Colorado government. The State Supreme Court cited two examples of protections in the governmental sphere that are now rescinded and may not be reintroduced. The first is Colorado Executive Order D0035 (1990), which forbids employment discrimination against "`all state employees, classified and exempt' on the basis of sexual orientation." 854 P. 2d, at 1284. Also repealed, and now forbidden, are "various provisions prohibiting discrimination based on sexual orientation at state colleges." Id., at 1284, 1285. The repeal of these measures and the prohibition against their future reenactment demonstrates that Amendment 2 has the same force and effect in Colorado's governmental sector as it does elsewhere and that it applies to policies as well as ordinary legislation.

          Amendment 2's reach may not be limited to specific laws passed for the benefit of gays and lesbians. It is a fair, if not necessary, inference from the broad language of the amendment that it deprives gays and lesbians even of the protection of general laws and policies that prohibit arbitrary discrimination in governmental and private settings. See, e.g., Colo. Rev. Stat. Section(s) 24-4-106(7) (1988) (agency action subject to judicial review under arbitrary and capricious standard); Section(s) 18-8-405 (making it a criminal offense for a public servant knowingly, arbitrarily or capriciously to refrain from performing a duty imposed on him by law); Section(s) 10-3-1104(1)(f) (prohibiting "unfair discrimination" in insurance); 4 Colo. Code of Regulations 801-1, Policy 11-1 (1983) (prohibiting discrimination in state employment on grounds of specified traits or "other non-merit factor"). At some point in the systematic administration of these laws, an official must determine whether homosexuality is an arbitrary and thus forbidden basis for decision. Yet a decision to that effect would itself amount to a policy prohibiting discrimination on the basis of homosexuality, and so would appear to be no more valid under Amendment 2 than the specific prohibitions against discrimination the state court held invalid.

          If this consequence follows from Amendment 2, as its broad language suggests, it would compound the constitutional difficulties the law creates. The state court did not decide whether the amendment has this effect, however, and neither need we. In the course of rejecting the argument that Amendment 2 is intended to conserve resources to fight discrimination against suspect classes, the Colorado Supreme Court made the limited observation that the amendment is not intended to affect many anti-discrimination laws protecting non-suspect classes, Romer II, 882 P. 2d at 1346, n. 9. In our view that does not resolve the issue. In any event, even if, as we doubt, homosexuals could find some safe harbor in laws of general application, we cannot accept the view that Amendment 2's prohibition on specific legal protections does no more than deprive homosexuals of special rights. To the contrary, the amendment imposes a special disability upon those persons alone. Homosexuals are forbidden the safeguards that others enjoy or may seek without constraint. They can obtain specific protection against discrimination only by enlisting the citizenry of Colorado to amend the state constitution or perhaps, on the State's view, by trying to pass helpful laws of general applicability. This is so no matter how local or discrete the harm, no matter how public and widespread the injury. We find nothing special in the protections Amendment 2 withholds. These are protections taken for granted by most people either because they already have them or do not need them; these are protections against exclusion from an almost limitless number of transactions and endeavors that constitute ordinary civic life in a free society.

          The Fourteenth Amendment's promise that no person shall be denied the equal protection of the laws must co-exist with the practical necessity that most legislation classifies for one purpose or another, with resulting disadvantage to various groups or persons. Personnel Administrator of Mass. v. Feeney, 442 U. S. 256, 271- 272 (1979); F. S. Royster Guano Co. v. Virginia, 253 U. S. 412, 415 (1920). We have attempted to reconcile the principle with the reality by stating that, if a law neither burdens a fundamental right nor targets a suspect class, we will uphold the legislative classification so long as it bears a rational relation to some legitimate end. See, e.g., Heller v. Doe, 509 U. S. ___, ___ (1993) (slip op., at 6).

          Amendment 2 fails, indeed defies, even this conventional inquiry. First, the amendment has the peculiar property of imposing a broad and undifferentiated disability on a single named group, an exceptional and, as we shall explain, invalid form of legislation. Second, its 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.

          Taking the first point, even in the ordinary equal protection case calling for the most deferential of standards, we insist on knowing the relation between the classification adopted and the object to be attained. The search for the link between classification and objective gives substance to the Equal Protection Clause; it provides guidance and discipline for the legislature, which is entitled to know what sorts of laws it can pass; and it marks the limits of our own authority. In the ordinary case, 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. See New Orleans v. Dukes, 427 U. S. 297 (1976) (tourism benefits justified classification favoring pushcart vendors of certain longevity); Williamson v. Lee Optical of Okla., Inc., 348 U. S. 483 (1955) (assumed health concerns justified law favoring optometrists over opticians); Railway Express Agency, Inc. v. New York, 336 U. S. 106 (1949) (potential traffic hazards justified exemption of vehicles advertising the owner's products from general advertising ban); Kotch v. Board of River Port Pilot Comm'rs for Port of New Orleans, 330 U. S. 552 (1947) (licensing scheme that disfavored persons unrelated to current river boat pilots justified by possible efficiency and safety benefits of a closely knit pilotage system). The laws challenged in the cases just cited were narrow enough in scope and grounded in a sufficient factual context for us to ascertain that there existed some relation between the classification and the purpose it served. By requiring that the classification bear a rational relationship to an independent and legitimate legislative end, we ensure that classifications are not drawn for the purpose of disadvantaging the group burdened by the law. See United States Railroad Retirement Bd. v. Fritz, 449 U. S. 166, 181 (1980) (Stevens, J., concurring) ("If the adverse impact on the disfavored class is an apparent aim of the legislature, its impartiality would be suspect.").

          Amendment 2 confounds this normal process of judicial review. It is at once too narrow and too broad. It identifies persons by a single trait and then denies them protection across the board. The resulting disqualification of a class of persons from the right to seek specific protection from the law is unprecedented in our jurisprudence. The absence of precedent for Amendment 2 is itself instructive; "[d]iscriminations of an unusual character especially suggest careful consideration to determine whether they are obnoxious to the constitutional provision." Louisville Gas & Elec. Co. v. Coleman, 277 U. S. 32, 37-38 (1928).

          It is not within our constitutional tradition to enact laws of this sort. Central both to the idea of the rule of law and to our own Constitution's guarantee of equal protection is the principle that government and each of its parts remain open on impartial terms to all who seek its assistance. "`Equal protection of the laws is not achieved through indiscriminate imposition of inequalities.'" Sweatt v. Painter, 339 U. S. 629, 635 (1950) (quoting Shelley v. Kraemer, 334 U. S. 1, 22 (1948)). Respect for this principle explains why laws singling out a certain class of citizens for disfavored legal status or general hardships are rare. A law declaring that in general it shall be more difficult for one group of citizens than for all others to seek aid from the government is itself a denial of equal protection of the laws in the most literal sense. "The guaranty of `equal protection of the laws is a pledge of the protection of equallaws.'" Skinner v. Oklahoma ex rel. Williamson, 316 U. S. 535, 541 (1942) (quoting Yick Wo v. Hopkins, 118 U. S. 356, 369 (1886)).

          Davis v. Beason, 133 U. S. 333 (1890), not cited by the parties but relied upon by the dissent, is not evidence that Amendment 2 is within our constitutional tradition, and any reliance upon it as authority for sustaining the amendment is misplaced. In Davis, the Court approved an Idaho territorial statute denying Mormons, polygamists, and advocates of polygamy the right to vote and to hold office because, as the Court construed the statute, it "simply excludes from the privilege of voting, or of holding any office of honor, trust or profit, those who have been convicted of certain offences, and those who advocate a practical resistance to the laws of the Territory and justify and approve the commission of crimes forbidden by it." Id., at 347. To the extent Davis held that persons advocating a certain practice may be denied the right to vote, it is no longer good law. Brandenburg v. Ohio, 395 U. S. 444 (1969) (per curiam). To the extent it held that the groups designated in the statute may be deprived of the right to vote because of their status, its ruling could not stand without surviving strict scrutiny, a most doubtful outcome. Dunn v. Blumstein, 405 U. S. 330, 337 (1972); cf. United States v. Brown, 381 U. S. 437 (1965); United States v. Robel, 389 U. S. 258 (1967). To the extent Davis held that a convicted felon may be denied the right to vote, its holding is not implicated by our decision and is unexceptionable. See Richardson v. Ramirez, 418 U. S. 24 (1974).

          A second and related point is that laws of the kind now before us raise the inevitable inference that the disadvantage imposed is born of animosity toward the class of persons affected. "[I]f the constitutional conception of `equal protection of the laws' means anything, it must at the very least mean that a bare . . . desire to harm a politically unpopular group cannot constitute a legitimate governmental interest." Department of Agriculture v. Moreno, 413 U. S. 528, 534 (1973). Even laws enacted for broad and ambitious purposes often can be explained by reference to legitimate public policies which justify the incidental disadvantages they impose on certain persons. Amendment 2, however, in making a general announcement that gays and lesbians shall not have any particular protections from the law, inflicts on them immediate, continuing, and real injuries that outrun and belie any legitimate justifications that may be claimed for it. We conclude that, in addition to the far-reaching deficiencies of Amendment 2 that we have noted, the principles it offends, in another sense, are conventional and venerable; a law must bear a rational relationship to a legitimate governmental purpose, Kadrmas v. Dickinson Public Schools, 487 U. S. 450, 462 (1988), and Amendment 2 does not.

          The primary rationale the State offers for Amendment 2 is respect for other citizens' freedom of association, and in particular the liberties of landlords or employers who have personal or religious objections to homosexuality. Colorado also cites its interest in conserving resources to fight discrimination against other groups. The breadth of the Amendment is so far removed from these particular justifications that we find it impossible to credit them. We cannot say that Amendment 2 is directed to any identifiable legitimate purpose or discrete objective. It is a status-based enactment divorced from any factual context from which we could discern a relationship to legitimate state interests; it is a classification of persons undertaken for its own sake, something the Equal Protection Clause does not permit. "[C]lass legislation . . . [is] obnoxious to the prohibitions of the Fourteenth Amendment . . . ." Civil Rights Cases, 109 U. S., at 24.

          We must conclude that Amendment 2 classifies homosexuals not to further a proper legislative end but to make them unequal to everyone else. This Colorado cannot do. A State cannot so deem a class of persons a stranger to its laws. Amendment 2 violates the Equal Protection Clause, and the judgment of the Supreme Court of Colorado is affirmed.

          It is so ordered.

djq Justice Scalia, with whom The Chief Justice and Justice Thomas join, dissenting.

          The Court has mistaken a Kulturkampf for a fit of spite. The constitutional amendment before us here is not the manifestation of a "`bare . . . desire to harm'" homosexuals, ante, at 13, but is rather a modest attempt by seemingly tolerant Coloradans to preserve traditional sexual mores against the efforts of a politically powerful minority to revise those mores through use of the laws. That objective, and the means chosen to achieve it, are not only unimpeachable under any constitutional doctrine hitherto pronounced (hence the opinion's heavy reliance upon principles of righteousness rather than judicial holdings); they have been specifically approved by the Congress of the United States and by this Court.

          In holding that homosexuality cannot be singled out for disfavorable treatment, the Court contradicts a decision, unchallenged here, pronounced only 10 years ago, see Bowers v. Hardwick, 478 U. S. 186 (1986), and places the prestige of this institution behind the proposition that opposition to homosexuality is as reprehensible as racial or religious bias. Whether it is or not is precisely the cultural debate that gave rise to the Colorado constitutional amendment (and to the preferential laws against which the amendment was directed). Since the Constitution of the United States says nothing about this subject, it is left to be resolved by normal democratic means, including the democratic adoption of provisions in state constitutions. This Court has no business imposing upon all Americans the resolution favored by the elite class from which the Members of this institution are selected, pronouncing that "animosity" toward homosexuality, ante, at 13, is evil. I vigorously dissent.

          Let me first discuss Part II of the Court's opinion, its longest section, which is devoted to rejecting the State's arguments that Amendment 2 "puts gays and lesbians in the same position as all other persons," and "does no more than deny homosexuals special rights," ante, at 4. The Court concludes that this reading of Amendment 2's language is "implausible" under the "authoritative construction" given Amendment 2 by the Supreme Court of Colorado. Ibid.

          In reaching this conclusion, the Court considers it unnecessary to decide the validity of the State's argument that Amendment 2 does not deprive homosexuals of the "protection [afforded by] general laws and policies that prohibit arbitrary discrimination in governmental and private settings." Ante, at 8. I agree that we need not resolve that dispute, because the Supreme Court of Colorado has resolved it for us. In Evans v. Romer, 882 P. 2d 1335 (1994), the Colorado court stated:

                    "[I]t is significant to note that Colorado law currently proscribes discrimination against persons who are not suspect classes, including discrimination based on age, Section(s) 24-34-402(1)(a), 10A C. R. S. (1994 Supp.); marital or family status, Section(s) 24-34-502(1)(a), 10A C. R. S. (1994 Supp.); veterans' status, Section(s) 28-3-506, 11B C. R. S. (1989); and for any legal, off-duty conduct such as smoking tobacco, Section(s) 24-34-402.5, 10A C. R. S. (1994 Supp.). Of course Amendment 2 is not intended to have any effect on this legislation, but seeks only to prevent the adoption of anti-discrimination laws intended to protect gays, lesbians, and bisexuals." Id., at 1346, n. 9 (emphasis added).

          The Court utterly fails to distinguish this portion ofthe Colorado court's opinion. Colorado Rev. Stat. Section(s) 24-34-402.5 (Supp. 1995), which this passage authoritatively declares not to be affected by Amendment 2, was respondents' primary example of a generally applicable law whose protections would be unavailable to homosexuals under Amendment 2. See Brief for Respondents Evans et al. 11-12. The clear import of the Colorado court's conclusion that it is not affected is that "general laws and policies that prohibit arbitrary discrimination" would continue to prohibit discrimination on the basis of homosexual conduct as well. This analysis, which is fully in accord with (indeed, follows inescapably from) the text of the constitutional provision, lays to rest such horribles, raised in the course of oral argument, as the prospect that assaults upon homosexuals could not be prosecuted. The amendment prohibits special treatment of homosexuals, and nothing more. It would not affect, for example, a requirement of state law that pensions be paid to all retiring state employees with a certain length of service; homosexual employees, as well as others, would be entitled to that benefit. But it would prevent the State or any municipality from making death-benefit payments to the "life partner" of a homosexual when it does not make such payments to the long-time roommate of a nonhomosexual employee. Or again, it does not affect the requirement of the State's general insurance laws that customers be afforded coverage without discrimination unrelated to anticipated risk. Thus, homosexuals could not be denied coverage, or charged a greater premium, with respect to auto collision insurance; but neither the State nor any municipality could require that distinctive health insurance risks associated with homosexuality (if there are any) be ignored.

          Despite all of its hand-wringing about the potential effect of Amendment 2 on general antidiscrimination laws, the Court's opinion ultimately does not dispute all this, but assumes it to be true. See ante, at 9. The only denial of equal treatment it contends homosexuals have suffered is this: They may not obtain preferential treatment without amending the state constitution. That is to say, the principle underlying the Court's opinion is that one who is accorded equal treatment under the laws, but cannot as readily as others obtain preferential treatment under the laws, has been denied equal protection of the laws. If merely stating this alleged "equal protection" violation does not suffice to refute it, our constitutional jurisprudence has achieved terminal silliness.

          The central thesis of the Court's reasoning is that any group is denied equal protection when, to obtain advantage (or, presumably, to avoid disadvantage), it must have recourse to a more general and hence more difficult level of political decisionmaking than others. The world has never heard of such a principle, which is why the Court's opinion is so long on emotive utterance and so short on relevant legal citation. And it seems to me most unlikely that any multilevel democracy can function under such a principle. For whenever a disadvantage is imposed, or conferral of a benefit is prohibited, at one of the higher levels of democratic decisionmaking (i.e., by the state legislature rather than local government, or by the people at large in the state constitution rather than the legislature), the affected group has (under this theory) been denied equal protection. To take the simplest of examples, consider a state law prohibiting the award of municipal contracts to relatives of mayors or city councilmen. Once such a law is passed, the group composed of such relatives must, in order to get the benefit of city contracts, persuade the state legislature-unlike all other citizens, who need only persuade the municipality. It is ridiculous to consider this a denial of equal protection, which is why the Court's theory is unheard-of.

          The Court might reply that the example I have given is not a denial of equal protection only because the same "rational basis" (avoidance of corruption) which renders constitutional the substantive discrimination against relatives (i.e., the fact that they alone cannot obtain city contracts) also automatically suffices to sustain what might be called the electoral-procedural discrimination against them (i.e., the fact that they must go to the state level to get this changed). This is of course a perfectly reasonable response, and would explain why "electoral-procedural discrimination" has not hitherto been heard of: a law that is valid in its substance is automatically valid in its level of enactment. But the Court cannot afford to make this argument, for as I shall discuss next, there is no doubt of a rational basis for the substance of the prohibition at issue here. The Court's entire novel theory rests upon the proposition that there is something special-something that cannot be justified by normal "rational basis" analysis-in making a disadvantaged group (or a nonpreferred group) resort to a higher decisionmaking level. That proposition finds no support in law or logic.

          I turn next to whether there was a legitimate rational basis for the substance of the constitutional amendment-for the prohibition of special protection for homosexuals. ¹ It is unsurprising that the Court avoids discussion of this question, since the answer is so obviously yes. The case most relevant to the issue before us today is not even mentioned in the Court's opinion: In Bowers v. Hardwick, 478 U. S. 186 (1986), we held that the Constitution does not prohibit what virtually all States had done from the founding of the Republic until very recent years-making homosexual conduct a crime. That holding is unassailable, except by those who think that the Constitution changes to suit current fashions. But in any event it is a given in the present case: Respondents' briefs did not urge overruling Bowers, and at oral argument respondents' counsel expressly disavowed any intent to seek such overruling, Tr. of Oral Arg. 53. If it is constitutionally permissible for a State to make homosexual conduct criminal, surely it is constitutionally permissible for a State to enact other laws merely disfavoring homosexual conduct. (As the Court of Appeals for the District of Columbia Circuit has aptly put it: "If the Court [in Bowers] 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." Padula v. Webster, 822 F. 2d 97, 103 (1987).) And a fortiori it is constitutionally permissible for a State to adopt a provision not even disfavoring homosexual conduct, but merely prohibiting all levels of state government from bestowing special protections upon homosexual conduct. Respondents (who, unlike the Court, cannot afford the luxury of ignoring inconvenient precedent) counter Bowers with the argument that a greater-includes-the-lesser rationale cannot justify Amendment 2's application to individuals who do not engage in homosexual acts, but are merely of homosexual "orientation." Some courts of appeals have concluded that, with respect to laws of this sort at least, that is a distinction without a difference. See Equality Foundation of Greater Cincinnati, Inc. v. Cincinnati, 54 F. 3d 261, 267 (CA6 1995) ("[F]or purposes of these proceedings, it is virtually impossible to distinguish or separate individuals of a particular orientation which predisposes them toward a particular sexual conduct from those who actually engage in that particular type of sexual conduct"); Steffan v. Perry, 41 F. 3d 677, 689-690 (CADC 1994). The Supreme Court of Colorado itself appears to be of this view. See 882 P. 2d, at 1349-1350 ("Amendment 2 targets this class of persons based on four characteristics: sexual orientation; conduct; practices; and relationships. Each characteristic provides a potentially different way of identifying that class of persons who are gay, lesbian, or bisexual. These four characteristics are not truly severable from one another because each provides nothing more than a different way of identifying the same class of persons") (emphasis added).

          But assuming that, in Amendment 2, a person of homosexual "orientation" is someone who does not engage in homosexual conduct but merely has a tendency or desire to do so, Bowers still suffices to establish a rational basis for the provision. If it is rational to criminalize the conduct, surely it is rational to deny special favor and protection to those with a self-avowed tendency or desire to engage in the conduct. Indeed, where criminal sanctions are not involved, homosexual "orientation" is an acceptable stand-in for homosexual conduct. A State "does not violate the Equal Protection Clause merely because the classifications made by its laws are imperfect," Dandridge v. Williams, 397 U. S. 471, 485 (1970). Just as a policy barring the hiring of methadone users as transit employees does not violate equal protection simply because some methadone users pose no threat to passenger safety, see New York City Transit Authority v. Beazer, 440 U. S. 568 (1979), and just as a mandatory retirement age of 50 for police officers does not violate equal protection even though it prematurely ends the careers of many policemen over 50 who still have the capacity to do the job, see Massachusetts Bd. of Retirement v. Murgia, 427 U. S. 307 (1976) (per curiam), Amendment 2 is not constitutionally invalid simply because it could have been drawn more precisely so as to withdraw special antidiscrimination protections only from those of homosexual "orientation" who actually engage in homosexual conduct. As Justice Kennedy wrote, when he was on the Court of Appeals, in a case involving discharge of homosexuals from the Navy: "Nearly any statute which classifies people may be irrational as applied in particular cases. Discharge of the particular plaintiffs before us would be rational, under minimal scrutiny, not because their particular cases present the dangers which justify Navy policy, but instead because the general policy of discharging all homosexuals is rational." Beller v. Middendorf, 632 F. 2d 788, 808-809, n. 20 (CA9 1980) (citation omitted). See also Ben-Shalom v. Marsh, 881 F. 2d 454, 464 (CA7 1989), cert. denied, 494 U. S. 1004 (1990).

          Moreover, even if the provision regarding homosexual "orientation" were invalid, respondents' challenge to Amendment 2-which is a facial challenge-must fail. "A facial challenge to a legislative Act is, of course, the most difficult challenge to mount successfully, since the challenger must establish that no set of circumstances exists under which the Act would be valid." United States v. Salerno, 481 U. S. 739, 745 (1987). It would not be enough for respondents to establish (if they could) that Amendment 2 is unconstitutional as applied to those of homosexual "orientation"; since, under Bowers, Amendment 2 is unquestionably constitutional as applied to those who engage in homosexual conduct, the facial challenge cannot succeed. Some individuals of homosexual "orientation" who do not engage in homosexual acts might successfully bring an as-applied challenge to Amendment 2, but so far as the record indicates, none of the respondents is such a person. See App. 4-5 (complaint describing each of the individual respondents as either "a gay man" or "a lesbian").2

          The foregoing suffices to establish what the Court's failure to cite any case remotely in point would lead one to suspect: No principle set forth in the Constitution, nor even any imagined by this Court in the past 200 years, prohibits what Colorado has done here. But the case for Colorado is much stronger than that. What it has done is not only unprohibited, but eminently reasonable, with close, congressionally approved precedent in earlier constitutional practice.

          First, as to its eminent reasonableness. The Court's opinion contains grim, disapproving hints that Coloradans have been guilty of "animus" or "animosity" toward homosexuality, as though that has been established as Unamerican. Of course it is our moral heritage that one should not hate any human being or class of human beings. But I had thought that one could consider certain conduct reprehensible-murder, for example, or polygamy, or cruelty to animals-and could exhibit even "animus" toward such conduct. Surely that is the only sort of "animus" at issue here: moral disapproval of homosexual conduct, the same sort of moral disapproval that produced the centuries-old criminal laws that we held constitutional in Bowers. The Colorado amendment does not, to speak entirely precisely, prohibit giving favored status to people who are homosexuals; they can be favored for many reasons-for example, because they are senior citizens or members of racial minorities. But it prohibits giving them favored status because of their homosexual conduct-that is, it prohibits favored status for homosexuality.

          But though Coloradans are, as I say, entitled to be hostile toward homosexual conduct, the fact is that the degree of hostility reflected by Amendment 2 is the smallest conceivable. The Court's portrayal of Coloradans as a society fallen victim to pointless, hate-filled "gay-bashing" is so false as to be comical. Colorado not only is one of the 25 States that have repealed their antisodomy laws, but was among the first to do so. See 1971 Colo. Sess. Laws, ch. 121, Section(s) 1. But the society that eliminates criminal punishment for homosexual acts does not necessarily abandon the view that homosexuality is morally wrong and socially harmful; often, abolition simply reflects the view that enforcement of such criminal laws involves unseemly intrusion into the intimate lives of citizens. Cf. Brief for Lambda Legal Defense and Education Fund, Inc., et al. as Amici Curiae in Bowers v. Hardwick, O. T. 1985, No. 85-140, p. 25, n. 21 (antisodomy statutes are "unenforceable by any but the most offensive snooping and wasteful allocation of law enforcement resources"); Kadish, The Crisis of Overcriminalization, 374 The Annals of the American Academy of Political and Social Science 157, 161 (1967) ("To obtain evidence [in sodomy cases], police are obliged to resort to behavior which tends to degrade and demean both themselves personally and law enforcement as an institution").

          There is a problem, however, which arises when criminal sanction of homosexuality is eliminated but moral and social disapprobation of homosexuality is meant to be retained. The Court cannot be unaware of that problem; it is evident in many cities of the country, and occasionally bubbles to the surface of the news, in heated political disputes over such matters as the introduction into local schools of books teaching that homosexuality is an optional and fully acceptable "alternate life style." The problem (a problem, that is, for those who wish to retain social disapprobation of homosexuality) is that, because those who engage in homosexual conduct tend to reside in disproportionate numbers in certain communities, see Record, Exh. MMM, have high disposable income, see ibid.; App. 254 (affidavit of Prof. James Hunter), 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. Quite understandably, they devote this political power to achieving not merely a grudging social toleration, but full social acceptance, of homosexuality. See, e.g., Jacobs, The Rhetorical Construction of Rights: The Case of the Gay Rights Movement, 1969-1991, 72 Neb. L. Rev. 723, 724 (1993) ("[T]he task of gay rights proponents is to move the center of public discourse along a continuum from the rhetoric of disapprobation, to rhetoric of tolerance, and finally to affirmation").

          By the time Coloradans were asked to vote on Amendment 2, their </B></I>exposure to homosexuals' quest for social endorsement was not limited to newspaper accounts of happenings in places such as New York, Los Angeles, San Francisco, and Key West. Three Colorado citiesAspen, Boulder, and Denver-had enacted ordinances that listed "sexual orientation" as an impermissible ground for discrimination, equating the moral disapproval of homosexual conduct with racial and religious bigotry. See Aspen Municipal Code Section(s) 13-98 (1977); Boulder Rev. Municipal Code Section(s) 12-1-1 to 12-1-11 (1987); Denver Rev. Municipal Code, Art. IV Section(s) 28-91 to 28-116 (1991). The phenomenon had even appeared statewide: the Governor of Colorado had signed an executive order pronouncing that "in the State of Colorado we recognize the diversity in our pluralistic society and strive to bring an end to discrimination in any form," and directing state agency-heads to "ensure non-discrimination" in hiring and promotion based on, among other things, "sexual orientation." Executive Order No. D0035 (Dec. 10, 1990). I do not mean to be critical of these legislative successes; homosexuals are as entitled to use the legal system for reinforcement of their moral sentiments as are the rest of society. But they are subject to being countered by lawful, democratic countermeasures as well.

          That is where Amendment 2 came in. It sought to counter both the geographic concentration and the disproportionate political power of homosexuals by (1) resolving the controversy at the statewide level, and (2) making the election a single-issue contest for both sides. It put directly, to all the citizens of the State, the question: Should homosexuality be given special protection? They answered no. The Court today asserts that this most democratic of procedures is unconstitutional. Lacking any cases to establish that facially absurd proposition, it simply asserts that it must be unconstitutional, because it has never happened before.

                    "[Amendment 2] identifies persons by a single trait and then denies them protection across the board. The resulting disqualification of a class of persons from the right to seek specific protection from the law is unprecedented in our jurisprudence. The absence of precedent for Amendment 2 is itselfinstructive . . . .

                    "It is not within our constitutional tradition to enact laws of this sort. Central both to the idea of the rule of law and to our own Constitution's guarantee of equal protection is the principle that government and each of its parts remain open on impartial terms to all who seek its assistance." Ante, at 11-12.

          As I have noted above, this is proved false every time a state law prohibiting or disfavoring certain conduct is passed, because such a law prevents the adversely affected group-whether drug addicts, or smokers, or gun owners, or motorcyclists-from changing the policy thus established in "each of [the] parts" of the State. What the Court says is even demonstrably false at the constitutional level. The Eighteenth Amendment to the Federal Constitution, for example, deprived those who drank alcohol not only of the power to alter the policy of prohibition locally or through state legislation, but even of the power to alter it through state constitutional amendment or federal legislation. The Establishment Clause of the First Amendment prevents theocrats from having their way by converting their fellow citizens at the local, state, or federal statutory level; as does the Republican Form of Government Clause prevent monarchists.

          But there is a much closer analogy, one that involves precisely the effort by the majority of citizens to preserve its view of sexual morality statewide, against the efforts of a geographically concentrated and politically powerful minority to undermine it. The constitutions of the States of Arizona, Idaho, New Mexico, Oklahoma, and Utah to this day contain provisions stating that polygamy is "forever prohibited." See Ariz. Const., Art. XX, par. 2; Idaho Const., Art. I, Section(s) 4; N. M. Const., Art. XXI, Section(s) 1; Okla. Const., Art. I, Section(s) 2; Utah Const., Art. III, Section(s) 1. Polygamists, and those who have a polygamous "orientation," have been "singled out" by these provisions for much more severe treatment than merely denial of favored status; and that treatment can only be changed by achieving amendment of the state constitutions. The Court's disposition today suggests that these provisions are unconstitutional, and that polygamy must be permitted in these States on a state-legislated, or perhaps even local-option, basis-unless, of course, polygamists for some reason have fewer constitutional rights than homosexuals.

          The United States Congress, by the way, required the inclusion of these antipolygamy provisions in the constitutions of Arizona, New Mexico, Oklahoma, and Utah, as a condition of their admission to statehood. See Arizona Enabling Act, 36 Stat. 569; New Mexico Enabling Act, 36 Stat. 558; Oklahoma Enabling Act, 34 Stat. 269; Utah Enabling Act, 28 Stat. 108. (For Arizona, New Mexico, and Utah, moreover, the Enabling Acts required that the antipolygamy provisions be "irrevocable without the consent of the United States and the people of said State"-so that not only were "each of [the] parts" of these States not "open on impartial terms" to polygamists, but even the States as a whole were not; polygamists would have to persuade the whole country to their way of thinking.) Idaho adopted the constitutional provision on its own, but the 51st Congress, which admitted Idaho into the Union, found its constitution to be "republican in form and . . . in conformity with the Constitution of the United States." Act of Admission of Idaho, 26 Stat. 215 (emphasis added). Thus, this "singling out" of the sexual practices of a single group for statewide, democratic vote-so utterly alien to our constitutional system, the Court would have us believe-has not only happened, but has received the explicit approval of the United States Congress.

          I cannot say that this Court has explicitly approved any of these state constitutional provisions; but it has approved a territorial statutory provision that went even further, depriving polygamists of the ability even to achieve a constitutional amendment, by depriving them of the power to vote. In Davis v. Beason, 133 U. S. 333 (1890), Justice Field wrote for a unanimous Court:

          "In our judgment, Section(s) 501 of the Revised Statutes of Idaho Territory, which provides that `no person . . . who is a bigamist or polygamist or who teaches, advises, counsels, or encourages any person or persons to become bigamists or polygamists, or to commit any other crime defined by law, or to enter into what is known as plural or celestial marriage, or who is a member of any order, organization or association which teaches, advises, counsels, or encourages its members or devotees or any other persons to commit the crime of bigamy or polygamy, or any other crime defined by law . . . is permitted to vote at any election, or to hold any position or office of honor, trust, or profit within this Territory,' is not open to any constitutional or legal objection." Id., at 346-347 (emphasis added).

          To the extent, if any, that this opinion permits the imposition of adverse consequences upon mere abstract advocacy of polygamy, it has of course been overruled by later cases. See Brandenburg v. Ohio, 395 U. S. 444 (1969) (per curiam). But the proposition that polygamy can be criminalized, and those engaging in that crime deprived of the vote, remains good law. See Richardson v. Ramirez, 418 U. S. 24, 53 (1974). Beason rejected the argument that "such discrimination is a denial of the equal protection of the laws." Brief for Appellant in Davis v. Beason, O. T. 1889, No. 1261, p. 41. Among the Justices joining in that rejection were the two whose views in other cases the Court today treats as equal-protection lodestars-Justice Harlan, who was to proclaim in Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (dissenting opinion), that the Constitution "neither knows nor tolerates classes among citizens," quoted ante, at 1, and Justice Bradley, who had earlier declared that "class legislation . . . [is] obnoxious to the prohibitions of the Fourteenth Amendment," Civil Rights Cases, 109 U. S. 3, 24 (1883), quoted ante, at 14. ³

          This Court cited Beason with approval as recently as 1993, in an opinion authored by the same Justice who writes for the Court today. That opinion said: "[A]dverse impact will not always lead to a finding of impermissible targeting. For example, a social harm may have been a legitimate concern of government for reasons quite apart from discrimination. . . . See, e.g., . . . Davis v. Beason, 133 U. S. 333 (1890)." Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U. S. 520, 535 (1993). It remains to be explained how Section(s) 501 of the Idaho Revised Statutes was not an "impermissible targeting" of polygamists, but (the much more mild) Amendment 2 is an "impermissible targeting" of homosexuals. Has the Court concluded that the perceived social harm of polygamy is a "legitimate concern of government," and the perceived social harm of homosexuality is not?

          I strongly suspect that the answer to the last question is yes, which leads me to the last point I wish to make: The Court today, announcing that Amendment 2 "defies . . . conventional [constitutional] inquiry," ante, at 10, and "confounds [the] normal process of judicial review," ante, at 11, employs a constitutional theory heretofore unknown to frustrate Colorado's reasonable effort to preserve traditional American moral values. The Court's stern disapproval of "animosity" towards homosexuality might be compared with what an earlier Court (including the revered Justices Harlan and Bradley) said in Murphy v. Ramsey, 114 U. S. 15 (1885), rejecting a constitutional challenge to a United States statute that denied the franchise in federal territories to those who engaged in polygamous cohabitation:

          "[C]ertainly no legislation can be supposed more wholesome and necessary in the founding of a free, self-governing commonwealth, fit to take rank as one of the co-ordinate States of the Union, than that which seeks to establish it on the basis of the idea of the family, as consisting in and springing from the union for life of one man and one woman in the holy estate of matrimony; the sure foundation of all that is stable and noble in our civilization; the best guaranty of that reverent morality which is the source of all beneficent progress in social and political improvement." Id., at 45.

          I would not myself indulge in such official praise for heterosexual monogamy, because I think it no business of the courts (as opposed to the political branches) to take sides in this culture war.

          But the Court today has done so, not only by inventing a novel and extravagant constitutional doctrine to take the victory away from traditional forces, but even by verbally disparaging as bigotry adherence to traditional attitudes. To suggest, for example, that this constitutional amendment springs from nothing more than "`a bare . . . desire to harm a politically unpopular group,'" ante, at 13, quoting Department of Agriculture v. Moreno, 413 U. S. 528, 534 (1973), is nothing short of insulting. (It is also nothing short of preposterous to call "politically unpopular" a group which enjoys enormous influence in American media and politics, and which, as the trial court here noted, though composing no more than 4% of the population had the support of 46% of the voters on Amendment 2, see App. to Pet. for Cert. C-18.)

          When the Court takes sides in the culture wars, it tends to be with the knights rather than the villeins-and more specifically with the Templars, reflecting the views and values of the lawyer class from which the Court's Members are drawn. How that class feels about homosexuality will be evident to anyone who wishes to interview job applicants at virtually any of the Nation's law schools. The interviewer may refuse to offer a job because the applicant is a Republican; because he is an adulterer; because he went to the wrong prep school or belongs to the wrong country club; because he eats snails; because he is a womanizer; because she wears real-animal fur; or even because he hates the Chicago Cubs. But if the interviewer should wish not to be an associate or partner of an applicant because he disapproves of the applicant's homosexuality, then he will have violated the pledge which the Association of American Law Schools requires all its member-schools to exact from job interviewers: "assurance of the employer's willingness" to hire homosexuals. Bylaws of the Association of American Law Schools, Inc. Section(s) 6-4(b); Executive Committee Regulations of the Association of American Law Schools Section(s) 6.19, in 1995 Handbook, Association of American Law Schools. This law-school view of what "prejudices" must be stamped out may be contrasted with the more plebeian attitudes that apparently still prevail in the United States Congress, which has been unresponsive to repeated attempts to extend to homosexuals the protections of federal civil rights laws, see, e.g., Employment Non-Discrimination Act of 1994, S. 2238, 103d Cong., 2d Sess. (1994); Civil Rights Amendments of 1975, H. R. 5452, 94th Cong., 1st Sess. (1975), and which took the pains to exclude them specifically from the Americans With Disabilities Act of 1990, see 42 U. S. C. Section(s) 12211(a) (1988 ed., Supp. V).

          Today's opinion has no foundation in American constitutional law, and barely pretends to. The people of Colorado have adopted an entirely reasonable provision which does not even disfavor homosexuals in any substantive sense, but merely denies them preferential treatment. Amendment 2 is designed to prevent piecemeal deterioration of the sexual morality favored by a majority of Coloradans, and is not only an appropriate means to that legitimate end, but a means that Americans have employed before. Striking it down is an act, not of judicial judgment, but of political will. I dissent.

^{* The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Lumber Co., 200 U. S. 321, 337.}

¹ The Court evidently agrees that "rational basis"-the normal test for compliance with the Equal Protection Clause-is the governing standard. The trial court rejected respondents' argument that homosexuals constitute a "suspect" or "quasi-suspect" class, and respondents elected not to appeal that ruling to the Supreme Court of Colorado. See Evans v. Romer, 882 P. 2d 1335, 1341, n. 3 (1994). And the Court implicitly rejects the Supreme Court of Colorado's holding, see Evans v. Romer, 854 P. 2d 1270, 1282 (1993), that Amendment 2 infringes upon a "fundamental right" of "independently identifiable class[es]" to "participate equally in the political process." Ante, at 4.

The Supreme Court of Colorado stated: "We hold that the portions of Amendment 2 that would remain if only the provision concerning sexual orientation were stricken are not autonomous and thus, not severable," 882 P. 2d, at 1349. That statement was premised, however, on the proposition that "[the] four characteristics [described in the Amendment-sexual orientation, conduct, practices, and relationships] are not truly severable from one another because each provides nothing more than a different way of identifying the same class of persons." Id., at 1349-1350 (emphasis added). As I have discussed above, if that premise is true-if the entire class affected by the Amendment takes part in homosexual conduct, practices and relationships-Bowers alone suffices to answer all constitutional objections. Separate consideration of persons of homosexual "orientation" is necessary only if one believes (as the Supreme Court of Colorado id not) that that is a distinct class.

The Court labors mightily to get around Beason, see ante, at 12-13, but cannot escape the central fact that this Court found the statute at issue-which went much further than Amendment 2, denying polygamists not merely special treatment but the right to vote-"not open to any constitutional or legal objection," rejecting the appellant's argument (much like the argument of respondents today) that the statute impermissibly "single[d] him out," Brief for Appellant in Davis v. Beason, O. T. 1889, No. 1261, p. 41. The Court adopts my conclusions that (a) insofar as Beason permits the imposition of adverse consequences based upon mere advocacy, it has been overruled by subsequent cases, and (b) insofar as Beason holds that convicted felons may be denied the right to vote, it remains good law. To these conclusions, it adds something new: the claim that "[t]o the extent [Beason] held that the groups designated in the statute may be deprived of the right to vote because of their status, its ruling could not stand without surviving strict scrutiny, a most doubtful outcome." Ante, at 12-13. But if that is so, it is only because we have declared the right to vote to be a "fundamental political right," see, e.g., Dunn v. Blumstein, 405 U. S. 330, 336 (1972), deprivation of which triggers strict scrutiny. Amendment 2, of course, does not deny the fundamental right to vote, and the Court rejects the Colorado court's view that there exists a fundamental right to participate in the political process. Strict scrutiny is thus not in play here. See ante, at 10. Finally, the Court's suggestion that Section(s) 501 of the Revised Statutes of Idaho, and Amendment 2, deny rights on account of "status" (rather than conduct) opens up a broader debate involving the significance of Bowers to this case, a debate which the Court is otherwise unwilling to join, see supra, at 6-9.

JUSTICE KENNEDY, delivered the opinion of the Court.

Two women then resident in New York were married in a lawful ceremony in Ontario, Canada, in 2007. Edith Windsor and Thea Spyer returned to their home in New York City. When Spyer died in 2009, she left her entire estate to Windsor. Windsor sought to claim the estate tax exemption for surviving spouses. She was barred from doing so, however, by a federal law, the Defense of Marriage Act, which excludes a same-sex partner from the definition of "spouse" as that term is used in federal statutes. Windsor paid the taxes but filed suit to challenge the constitutionality of this provision. The United States District Court and the Court of Appeals ruled that this portion of the statute is unconstitutional and ordered the United States to pay Windsor a refund. This Court granted certiorari and now affirms the judgment in Windsor's favor.

I

In 1996, as some States were beginning to consider the concept of same-sex marriage, see, e.g., Baehr v. Lewin, 74 Haw. 530, 852 P. 2d 44 (1993), and before any State had acted to permit it, Congress enacted the Defense of Marriage Act (DOMA), 110 Stat. 2419. DOMA contains two operative sections: Section 2, which has not been challenged here, allows States to refuse to recognize same-sex marriages performed under the laws of other States. See 28 U. S. C. §1738C.

Section 3 is at issue here. It amends the Dictionary Act in Title 1, §7, of the United States Code to provide a federal definition of "marriage" and "spouse." Section 3 of DOMA provides as follows:

"In determining the meaning of any Act of Congress, or of any ruling, regulation, or interpretation of the various administrative bureaus and agencies of the United States, the word `marriage' means only a legal union between one man and one woman as husband and wife, and the word `spouse' refers only to a person of the opposite sex who is a husband or a wife." 1 U. S. C. §7.

The definitional provision does not by its terms forbid States from enacting laws permitting same-sex marriages or civil unions or providing state benefits to residents in that status. The enactment's comprehensive definition of marriage for purposes of all federal statutes and other regulations or directives covered by its terms, however, does control over 1,000 federal laws in which marital or spousal status is addressed as a matter of federal law. See GAO, D. Shah, Defense of Marriage Act: Update to Prior Report 1 (GAO-04-353R, 2004).

Edith Windsor and Thea Spyer met in New York City in 1963 and began a long-term relationship. Windsor and Spyer registered as domestic partners when New York City gave that right to same-sex couples in 1993. Concerned about Spyer's health, the couple made the 2007 trip to Canada for their marriage, but they continued to reside in New York City. The State of New York deems their Ontario marriage to be a valid one. See 699 F. 3d 169, 177-178 (CA2 2012).

Spyer died in February 2009, and left her entire estate to Windsor. Because DOMA denies federal recognition to same-sex spouses, Windsor did not qualify for the marital exemption from the federal estate tax, which excludes from taxation "any interest in property which passes or has passed from the decedent to his surviving spouse." 26 U. S. C. §2056(a). Windsor paid $363,053 in estate taxes and sought a refund. The Internal Revenue Service denied the refund, concluding that, under DOMA, Windsor was not a "surviving spouse." Windsor commenced this refund suit in the United States District Court for the Southern District of New York. She contended that DOMA violates the guarantee of equal protection, as applied to the Federal Government through the Fifth Amendment.

While the tax refund suit was pending, the Attorney General of the United States notified the Speaker of the House of Representatives, pursuant to 28 U. S. C. §530D, that the Department of Justice would no longer defend the constitutionality of DOMA's §3. Noting that "the Department has previously defended DOMA against . . . challenges involving legally married same-sex couples," App. 184, the Attorney General informed Congress that "the President has concluded that given a number of factors, including a documented history of discrimination, classifications based on sexual orientation should be subject to a heightened standard of scrutiny." Id., at 191. The Department of Justice has submitted many §530D letters over the years refusing to defend laws it deems unconstitutional, when, for instance, a federal court has rejected the Government's defense of a statute and has issued a judgment against it. This case is unusual, however, because the §530D letter was not preceded by an adverse judgment. The letter instead reflected the Executive's own conclusion, relying on a definition still being debated and considered in the courts, that heightened equal protection scrutiny should apply to laws that classify on the basis of sexual orientation.

Although "the President . . . instructed the Department not to defend the statute in Windsor," he also decided "that Section 3 will continue to be enforced by the Executive Branch" and that the United States had an "interest in providing Congress a full and fair opportunity to participate in the litigation of those cases." Id., at 191-193. The stated rationale for this dual-track procedure (determination of unconstitutionality coupled with ongoing enforcement) was to "recogniz[e] the judiciary as the final arbiter of the constitutional claims raised." Id., at 192.

In response to the notice from the Attorney General, the Bipartisan Legal Advisory Group (BLAG) of the House of Representatives voted to intervene in the litigation to defend the constitutionality of §3 of DOMA. The Department of Justice did not oppose limited intervention by BLAG. The District Court denied BLAG's motion to enter the suit as of right, on the rationale that the United States already was represented by the Department of Justice. The District Court, however, did grant intervention by BLAG as an interested party. See Fed. Rule Civ. Proc. 24(a)(2).

On the merits of the tax refund suit, the District Court ruled against the United States. It held that §3 of DOMA is unconstitutional and ordered the Treasury to refund the tax with interest. Both the Justice Department and BLAG filed notices of appeal, and the Solicitor General filed a petition for certiorari before judgment. Before this Court acted on the petition, the Court of Appeals for the Second Circuit affirmed the District Court's judgment. It applied heightened scrutiny to classifications based on sexual orientation, as both the Department and Windsor had urged. The United States has not complied with the judgment. Windsor has not received her refund, and the Executive Branch continues to enforce §3 of DOMA.

In granting certiorari on the question of the constitutionality of §3 of DOMA, the Court requested argument on two additional questions: whether the United States' agreement with Windsor's legal position precludes further review and whether BLAG has standing to appeal the case. All parties agree that the Court has jurisdiction to decide this case; and, with the case in that framework, the Court appointed Professor Vicki Jackson as amicus curiae to argue the position that the Court lacks jurisdiction to hear the dispute. 568 U. S. ___ (2012). She has ably discharged her duties.

In an unrelated case, the United States Court of Appeals for the First Circuit has also held §3 of DOMA to be unconstitutional. A petition for certiorari has been filed in that case. Pet. for Cert. in Bipartisan Legal Advisory Group v. Gill, O. T. 2012, No. 12-13.

II

It is appropriate to begin by addressing whether either the Government or BLAG, or both of them, were entitled to appeal to the Court of Appeals and later to seek certiorari and appear as parties here.

There is no dispute that when this case was in the District Court it presented a concrete disagreement between opposing parties, a dispute suitable for judicial resolution. "[A] taxpayer has standing to challenge the collection of a specific tax assessment as unconstitutional; being forced to pay such a tax causes a real and immediate economic injury to the individual taxpayer." Hein v. Freedom From Religion Foundation, Inc., 551 U. S. 587, 599 (2007) (plurality opinion) (emphasis deleted). Windsor suffered a redressable injury when she was required to pay estate taxes from which, in her view, she was exempt but for the alleged invalidity of §3 of DOMA.

The decision of the Executive not to defend the constitutionality of §3 in court while continuing to deny refunds and to assess deficiencies does introduce a complication. Even though the Executive's current position was announced before the District Court entered its judgment, the Government's agreement with Windsor's position would not have deprived the District Court of jurisdiction to entertain and resolve the refund suit; for her injury (failure to obtain a refund allegedly required by law) was concrete, persisting, and unredressed. The Government's position—agreeing with Windsor's legal contention but refusing to give it effect—meant that there was a justiciable controversy between the parties, despite what the claimant would find to be an inconsistency in that stance. Windsor, the Government, BLAG, and the amicus appear to agree upon that point. The disagreement is over the standing of the parties, or aspiring parties, to take an appeal in the Court of Appeals and to appear as parties in further proceedings in this Court.

The amicus' position is that, given the Government's concession that §3 is unconstitutional, once the District Court ordered the refund the case should have ended; and the amicus argues the Court of Appeals should have dismissed the appeal. The amicus submits that once the President agreed with Windsor's legal position and the District Court issued its judgment, the parties were no longer adverse. From this standpoint the United States was a prevailing party below, just as Windsor was. Accordingly, the amicus reasons, it is inappropriate for this Court to grant certiorari and proceed to rule on the merits; for the United States seeks no redress from the judgment entered against it.

This position, however, elides the distinction between two principles: the jurisdictional requirements of Article III and the prudential limits on its exercise. See Warth v. Seldin, 422 U. S. 490, 498 (1975). The latter are "essentially matters of judicial self-governance." Id., at 500. The Court has kept these two strands separate: "Article III standing, which enforces the Constitution's case-orcontroversy requirement, see Lujan v. Defenders of Wildlife, 504 U. S. 555, 559-562 (1992); and prudential standing, which embodies `judicially self-imposed limits on the exercise of federal jurisdiction,' Allen [v. Wright,] 468 U. S. [737,] 751 [(1984)]." Elk Grove Unified School Dist. v. Newdow, 542 U. S. 1, 11-12 (2004).

The requirements of Article III standing are familiar:

"First, the plaintiff must have suffered an `injury in fact'—an invasion of a legally protected interest which is (a) concrete and particularized, and (b) `actual or imminent, not "conjectural or hypothetical."' Second, there must be a causal connection between the injury and the conduct complained of—the injury has to be `fairly . . . trace[able] to the challenged action of the defendant, and not . . . th[e] result [of] the independent action of some third party not before the court.' Third, it must be `likely,' as opposed to merely `speculative,' that the injury will be `redressed by a favorable decision.'" Lujan, supra, at 560-561 (footnote and citations omitted).

Rules of prudential standing, by contrast, are more flexible "rule[s] . . . of federal appellate practice," Deposit Guaranty Nat. Bank v. Roper, 445 U. S. 326, 333 (1980), designed to protect the courts from "decid[ing] abstract questions of wide public significance even [when] other governmental institutions may be more competent to address the questions and even though judicial intervention may be unnecessary to protect individual rights." Warth, supra, at 500.

In this case the United States retains a stake sufficient to support Article III jurisdiction on appeal and in proceedings before this Court. The judgment in question orders the United States to pay Windsor the refund she seeks. An order directing the Treasury to pay money is "a real and immediate economic injury," Hein, 551 U. S., at 599, indeed as real and immediate as an order directing an individual to pay a tax. That the Executive may welcome this order to pay the refund if it is accompanied by the constitutional ruling it wants does not eliminate the injury to the national Treasury if payment is made, or to the taxpayer if it is not. The judgment orders the United States to pay money that it would not disburse but for the court's order. The Government of the United States has a valid legal argument that it is injured even if the Executive disagrees with §3 of DOMA, which results in Windsor's liability for the tax. Windsor's ongoing claim for funds that the United States refuses to pay thus establishes a controversy sufficient for Article III jurisdiction. It would be a different case if the Executive had taken the further step of paying Windsor the refund to which she was entitled under the District Court's ruling.

This Court confronted a comparable case in INS v. Chadha, 462 U. S. 919 (1983). A statute by its terms allowed one House of Congress to order the Immigration and Naturalization Service (INS) to deport the respondent Chadha. There, as here, the Executive determined that the statute was unconstitutional, and "the INS presented the Executive's views on the constitutionality of the House action to the Court of Appeals." Id., at 930. The INS, however, continued to abide by the statute, and "the INS brief to the Court of Appeals did not alter the agency's decision to comply with the House action ordering deportation of Chadha." Ibid. This Court held "that the INS was sufficiently aggrieved by the Court of Appeals decision prohibiting it from taking action it would otherwise take," ibid., regardless of whether the agency welcomed the judgment. The necessity of a "case or controversy" to satisfy Article III was defined as a requirement that the Court's "`decision will have real meaning: if we rule for Chadha, he will not be deported; if we uphold [the statute], the INS will execute its order and deport him.'" Id., at 939-940 (quoting Chadha v. INS, 634 F. 2d 408, 419 (CA9 1980)). This conclusion was not dictum. It was a necessary predicate to the Court's holding that "prior to Congress' intervention, there was adequate Art. III adverseness." 462 U. S., at 939. The holdings of cases are instructive, and the words of Chadha make clear its holding that the refusal of the Executive to provide the relief sought suffices to preserve a justiciable dispute as required by Article III. In short, even where "the Government largely agree[s] with the opposing party on the merits of the controversy," there is sufficient adverseness and an "adequate basis for jurisdiction in the fact that the Government intended to enforce the challenged law against that party." Id., at 940, n. 12.

It is true that "[a] party who receives all that he has sought generally is not aggrieved by the judgment affording the relief and cannot appeal from it." Roper, supra, at 333, see also Camreta v. Greene, 563 U. S. ___, ___ (2011) (slip op., at 8) ("As a matter of practice and prudence, we have generally declined to consider cases at the request of a prevailing party, even when the Constitution allowed us to do so"). But this rule "does not have its source in the jurisdictional limitations of Art. III. In an appropriate case, appeal may be permitted . . . at the behest of the party who has prevailed on the merits, so long as that party retains a stake in the appeal satisfying the requirements of Art. III." Roper, supra, at 333-334.

While these principles suffice to show that this case presents a justiciable controversy under Article III, the prudential problems inherent in the Executive's unusual position require some further discussion. The Executive's agreement with Windsor's legal argument raises the risk that instead of a "`real, earnest and vital controversy,'" the Court faces a "friendly, non-adversary, proceeding . . . [in which] `a party beaten in the legislature [seeks to] transfer to the courts an inquiry as to the constitutionality of the legislative act.'" Ashwander v. TVA, 297 U. S. 288, 346 (1936) (Brandeis, J., concurring) (quoting Chicago & Grand Trunk R. Co. v. Wellman, 143 U. S. 339, 345 (1892)). Even when Article III permits the exercise of federal jurisdiction, prudential considerations demand that the Court insist upon "that concrete adverseness which sharpens the presentation of issues upon which the court so largely depends for illumination of difficult constitutional questions." Baker v. Carr, 369 U. S. 186, 204 (1962).

There are, of course, reasons to hear a case and issue a ruling even when one party is reluctant to prevail in its position. Unlike Article III requirements—which must be satisfied by the parties before judicial consideration is appropriate—the relevant prudential factors that counsel against hearing this case are subject to "countervailing considerations [that] may outweigh the concerns underlying the usual reluctance to exert judicial power." Warth, 422 U. S., at 500-501. One consideration is the extent to which adversarial presentation of the issues is assured by the participation of amici curiae prepared to defend with vigor the constitutionality of the legislative act. With respect to this prudential aspect of standing as well, the Chadha Court encountered a similar situation. It noted that "there may be prudential, as opposed to Art. III, concerns about sanctioning the adjudication of [this case] in the absence of any participant supporting the validity of [the statute]. The Court of Appeals properly dispelled any such concerns by inviting and accepting briefs from both Houses of Congress." 462 U. S., at 940. Chadha was not an anomaly in this respect. The Court adopts the practice of entertaining arguments made by an amicus when the Solicitor General confesses error with respect to a judgment below, even if the confession is in effect an admission that an Act of Congress is unconstitutional. See, e.g., Dickerson v. United States, 530 U. S. 428 (2000).

In the case now before the Court the attorneys for BLAG present a substantial argument for the constitutionality of §3 of DOMA. BLAG's sharp adversarial presentation of the issues satisfies the prudential concerns that otherwise might counsel against hearing an appeal from a decision with which the principal parties agree. Were this Court to hold that prudential rules require it to dismiss the case, and, in consequence, that the Court of Appeals erred in failing to dismiss it as well, extensive litigation would ensue. The district courts in 94 districts throughout the Nation would be without precedential guidance not only in tax refund suits but also in cases involving the whole of DOMA's sweep involving over 1,000 federal statutes and a myriad of federal regulations. For instance, the opinion of the Court of Appeals for the First Circuit, addressing the validity of DOMA in a case involving regulations of the Department of Health and Human Services, likely would be vacated with instructions to dismiss, its ruling and guidance also then erased. See Massachusetts v. United States Dept. of Health and Human Servs., 682 F. 3d 1 (CA1 2012). Rights and privileges of hundreds of thousands of persons would be adversely affected, pending a case in which all prudential concerns about justiciability are absent. That numerical prediction may not be certain, but it is certain that the cost in judicial resources and expense of litigation for all persons adversely affected would be immense. True, the very extent of DOMA's mandate means that at some point a case likely would arise without the prudential concerns raised here; but the costs, uncertainties, and alleged harm and injuries likely would continue for a time measured in years before the issue is resolved. In these unusual and urgent circumstances, the very term "prudential" counsels that it is a proper exercise of the Court's responsibility to take jurisdiction. For these reasons, the prudential and Article III requirements are met here; and, as a consequence, the Court need not decide whether BLAG would have standing to challenge the District Court's ruling and its affirmance in the Court of Appeals on BLAG's own authority.

The Court's conclusion that this petition may be heard on the merits does not imply that no difficulties would ensue if this were a common practice in ordinary cases. The Executive's failure to defend the constitutionality of an Act of Congress based on a constitutional theory not yet established in judicial decisions has created a procedural dilemma. On the one hand, as noted, the Government's agreement with Windsor raises questions about the propriety of entertaining a suit in which it seeks affirmance of an order invalidating a federal law and ordering the United States to pay money. On the other hand, if the Executive's agreement with a plaintiff that a law is unconstitutional is enough to preclude judicial review, then the Supreme Court's primary role in determining the constitutionality of a law that has inflicted real injury on a plaintiff who has brought a justiciable legal claim would become only secondary to the President's. This would undermine the clear dictate of the separation-of-powers principle that "when an Act of Congress is alleged to conflict with the Constitution, `[i]t is emphatically the province and duty of the judicial department to say what the law is.'" Zivotofsky v. Clinton, 566 U. S. ___, ___ (2012) (slip op., at 7) (quoting Marbury v. Madison, 1 Cranch 137, 177 (1803)). Similarly, with respect to the legislative power, when Congress has passed a statute and a President has signed it, it poses grave challenges to the separation of powers for the Executive at a particular moment to be able to nullify Congress' enactment solely on its own initiative and without any determination from the Court.

The Court's jurisdictional holding, it must be underscored, does not mean the arguments for dismissing this dispute on prudential grounds lack substance. Yet the difficulty the Executive faces should be acknowledged. When the Executive makes a principled determination that a statute is unconstitutional, it faces a difficult choice. Still, there is no suggestion here that it is appropriate for the Executive as a matter of course to challenge statutes in the judicial forum rather than making the case to Congress for their amendment or repeal. The integrity of the political process would be at risk if difficult constitutional issues were simply referred to the Court as a routine exercise. But this case is not routine. And the capable defense of the law by BLAG ensures that these prudential issues do not cloud the merits question, which is one of immediate importance to the Federal Government and to hundreds of thousands of persons. These circumstances support the Court's decision to proceed to the merits.

III

When at first Windsor and Spyer longed to marry, neither New York nor any other State granted them that right. After waiting some years, in 2007 they traveled to Ontario to be married there. It seems fair to conclude that, until recent years, many citizens had not even considered the possibility that two persons of the same sex might aspire to occupy the same status and dignity as that of a man and woman in lawful marriage. For marriage between a man and a woman no doubt had been thought of by most people as essential to the very definition of that term and to its role and function throughout the history of civilization. That belief, for many who long have held it, became even more urgent, more cherished when challenged. For others, however, came the beginnings of a new perspective, a new insight. Accordingly some States concluded that same-sex marriage ought to be given recognition and validity in the law for those same-sex couples who wish to define themselves by their commitment to each other. The limitation of lawful marriage to heterosexual couples, which for centuries had been deemed both necessary and fundamental, came to be seen in New York and certain other States as an unjust exclusion.

Slowly at first and then in rapid course, the laws of New York came to acknowledge the urgency of this issue for same-sex couples who wanted to affirm their commitment to one another before their children, their family, their friends, and their community. And so New York recognized same-sex marriages performed elsewhere; and then it later amended its own marriage laws to permit samesex marriage. New York, in common with, as of this writing, 11 other States and the District of Columbia, decided that same-sex couples should have the right to marry and so live with pride in themselves and their union and in a status of equality with all other married persons. After a statewide deliberative process that enabled its citizens to discuss and weigh arguments for and against samesex marriage, New York acted to enlarge the definition of marriage to correct what its citizens and elected representatives perceived to be an injustice that they had not earlier known or understood. See Marriage Equality Act, 2011 N. Y. Laws 749 (codified at N. Y. Dom. Rel. Law Ann. §§10-a, 10-b, 13 (West 2013)).

Against this background of lawful same-sex marriage in some States, the design, purpose, and effect of DOMA should be considered as the beginning point in deciding whether it is valid under the Constitution. By history and tradition the definition and regulation of marriage, as will be discussed in more detail, has been treated as being within the authority and realm of the separate States. Yet it is further established that Congress, in enacting discrete statutes, can make determinations that bear on marital rights and privileges. Just this Term the Court upheld the authority of the Congress to pre-empt state laws, allowing a former spouse to retain life insurance proceeds under a federal program that gave her priority, because of formal beneficiary designation rules, over the wife by a second marriage who survived the husband. Hillman v. Maretta, 569 U. S. ___ (2013); see also Ridgway v. Ridgway, 454 U. S. 46 (1981); Wissner v. Wissner, 338 U. S. 655 (1950). This is one example of the general principle that when the Federal Government acts in the exercise of its own proper authority, it has a wide choice of the mechanisms and means to adopt. See McCulloch v. Maryland, 4 Wheat. 316, 421 (1819). Congress has the power both to ensure efficiency in the administration of its programs and to choose what larger goals and policies to pursue.

Other precedents involving congressional statutes which affect marriages and family status further illustrate this point. In addressing the interaction of state domestic relations and federal immigration law Congress determined that marriages "entered into for the purpose of procuring an alien's admission [to the United States] as an immigrant" will not qualify the noncitizen for that status, even if the noncitizen's marriage is valid and proper for state-law purposes. 8 U. S. C. §1186a(b)(1) (2006 ed. and Supp. V). And in establishing income-based criteria for Social Security benefits, Congress decided that although state law would determine in general who qualifies as an applicant's spouse, common-law marriages also should be recognized, regardless of any particular State's view on these relationships. 42 U. S. C. §1382c(d)(2).

Though these discrete examples establish the constitutionality of limited federal laws that regulate the meaning of marriage in order to further federal policy, DOMA has a far greater reach; for it enacts a directive applicable to over 1,000 federal statutes and the whole realm of federal regulations. And its operation is directed to a class of persons that the laws of New York, and of 11 other States, have sought to protect. See Goodridge v. Department of Public Health, 440 Mass. 309, 798 N. E. 2d 941 (2003); An Act Implementing the Guarantee of Equal Protection Under the Constitution of the State for Same Sex Couples, 2009 Conn. Pub. Acts no. 09-13; Varnum v. Brien, 763 N. W. 2d 862 (Iowa 2009); Vt. Stat. Ann., Tit. 15, §8 (2010); N. H. Rev. Stat. Ann. §457:1-a (West Supp. 2012); Religious Freedom and Civil Marriage Equality Amendment Act of 2009, 57 D. C. Reg. 27 (Dec. 18, 2009); N. Y. Dom. Rel. Law Ann. §10-a (West Supp. 2013); Wash. Rev. Code §26.04.010 (2012); Citizen Initiative, Same-Sex Marriage, Question 1 (Me. 2012) (results online at http://www.maine.gov/sos/cec/elec/2012/tab-ref-2012.html (all Internet sources as visited June 18, 2013, and available in Clerk of Court's case file)); Md. Fam. Law Code Ann. §2-201 (Lexis 2012); An Act to Amend Title 13 of the Delaware Code Relating to Domestic Relations to Provide for Same-Gender Civil Marriage and to Convert Existing Civil Unions to Civil Marriages, 79 Del. Laws ch. 19 (2013); An act relating to marriage; providing for civil marriage between two persons; providing for exemptions and protections based on religious association, 2013 Minn. Laws ch. 74; An Act Relating to Domestic Relations— Persons Eligible to Marry, 2013 R. I. Laws ch. 4.

In order to assess the validity of that intervention it is necessary to discuss the extent of the state power and authority over marriage as a matter of history and tradition. State laws defining and regulating marriage, of course, must respect the constitutional rights of persons, see, e.g., Loving v. Virginia, 388 U. S. 1 (1967); but, subject to those guarantees, "regulation of domestic relations" is "an area that has long been regarded as a virtually exclusive province of the States." Sosna v. Iowa, 419 U. S. 393, 404 (1975).

The recognition of civil marriages is central to state domestic relations law applicable to its residents and citizens. See Williams v. North Carolina, 317 U. S. 287, 298 (1942) ("Each state as a sovereign has a rightful and legitimate concern in the marital status of persons domiciled within its borders"). The definition of marriage is the foundation of the State's broader authority to regulate the subject of domestic relations with respect to the "[p]rotection of offspring, property interests, and the enforcement of marital responsibilities." Ibid. "[T]he states, at the time of the adoption of the Constitution, possessed full power over the subject of marriage and divorce. . . [and] the Constitution delegated no authority to the Government of the United States on the subject of marriage and divorce." Haddock v. Haddock, 201 U. S. 562, 575 (1906); see also In re Burrus, 136 U. S. 586, 593-594 (1890) ("The whole subject of the domestic relations of husband and wife, parent and child, belongs to the laws of the States and not to the laws of the United States").

Consistent with this allocation of authority, the Federal Government, through our history, has deferred to statelaw policy decisions with respect to domestic relations. In De Sylva v. Ballentine, 351 U. S. 570 (1956), for example, the Court held that, "[t]o decide who is the widow or widower of a deceased author, or who are his executors or next of kin," under the Copyright Act "requires a reference to the law of the State which created those legal relationships" because "there is no federal law of domestic relations." Id., at 580. In order to respect this principle, the federal courts, as a general rule, do not adjudicate issues of marital status even when there might otherwise be a basis for federal jurisdiction. See Ankenbrandt v. Richards, 504 U. S. 689, 703 (1992). Federal courts will not hear divorce and custody cases even if they arise in diversity because of "the virtually exclusive primacy . . . of the States in the regulation of domestic relations." Id., at 714 (Blackmun, J., concurring in judgment).

The significance of state responsibilities for the definition and regulation of marriage dates to the Nation's beginning; for "when the Constitution was adopted the common understanding was that the domestic relations of husband and wife and parent and child were matters reserved to the States." Ohio ex rel. Popovici v. Agler, 280 U. S. 379, 383-384 (1930). Marriage laws vary in some respects from State to State. For example, the required minimum age is 16 in Vermont, but only 13 in New Hampshire. Compare Vt. Stat. Ann., Tit. 18, §5142 (2012), with N. H. Rev. Stat. Ann. §457:4 (West Supp. 2012). Likewise the permissible degree of consanguinity can vary (most States permit first cousins to marry, but a handful— such as Iowa and Washington, see Iowa Code §595.19 (2009); Wash. Rev. Code §26.04.020 (2012)—prohibit the practice). But these rules are in every event consistent within each State.

Against this background DOMA rejects the long-established precept that the incidents, benefits, and obligations of marriage are uniform for all married couples within each State, though they may vary, subject to constitutional guarantees, from one State to the next. Despite these considerations, it is unnecessary to decide whether this federal intrusion on state power is a violation of the Constitution because it disrupts the federal balance. The State's power in defining the marital relation is of central relevance in this case quite apart from principles of federalism. Here the State's decision to give this class of persons the right to marry conferred upon them a dignity and status of immense import. When the State used its historic and essential authority to define the marital relation in this way, its role and its power in making the decision enhanced the recognition, dignity, and protection of the class in their own community. DOMA, because of its reach and extent, departs from this history and tradition of reliance on state law to define marriage. "`[D]iscriminations of an unusual character especially suggest careful consideration to determine whether they are obnoxious to the constitutional provision.'" Romer v. Evans, 517 U. S. 620, 633 (1996) (quoting Louisville Gas & Elec. Co. v. Coleman, 277 U. S. 32, 37-38 (1928)).

The Federal Government uses this state-defined class for the opposite purpose—to impose restrictions and disabilities. That result requires this Court now to address whether the resulting injury and indignity is a deprivation of an essential part of the liberty protected by the Fifth Amendment. What the State of New York treats as alike the federal law deems unlike by a law designed to injure the same class the State seeks to protect.

In acting first to recognize and then to allow same-sex marriages, New York was responding "to the initiative of those who [sought] a voice in shaping the destiny of their own times." Bond v. United States, 564 U. S. ___, ___ (2011) (slip op., at 9). These actions were without doubt a proper exercise of its sovereign authority within our federal system, all in the way that the Framers of the Constitution intended. The dynamics of state government in the federal system are to allow the formation of consensus respecting the way the members of a discrete community treat each other in their daily contact and constant interaction with each other.

The States' interest in defining and regulating the marital relation, subject to constitutional guarantees, stems from the understanding that marriage is more than a routine classification for purposes of certain statutory benefits. Private, consensual sexual intimacy between two adult persons of the same sex may not be punished by the State, and it can form "but one element in a personal bond that is more enduring." Lawrence v. Texas, 539 U. S. 558, 567 (2003). By its recognition of the validity of same-sex marriages performed in other jurisdictions and then by authorizing same-sex unions and same-sex marriages, New York sought to give further protection and dignity to that bond. For same-sex couples who wished to be married, the State acted to give their lawful conduct a lawful status. This status is a far-reaching legal acknowledgment of the intimate relationship between two people, a relationship deemed by the State worthy of dignity in the community equal with all other marriages. It reflects both the community's considered perspective on the historical roots of the institution of marriage and its evolving understanding of the meaning of equality.

IV

DOMA seeks to injure the very class New York seeks to protect. By doing so it violates basic due process and equal protection principles applicable to the Federal Government. See U. S. Const., Amdt. 5; Bolling v. Sharpe, 347 U. S. 497 (1954). The Constitution's guarantee of equality "must at the very least mean that a bare congressional desire to harm a politically unpopular group cannot" justify disparate treatment of that group. Department of Agriculture v. Moreno, 413 U. S. 528, 534-535 (1973). In determining whether a law is motived by an improper animus or purpose, "`[d]iscriminations of an unusual character'" especially require careful consideration. Supra, at 19 (quoting Romer, supra, at 633). DOMA cannot survive under these principles. The responsibility of the States for the regulation of domestic relations is an important indicator of the substantial societal impact the State's classifications have in the daily lives and customs of its people. DOMA's unusual deviation from the usual tradition of recognizing and accepting state definitions of marriage here operates to deprive same-sex couples of the benefits and responsibilities that come with the federal recognition of their marriages. This is strong evidence of a law having the purpose and effect of disapproval of that class. The avowed purpose and practical effect of the law here in question are to impose a disadvantage, a separate status, and so a stigma upon all who enter into same-sex marriages made lawful by the unquestioned authority of the States.

The history of DOMA's enactment and its own text demonstrate that interference with the equal dignity of same-sex marriages, a dignity conferred by the States in the exercise of their sovereign power, was more than an incidental effect of the federal statute. It was its essence. The House Report announced its conclusion that "it is both appropriate and necessary for Congress to do what it can to defend the institution of traditional heterosexual marriage.. . . H. R. 3396 is appropriately entitled the `Defense of Marriage Act.' The effort to redefine `marriage' to extend to homosexual couples is a truly radical proposal that would fundamentally alter the institution of marriage." H. R. Rep. No. 104-664, pp. 12-13 (1996). The House concluded that DOMA expresses "both moral disapproval of homosexuality, and a moral conviction that heterosexuality better comports with traditional (especially Judeo-Christian) morality." Id., at 16 (footnote deleted). The stated purpose of the law was to promote an "interest in protecting the traditional moral teachings reflected in heterosexual-only marriage laws." Ibid. Were there any doubt of this far-reaching purpose, the title of the Act confirms it: The Defense of Marriage.

The arguments put forward by BLAG are just as candid about the congressional purpose to influence or interfere with state sovereign choices about who may be married. As the title and dynamics of the bill indicate, its purpose is to discourage enactment of state same-sex marriage laws and to restrict the freedom and choice of couples married under those laws if they are enacted. The congressional goal was "to put a thumb on the scales and influence a state's decision as to how to shape its own marriage laws." Massachusetts, 682 F. 3d, at 12-13. The Act's demonstrated purpose is to ensure that if any State decides to recognize same-sex marriages, those unions will be treated as second-class marriages for purposes of federal law. This raises a most serious question under the Constitution's Fifth Amendment.

DOMA's operation in practice confirms this purpose. When New York adopted a law to permit same-sex marriage, it sought to eliminate inequality; but DOMA frustrates that objective through a system-wide enactment with no identified connection to any particular area of federal law. DOMA writes inequality into the entire United States Code. The particular case at hand concerns the estate tax, but DOMA is more than a simple determination of what should or should not be allowed as an estate tax refund. Among the over 1,000 statutes and numerous federal regulations that DOMA controls are laws pertaining to Social Security, housing, taxes, criminal sanctions, copyright, and veterans' benefits.

DOMA's principal effect is to identify a subset of statesanctioned marriages and make them unequal. The principal purpose is to impose inequality, not for other reasons like governmental efficiency. Responsibilities, as well as rights, enhance the dignity and integrity of the person. And DOMA contrives to deprive some couples married under the laws of their State, but not other couples, of both rights and responsibilities. By creating two contradictory marriage regimes within the same State, DOMA forces same-sex couples to live as married for the purpose of state law but unmarried for the purpose of federal law, thus diminishing the stability and predictability of basic personal relations the State has found it proper to acknowledge and protect. By this dynamic DOMA undermines both the public and private significance of statesanctioned same-sex marriages; for it tells those couples, and all the world, that their otherwise valid marriages are unworthy of federal recognition. This places same-sex couples in an unstable position of being in a second-tier marriage. The differentiation demeans the couple, whose moral and sexual choices the Constitution protects, see Lawrence, 539 U. S. 558, and whose relationship the State has sought to dignify. And it humiliates tens of thousands of children now being raised by same-sex couples. The law in question makes it even more difficult for the children to understand the integrity and closeness of their own family and its concord with other families in their community and in their daily lives.

Under DOMA, same-sex married couples have their lives burdened, by reason of government decree, in visible and public ways. By its great reach, DOMA touches many aspects of married and family life, from the mundane to the profound. It prevents same-sex married couples from obtaining government healthcare benefits they would otherwise receive. See 5 U. S. C. §§8901(5), 8905. It deprives them of the Bankruptcy Code's special protections for domestic-support obligations. See 11 U. S. C. §§101(14A), 507(a)(1)(A), 523(a)(5), 523(a)(15). It forces them to follow a complicated procedure to file their state and federal taxes jointly. Technical Bulletin TB-55, 2010 Vt. Tax LEXIS 6 (Oct. 7, 2010); Brief for Federalism Scholars as Amici Curiae 34. It prohibits them from being buried together in veterans' cemeteries. National Cemetery Administration Directive 3210/1, p. 37 (June 4, 2008).

For certain married couples, DOMA's unequal effects are even more serious. The federal penal code makes it a crime to "assaul[t], kidna[p], or murde[r] . . . a member of the immediate family" of "a United States official, a United States judge, [or] a Federal law enforcement officer," 18 U. S. C. §115(a)(1)(A), with the intent to influence or retaliate against that official, §115(a)(1). Although a "spouse" qualifies as a member of the officer's "immediate family," §115(c)(2), DOMA makes this protection inapplicable to same-sex spouses.

DOMA also brings financial harm to children of samesex couples. It raises the cost of health care for families by taxing health benefits provided by employers to their workers' same-sex spouses. See 26 U. S. C. §106; Treas. Reg. §1.106-1, 26 CFR §1.106-1 (2012); IRS Private Letter Ruling 9850011 (Sept. 10, 1998). And it denies or reduces benefits allowed to families upon the loss of a spouse and parent, benefits that are an integral part of family security. See Social Security Administration, Social Security Survivors Benefits 5 (2012) (benefits available to a surviving spouse caring for the couple's child), online at http://www.ssa.gov/pubs/EN-05-10084.pdf.

DOMA divests married same-sex couples of the duties and responsibilities that are an essential part of married life and that they in most cases would be honored to accept were DOMA not in force. For instance, because it is expected that spouses will support each other as they pursue educational opportunities, federal law takes into consideration a spouse's income in calculating a student's federal financial aid eligibility. See 20 U. S. C. §1087nn(b). Same-sex married couples are exempt from this requirement. The same is true with respect to federal ethics rules. Federal executive and agency officials are prohibited from "participat[ing] personally and substantially" in matters as to which they or their spouses have a financial interest. 18 U. S. C. §208(a). A similar statute prohibits Senators, Senate employees, and their spouses from accepting high-value gifts from certain sources, see 2 U. S. C. §31-2(a)(1), and another mandates detailed financial disclosures by numerous high-ranking officials and their spouses. See 5 U. S. C. App. §§102(a), (e). Under DOMA, however, these Government-integrity rules do not apply to same-sex spouses.

* * *

The power the Constitution grants it also restrains. And though Congress has great authority to design laws to fit its own conception of sound national policy, it cannot deny the liberty protected by the Due Process Clause of the Fifth Amendment.

What has been explained to this point should more than suffice to establish that the principal purpose and the necessary effect of this law are to demean those persons who are in a lawful same-sex marriage. This requires the Court to hold, as it now does, that DOMA is unconstitutional as a deprivation of the liberty of the person protected by the Fifth Amendment of the Constitution.

The liberty protected by the Fifth Amendment's Due Process Clause contains within it the prohibition against denying to any person the equal protection of the laws. See Bolling, 347 U. S., at 499-500; Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 217-218 (1995). While the Fifth Amendment itself withdraws from Government the power to degrade or demean in the way this law does, the equal protection guarantee of the Fourteenth Amendment makes that Fifth Amendment right all the more specific and all the better understood and preserved.

The class to which DOMA directs its restrictions and restraints are those persons who are joined in same-sex marriages made lawful by the State. DOMA singles out a class of persons deemed by a State entitled to recognition and protection to enhance their own liberty. It imposes a disability on the class by refusing to acknowledge a status the State finds to be dignified and proper. DOMA instructs all federal officials, and indeed all persons with whom same-sex couples interact, including their own children, that their marriage is less worthy than the marriages of others. The federal statute is invalid, for no legitimate purpose overcomes the purpose and effect to disparage and to injure those whom the State, by its marriage laws, sought to protect in personhood and dignity. By seeking to displace this protection and treating those persons as living in marriages less respected than others, the federal statute is in violation of the Fifth Amendment. This opinion and its holding are confined to those lawful marriages.

The judgment of the Court of Appeals for the Second Circuit is affirmed.

It is so ordered.

CHIEF JUSTICE ROBERTS, dissenting.

I agree with JUSTICE SCALIA that this Court lacks jurisdiction to review the decisions of the courts below. On the merits of the constitutional dispute the Court decides to decide, I also agree with JUSTICE SCALIA that Congress acted constitutionally in passing the Defense of Marriage Act (DOMA). Interests in uniformity and stability amply justified Congress's decision to retain the definition of marriage that, at that point, had been adopted by every State in our Nation, and every nation in the world. Post, at 19-20 (dissenting opinion).

The majority sees a more sinister motive, pointing out that the Federal Government has generally (though not uniformly) deferred to state definitions of marriage in the past. That is true, of course, but none of those prior stateby-state variations had involved differences over something—as the majority puts it—"thought of by most people as essential to the very definition of [marriage] and to its role and function throughout the history of civilization." Ante, at 13. That the Federal Government treated this fundamental question differently than it treated variations over consanguinity or minimum age is hardly surprising— and hardly enough to support a conclusion that the "principal purpose," ante, at 22, of the 342 Representatives and 85 Senators who voted for it, and the President who signed it, was a bare desire to harm. Nor do the snippets of legislative history and the banal title of the Act to which the majority points suffice to make such a showing. At least without some more convincing evidence that the Act's principal purpose was to codify malice, and that it furthered no legitimate government interests, I would not tar the political branches with the brush of bigotry.

But while I disagree with the result to which the majority's analysis leads it in this case, I think it more important to point out that its analysis leads no further. The Court does not have before it, and the logic of its opinion does not decide, the distinct question whether the States, in the exercise of their "historic and essential authority to define the marital relation," ante, at 18, may continue to utilize the traditional definition of marriage.

The majority goes out of its way to make this explicit in the penultimate sentence of its opinion. It states that "[t]his opinion and its holding are confined to those lawful marriages," ante, at 26—referring to same-sex marriages that a State has already recognized as a result of the local "community's considered perspective on the historical roots of the institution of marriage and its evolving understanding of the meaning of equality." Ante, at 20. JUSTICE SCALIA believes this is a "`bald, unreasoned disclaime[r].'" Post, at 22. In my view, though, the disclaimer is a logical and necessary consequence of the argument the majority has chosen to adopt. The dominant theme of the majority opinion is that the Federal Government's intrusion into an area "central to state domestic relations law applicable to its residents and citizens" is sufficiently "unusual" to set off alarm bells. Ante, at 17, 20. I think the majority goes off course, as I have said, but it is undeniable that its judgment is based on federalism.

The majority extensively chronicles DOMA's departure from the normal allocation of responsibility between State and Federal Governments, emphasizing that DOMA "rejects the long-established precept that the incidents, benefits, and obligations of marriage are uniform for all married couples within each State." Ante, at 18. But there is no such departure when one State adopts or keeps a definition of marriage that differs from that of its neighbor, for it is entirely expected that state definitions would "vary, subject to constitutional guarantees, from one State to the next." Ibid. Thus, while "[t]he State's power in defining the marital relation is of central relevance" to the majority's decision to strike down DOMA here, ibid., that power will come into play on the other side of the board in future cases about the constitutionality of state marriage definitions. So too will the concerns for state diversity and sovereignty that weigh against DOMA's constitutionality in this case. See ante, at 19.

It is not just this central feature of the majority's analysis that is unique to DOMA, but many considerations on the periphery as well. For example, the majority focuses on the legislative history and title of this particular Act, ante, at 21; those statute-specific considerations will, of course, be irrelevant in future cases about different statutes. The majority emphasizes that DOMA was a "systemwide enactment with no identified connection to any particular area of federal law," but a State's definition of marriage "is the foundation of the State's broader authority to regulate the subject of domestic relations with respect to the `[p]rotection of offspring, property interests, and the enforcement of marital responsibilities.'" Ante, at 22, 17. And the federal decision undermined (in the majority's view) the "dignity [already] conferred by the States in the exercise of their sovereign power," ante, at 21, whereas a State's decision whether to expand the definition of marriage from its traditional contours involves no similar concern.

We may in the future have to resolve challenges to state marriage definitions affecting same-sex couples. That issue, however, is not before us in this case, and we hold today that we lack jurisdiction to consider it in the particular context of Hollingsworth v. Perry, ante, p. ___. I write only to highlight the limits of the majority's holding and reasoning today, lest its opinion be taken to resolve not only a question that I believe is not properly before us—DOMA's constitutionality—but also a question that all agree, and the Court explicitly acknowledges, is not at issue.

JUSTICE SCALIA, with whom JUSTICE THOMAS joins, and with whom THE CHIEF JUSTICE joins as to Part I, dissenting.

This case is about power in several respects. It is about the power of our people to govern themselves, and the power of this Court to pronounce the law. Today's opinion aggrandizes the latter, with the predictable consequence of diminishing the former. We have no power to decide this case. And even if we did, we have no power under the Constitution to invalidate this democratically adopted legislation. The Court's errors on both points spring forth from the same diseased root: an exalted conception of the role of this institution in America.

I

A

The Court is eager—hungry—to tell everyone its view of the legal question at the heart of this case. Standing in the way is an obstacle, a technicality of little interest to anyone but the people of We the People, who created it as a barrier against judges' intrusion into their lives. They gave judges, in Article III, only the "judicial Power," a power to decide not abstract questions but real, concrete "Cases" and "Controversies." Yet the plaintiff and the Government agree entirely on what should happen in this lawsuit. They agree that the court below got it right; and they agreed in the court below that the court below that one got it right as well. What, then, are we doing here?

The answer lies at the heart of the jurisdictional portion of today's opinion, where a single sentence lays bare the majority's vision of our role. The Court says that we have the power to decide this case because if we did not, then our "primary role in determining the constitutionality of a law" (at least one that "has inflicted real injury on a plaintiff ") would "become only secondary to the President's." Ante, at 12. But wait, the reader wonders—Windsor won below, and so cured her injury, and the President was glad to see it. True, says the majority, but judicial review must march on regardless, lest we "undermine the clear dictate of the separation-of-powers principle that when an Act of Congress is alleged to conflict with the Constitution, it is emphatically the province and duty of the judicial department to say what the law is." Ibid. (internal quotation marks and brackets omitted).

That is jaw-dropping. It is an assertion of judicial supremacy over the people's Representatives in Congress and the Executive. It envisions a Supreme Court standing (or rather enthroned) at the apex of government, empowered to decide all constitutional questions, always and everywhere "primary" in its role.

This image of the Court would have been unrecognizable to those who wrote and ratified our national charter. They knew well the dangers of "primary" power, and so created branches of government that would be "perfectly coordinate by the terms of their common commission," none of which branches could "pretend to an exclusive or superior right of settling the boundaries between their respective powers." The Federalist, No. 49, p. 314 (C. Rossiter ed. 1961) (J. Madison). The people did this to protect themselves. They did it to guard their right to self-rule against the black-robed supremacy that today's majority finds so attractive. So it was that Madison could confidently state, with no fear of contradiction, that there was nothing of "greater intrinsic value" or "stamped with the authority of more enlightened patrons of liberty" than a government of separate and coordinate powers. Id., No. 47, at 301.

For this reason we are quite forbidden to say what the law is whenever (as today's opinion asserts) "`an Act of Congress is alleged to conflict with the Constitution.'" Ante, at 12. We can do so only when that allegation will determine the outcome of a lawsuit, and is contradicted by the other party. The "judicial Power" is not, as the majority believes, the power "`to say what the law is,'" ibid., giving the Supreme Court the "primary role in determining the constitutionality of laws." The majority must have in mind one of the foreign constitutions that pronounces such primacy for its constitutional court and allows that primacy to be exercised in contexts other than a lawsuit. See, e.g., Basic Law for the Federal Republic of Germany, Art. 93. The judicial power as Americans have understood it (and their English ancestors before them) is the power to adjudicate, with conclusive effect, disputed government claims (civil or criminal) against private persons, and disputed claims by private persons against the government or other private persons. Sometimes (though not always) the parties before the court disagree not with regard to the facts of their case (or not only with regard to the facts) but with regard to the applicable law—in which event (and only in which event) it becomes the "`province and duty of the judicial department to say what the law is.'" Ante, at 12.

In other words, declaring the compatibility of state or federal laws with the Constitution is not only not the "primary role" of this Court, it is not a separate, freestanding role at all. We perform that role incidentally—by accident, as it were—when that is necessary to resolve the dispute before us. Then, and only then, does it become "`the province and duty of the judicial department to say what the law is.'" That is why, in 1793, we politely declined the Washington Administration's request to "say what the law is" on a particular treaty matter that was not the subject of a concrete legal controversy. 3 Correspondence and Public Papers of John Jay 486-489 (H. Johnston ed. 1893). And that is why, as our opinions have said, some questions of law will never be presented to this Court, because there will never be anyone with standing to bring a lawsuit. See Schlesinger v. Reservists Comm. to Stop the War, 418 U. S. 208, 227 (1974); United States v. Richardson, 418 U. S. 166, 179 (1974). As Justice Brandeis put it, we cannot "pass upon the constitutionality of legislation in a friendly, non-adversary, proceeding"; absent a "`real, earnest and vital controversy between individuals,'" we have neither any work to do nor any power to do it. Ashwander v. TVA, 297 U. S. 288, 346 (1936) (concurring opinion) (quoting Chicago & Grand Trunk R. Co. v. Wellman, 143 U. S. 339, 345 (1892)). Our authority begins and ends with the need to adjudge the rights of an injured party who stands before us seeking redress. Lujan v. Defenders of Wildlife, 504 U. S. 555, 560 (1992).

That is completely absent here. Windsor's injury was cured by the judgment in her favor. And while, in ordinary circumstances, the United States is injured by a directive to pay a tax refund, this suit is far from ordinary. Whatever injury the United States has suffered will surely not be redressed by the action that it, as a litigant, asks us to take. The final sentence of the Solicitor General's brief on the merits reads: "For the foregoing reasons, the judgment of the court of appeals should be affirmed." Brief for United States (merits) 54 (emphasis added). That will not cure the Government's injury, but carve it into stone. One could spend many fruitless afternoons ransacking our library for any other petitioner's brief seeking an affirmance of the judgment against it.^[1] What the petitioner United States asks us to do in the case before us is exactly what the respondent Windsor asks us to do: not to provide relief from the judgment below but to say that that judgment was correct. And the same was true in the Court of Appeals: Neither party sought to undo the judgment for Windsor, and so that court should have dismissed the appeal (just as we should dismiss) for lack of jurisdiction. Since both parties agreed with the judgment of the District Court for the Southern District of New York, the suit should have ended there. The further proceedings have been a contrivance, having no object in mind except to elevate a District Court judgment that has no precedential effect in other courts, to one that has precedential effect throughout the Second Circuit, and then (in this Court) precedential effect throughout the United States.

We have never before agreed to speak—to "say what the law is"—where there is no controversy before us. In the more than two centuries that this Court has existed as an institution, we have never suggested that we have the power to decide a question when every party agrees with both its nominal opponent and the court below on that question's answer. The United States reluctantly conceded that at oral argument. See Tr. of Oral Arg. 19-20.

The closest we have ever come to what the Court blesses today was our opinion in INS v. Chadha, 462 U. S. 919 (1983). But in that case, two parties to the litigation disagreed with the position of the United States and with the court below: the House and Senate, which had intervened in the case. Because Chadha concerned the validity of a mode of congressional action—the one-house legislative veto—the House and Senate were threatened with destruction of what they claimed to be one of their institutional powers. The Executive choosing not to defend that power,^[2] we permitted the House and Senate to intervene. Nothing like that is present here.

To be sure, the Court in Chadha said that statutory aggrieved-party status was "not altered by the fact that the Executive may agree with the holding that the statute in question is unconstitutional." Id., at 930-931. But in a footnote to that statement, the Court acknowledged Article III's separate requirement of a "justiciable case or controversy," and stated that this requirement was satisfied "because of the presence of the two Houses of Congress as adverse parties." Id., at 931, n. 6. Later in its opinion, the Chadha Court remarked that the United States' announced intention to enforce the statute also sufficed to permit judicial review, even absent congressional participation. Id., at 939. That remark is true, as a description of the judicial review conducted in the Court of Appeals, where the Houses of Congress had not intervened. (The case originated in the Court of Appeals, since it sought review of agency action under 8 U. S. C. §1105a(a) (1976 ed.).) There, absent a judgment setting aside the INS order, Chadha faced deportation. This passage of our opinion seems to be addressing that initial standing in the Court of Appeals, as indicated by its quotation from the lower court's opinion, 462 U. S., at 939-940. But if it was addressing standing to pursue the appeal, the remark was both the purest dictum (as congressional intervention at that point made the required adverseness "beyond doubt," id., at 939), and quite incorrect. When a private party has a judicial decree safely in hand to prevent his injury, additional judicial action requires that a party injured by the decree seek to undo it. In Chadha, the intervening House and Senate fulfilled that requirement. Here no one does.

The majority's discussion of the requirements of Article III bears no resemblance to our jurisprudence. It accuses the amicus (appointed to argue against our jurisdiction) of "elid[ing] the distinction between . . . the jurisdictional requirements of Article III and the prudential limits on its exercise." Ante, at 6. It then proceeds to call the requirement of adverseness a "prudential" aspect of standing. Of standing. That is incomprehensible. A plaintiff (or appellant) can have all the standing in the world—satisfying all three standing requirements of Lujan that the majority so carefully quotes, ante, at 7—and yet no Article III controversy may be before the court. Article III requires not just a plaintiff (or appellant) who has standing to complain but an opposing party who denies the validity of the complaint. It is not the amicus that has done the eliding of distinctions, but the majority, calling the quite separate Article III requirement of adverseness between the parties an element (which it then pronounces a "prudential" element) of standing. The question here is not whether, as the majority puts it, "the United States retains a stake sufficient to support Article III jurisdiction," ibid. the question is whether there is any controversy (which requires contradiction) between the United States and Ms. Windsor. There is not.

I find it wryly amusing that the majority seeks to dismiss the requirement of party-adverseness as nothing more than a "prudential" aspect of the sole Article III requirement of standing. (Relegating a jurisdictional requirement to "prudential" status is a wondrous device, enabling courts to ignore the requirement whenever they believe it "prudent"—which is to say, a good idea.) Half a century ago, a Court similarly bent upon announcing its view regarding the constitutionality of a federal statute achieved that goal by effecting a remarkably similar but completely opposite distortion of the principles limiting our jurisdiction. The Court's notorious opinion in Flast v. Cohen, 392 U. S. 83, 98-101 (1968), held that standing was merely an element (which it pronounced to be a "prudential" element) of the sole Article III requirement of adverseness. We have been living with the chaos created by that power-grabbing decision ever since, see Hein v. Freedom From Religion Foundation, Inc., 551 U. S. 587 (2007), as we will have to live with the chaos created by this one.

The authorities the majority cites fall miles short of supporting the counterintuitive notion that an Article III "controversy" can exist without disagreement between the parties. In Deposit Guaranty Nat. Bank v. Roper, 445 U. S. 326 (1980), the District Court had entered judgment in the individual plaintiff's favor based on the defendant bank's offer to pay the full amount claimed. The plaintiff, however, sought to appeal the District Court's denial of class certification under Federal Rule of Civil Procedure 23. There was a continuing dispute between the parties concerning the issue raised on appeal. The same is true of the other case cited by the majority, Camreta v. Greene, 563 U. S. ___ (2011). There the District Court found that the defendant state officers had violated the Fourth Amendment, but rendered judgment in their favor because they were entitled to official immunity, application of the Fourth Amendment to their conduct not having been clear at the time of violation. The officers sought to appeal the holding of Fourth Amendment violation, which would circumscribe their future conduct; the plaintiff continued to insist that a Fourth Amendment violation had occurred. The "prudential" discretion to which both those cases refer was the discretion to deny an appeal even when a live controversy exists—not the discretion to grant one when it does not. The majority can cite no case in which this Court entertained an appeal in which both parties urged us to affirm the judgment below. And that is because the existence of a controversy is not a "prudential" requirement that we have invented, but an essential element of an Article III case or controversy. The majority's notion that a case between friendly parties can be entertained so long as "adversarial presentation of the issues is assured by the participation of amici curiae prepared to defend with vigor" the other side of the issue, ante, at 10, effects a breathtaking revolution in our Article III jurisprudence.

It may be argued that if what we say is true some Presidential determinations that statutes are unconstitutional will not be subject to our review. That is as it should be, when both the President and the plaintiff agree that the statute is unconstitutional. Where the Executive is enforcing an unconstitutional law, suit will of course lie; but if, in that suit, the Executive admits the unconstitutionality of the law, the litigation should end in an order or a consent decree enjoining enforcement. This suit saw the light of day only because the President enforced the Act (and thus gave Windsor standing to sue) even though he believed it unconstitutional. He could have equally chosen (more appropriately, some would say) neither to enforce nor to defend the statute he believed to be unconstitutional, see Presidential Authority to Decline to Execute Unconstitutional Statutes, 18 Op. Off. Legal Counsel 199 (Nov. 2, 1994)—in which event Windsor would not have been injured, the District Court could not have refereed this friendly scrimmage, and the Executive's determination of unconstitutionality would have escaped this Court's desire to blurt out its view of the law. The matter would have been left, as so many matters ought to be left, to a tug of war between the President and the Congress, which has innumerable means (up to and including impeachment) of compelling the President to enforce the laws it has written. Or the President could have evaded presentation of the constitutional issue to this Court simply by declining to appeal the District Court and Court of Appeals dispositions he agreed with. Be sure of this much: If a President wants to insulate his judgment of unconstitutionality from our review, he can. What the views urged in this dissent produce is not insulation from judicial review but insulation from Executive contrivance.

The majority brandishes the famous sentence from Marbury v. Madison, 1 Cranch 137, 177 (1803) that "[i]t is emphatically the province and duty of the judicial department to say what the law is." Ante, at 12 (internal quotation marks omitted). But that sentence neither says nor implies that it is always the province and duty of the Court to say what the law is—much less that its responsibility in that regard is a "primary" one. The very next sentence of Chief Justice Marshall's opinion makes the crucial qualification that today's majority ignores: "Those who apply the rule to particular cases, must of necessity expound and interpret that rule." 1 Cranch, at 177 (emphasis added). Only when a "particular case" is before us—that is, a controversy that it is our business to resolve under Article III—do we have the province and duty to pronounce the law. For the views of our early Court more precisely addressing the question before us here, the majority ought instead to have consulted the opinion of Chief Justice Taney in Lord v. Veazie, 8 How. 251 (1850):

"The objection in the case before us is . . . that the plaintiff and defendant have the same interest, and that interest adverse and in conflict with the interest of third persons, whose rights would be seriously affected if the question of law was decided in the manner that both of the parties to this suit desire it to be.

"A judgment entered under such circumstances, and for such purposes, is a mere form. The whole proceeding was in contempt of the court, and highly reprehensible. . . . A judgment in form, thus procured, in the eye of the law is no judgment of the court. It is a nullity, and no writ of error will lie upon it. This writ is, therefore, dismissed." Id., at 255-256.

There is, in the words of Marbury, no "necessity [to] expound and interpret" the law in this case; just a desire to place this Court at the center of the Nation's life. 1 Cranch, at 177.

B

A few words in response to the theory of jurisdiction set forth in JUSTICE ALITO's dissent: Though less far reaching in its consequences than the majority's conversion of constitutionally required adverseness into a discretionary element of standing, the theory of that dissent similarly elevates the Court to the "primary" determiner of constitutional questions involving the separation of powers, and, to boot, increases the power of the most dangerous branch: the "legislative department," which by its nature "draw[s] all power into its impetuous vortex." The Federalist, No. 48, at 309 (J. Madison). Heretofore in our national history, the President's failure to "take Care that the Laws be faithfully executed," U. S. Const., Art. II, §3, could only be brought before a judicial tribunal by someone whose concrete interests were harmed by that alleged failure. JUSTICE ALITO would create a system in which Congress can hale the Executive before the courts not only to vindicate its own institutional powers to act, but to correct a perceived inadequacy in the execution of its laws.^[3] This would lay to rest Tocqueville's praise of our judicial system as one which "intimately bind[s] the case made for the law with the case made for one man," one in which legislation is "no longer exposed to the daily aggression of the parties," and in which "[t]he political question that [the judge] must resolve is linked to the interest" of private litigants. A. de Tocqueville, Democracy in America 97 (H. Mansfield & D. Winthrop eds. 2000). That would be replaced by a system in which Congress and the Executive can pop immediately into court, in their institutional capacity, whenever the President refuses to implement a statute he believes to be unconstitutional, and whenever he implements a law in a manner that is not to Congress's liking.

JUSTICE ALITO's notion of standing will likewise enormously shrink the area to which "judicial censure, exercised by the courts on legislation, cannot extend," ibid. For example, a bare majority of both Houses could bring into court the assertion that the Executive's implementation of welfare programs is too generous—a failure that no other litigant would have standing to complain about. Moreover, as we indicated in Raines v. Byrd, 521 U. S. 811, 828 (1997), if Congress can sue the Executive for the erroneous application of the law that "injures" its power to legislate, surely the Executive can sue Congress for its erroneous adoption of an unconstitutional law that "injures" the Executive's power to administer—or perhaps for its protracted failure to act on one of his nominations. The opportunities for dragging the courts into disputes hitherto left for political resolution are endless.

JUSTICE ALITO's dissent is correct that Raines did not formally decide this issue, but its reasoning does. The opinion spends three pages discussing famous, decadeslong disputes between the President and Congress— regarding congressional power to forbid the Presidential removal of executive officers, regarding the legislative veto, regarding congressional appointment of executive officers, and regarding the pocket veto—that would surely have been promptly resolved by a Congress-vs.-the-President lawsuit if the impairment of a branch's powers alone conferred standing to commence litigation. But it does not, and never has; the "enormous power that the judiciary would acquire" from the ability to adjudicate such suits "would have made a mockery of [Hamilton's] quotation of Montesquieu to the effect that `of the three powers above mentioned . . . the JUDICIARY is next to nothing.'" Barnes v. Kline, 759 F. 2d 21, 58 (CADC 1985) (Bork, J., dissenting) (quoting The Federalist No. 78 (A. Hamilton)).

To be sure, if Congress cannot invoke our authority in the way that JUSTICE ALITO proposes, then its only recourse is to confront the President directly. Unimaginable evil this is not. Our system is designed for confrontation. That is what "[a]mbition . . . counteract[ing] ambition," The Federalist, No. 51, at 322 (J. Madison), is all about. If majorities in both Houses of Congress care enough about the matter, they have available innumerable ways to compel executive action without a lawsuit—from refusing to confirm Presidential appointees to the elimination of funding. (Nothing says "enforce the Act" quite like ". . . or you will have money for little else.") But the condition is crucial; Congress must care enough to act against the President itself, not merely enough to instruct its lawyers to ask us to do so. Placing the Constitution's entirely anticipated political arm wrestling into permanent judicial receivership does not do the system a favor. And by the way, if the President loses the lawsuit but does not faithfully implement the Court's decree, just as he did not faithfully implement Congress's statute, what then? Only Congress can bring him to heel by . . . what do you think? Yes: a direct confrontation with the President.

II

For the reasons above, I think that this Court has, and the Court of Appeals had, no power to decide this suit. We should vacate the decision below and remand to the Court of Appeals for the Second Circuit, with instructions to dismiss the appeal. Given that the majority has volunteered its view of the merits, however, I proceed to discuss that as well.

A

There are many remarkable things about the majority's merits holding. The first is how rootless and shifting its justifications are. For example, the opinion starts with seven full pages about the traditional power of States to define domestic relations—initially fooling many readers, I am sure, into thinking that this is a federalism opinion. But we are eventually told that "it is unnecessary to decide whether this federal intrusion on state power is a violation of the Constitution," and that "[t]he State's power in defining the marital relation is of central relevance in this case quite apart from principles of federalism" because "the State's decision to give this class of persons the right to marry conferred upon them a dignity and status of immense import." Ante, at 18. But no one questions the power of the States to define marriage (with the concomitant conferral of dignity and status), so what is the point of devoting seven pages to describing how long and well established that power is? Even after the opinion has formally disclaimed reliance upon principles of federalism, mentions of "the usual tradition of recognizing and accepting state definitions of marriage" continue. See, e.g., ante, at 20. What to make of this? The opinion never explains. My guess is that the majority, while reluctant to suggest that defining the meaning of "marriage" in federal statutes is unsupported by any of the Federal Government's enumerated powers,^[4] nonetheless needs some rhetorical basis to support its pretense that today's prohibition of laws excluding same-sex marriage is confined to the Federal Government (leaving the second, state-law shoe to be dropped later, maybe next Term). But I am only guessing.

Equally perplexing are the opinion's references to "the Constitution's guarantee of equality." Ibid. Near the end of the opinion, we are told that although the "equal protection guarantee of the Fourteenth Amendment makes [the] Fifth Amendment [due process] right all the more specific and all the better understood and preserved"—what can that mean?—"the Fifth Amendment itself withdraws from Government the power to degrade or demean in the way this law does." Ante, at 25. The only possible interpretation of this statement is that the Equal Protection Clause, even the Equal Protection Clause as incorporated in the Due Process Clause, is not the basis for today's holding. But the portion of the majority opinion that explains why DOMA is unconstitutional (Part IV) begins by citing Bolling v. Sharpe, 347 U. S. 497 (1954), Department of Agriculture v. Moreno, 413 U. S. 528 (1973), and Romer v. Evans, 517 U. S. 620 (1996)—all of which are equalprotection cases.^[5] And those three cases are the only authorities that the Court cites in Part IV about the Constitution's meaning, except for its citation of Lawrence v. Texas, 539 U. S. 558 (2003) (not an equal-protection case) to support its passing assertion that the Constitution protects the "moral and sexual choices" of same-sex couples, ante, at 23.

Moreover, if this is meant to be an equal-protection opinion, it is a confusing one. The opinion does not resolve and indeed does not even mention what had been the central question in this litigation: whether, under the Equal Protection Clause, laws restricting marriage to a man and a woman are reviewed for more than mere rationality. That is the issue that divided the parties and the court below, compare Brief for Respondent Bipartisan Legal Advisory Group of U. S. House of Representatives (merits) 24-28 (no), with Brief for Respondent Windsor (merits) 17-31 and Brief for United States (merits) 18-36 (yes); and compare 699 F. 3d 169, 180-185 (CA2 2012) (yes), with id., at 208-211 (Straub, J., dissenting in part and concurring in part) (no). In accord with my previously expressed skepticism about the Court's "tiers of scrutiny" approach, I would review this classification only for its rationality. See United States v. Virginia, 518 U. S. 515, 567-570 (1996) (SCALIA, J., dissenting). As nearly as I can tell, the Court agrees with that; its opinion does not apply strict scrutiny, and its central propositions are taken from rational-basis cases like Moreno. But the Court certainly does not apply anything that resembles that deferential framework. See Heller v. Doe, 509 U. S. 312, 320 (1993) (a classification "`must be upheld . . . if there is any reasonably conceivable state of facts'" that could justify it).

The majority opinion need not get into the strict-vs.rational-basis scrutiny question, and need not justify its holding under either, because it says that DOMA is unconstitutional as "a deprivation of the liberty of the person protected by the Fifth Amendment of the Constitution," ante, at 25; that it violates "basic due process" principles, ante, at 20; and that it inflicts an "injury and indignity" of a kind that denies "an essential part of the liberty protected by the Fifth Amendment," ante, at 19. The majority never utters the dread words "substantive due process," perhaps sensing the disrepute into which that doctrine has fallen, but that is what those statements mean. Yet the opinion does not argue that same-sex marriage is "deeply rooted in this Nation's history and tradition," Washington v. Glucksberg, 521 U. S. 702, 720-721 (1997), a claim that would of course be quite absurd. So would the further suggestion (also necessary, under our substantive-due-process precedents) that a world in which DOMA exists is one bereft of "`ordered liberty.'" Id., at 721 (quoting Palko v. Connecticut, 302 U. S. 319, 325 (1937)).

Some might conclude that this loaf could have used a while longer in the oven. But that would be wrong; it is already overcooked. The most expert care in preparation cannot redeem a bad recipe. The sum of all the Court's nonspecific hand-waving is that this law is invalid (maybe on equal-protection grounds, maybe on substantive-dueprocess grounds, and perhaps with some amorphous federalism component playing a role) because it is motivated by a "`bare . . . desire to harm'" couples in same-sex marriages. Ante, at 20. It is this proposition with which I will therefore engage.

B

As I have observed before, the Constitution does not forbid the government to enforce traditional moral and sexual norms. See Lawrence v. Texas, 539 U. S. 558, 599 (2003) (SCALIA, J., dissenting). I will not swell the U. S. Reports with restatements of that point. It is enough to say that the Constitution neither requires nor forbids our society to approve of same-sex marriage, much as it neither requires nor forbids us to approve of no-fault divorce, polygamy, or the consumption of alcohol.

However, even setting aside traditional moral disapproval of same-sex marriage (or indeed same-sex sex), there are many perfectly valid—indeed, downright boring—justifying rationales for this legislation. Their existence ought to be the end of this case. For they give the lie to the Court's conclusion that only those with hateful hearts could have voted "aye" on this Act. And more importantly, they serve to make the contents of the legislators' hearts quite irrelevant: "It is a familiar principle of constitutional law that this Court will not strike down an otherwise constitutional statute on the basis of an alleged illicit legislative motive." United States v. O'Brien, 391 U. S. 367, 383 (1968). Or at least it was a familiar principle. By holding to the contrary, the majority has declared open season on any law that (in the opinion of the law's opponents and any panel of like-minded federal judges) can be characterized as mean-spirited.

The majority concludes that the only motive for this Act was the "bare . . . desire to harm a politically unpopular group." Ante, at 20. Bear in mind that the object of this condemnation is not the legislature of some once-Confederate Southern state (familiar objects of the Court's scorn, see, e.g., Edwards v. Aguillard, 482 U. S. 578 (1987)), but our respected coordinate branches, the Congress and Presidency of the United States. Laying such a charge against them should require the most extraordinary evidence, and I would have thought that every attempt would be made to indulge a more anodyne explanation for the statute. The majority does the opposite— affirmatively concealing from the reader the arguments that exist in justification. It makes only a passing mention of the "arguments put forward" by the Act's defenders, and does not even trouble to paraphrase or describe them. See ante, at 21. I imagine that this is because it is harder to maintain the illusion of the Act's supporters as unhinged members of a wild-eyed lynch mob when one first describes their views as they see them.

To choose just one of these defenders' arguments, DOMA avoids difficult choice-of-law issues that will now arise absent a uniform federal definition of marriage. See, e.g., Baude, Beyond DOMA: Choice of State Law in Federal Statutes, 64 Stan. L. Rev. 1371 (2012). Imagine a pair of women who marry in Albany and then move to Alabama, which does not "recognize as valid any marriage of parties of the same sex." Ala. Code §30-1-19(e) (2011). When the couple files their next federal tax return, may it be a joint one? Which State's law controls, for federal-law purposes: their State of celebration (which recognizes the marriage) or their State of domicile (which does not)? (Does the answer depend on whether they were just visiting in Albany?) Are these questions to be answered as a matter of federal common law, or perhaps by borrowing a State's choice-of-law rules? If so, which State's? And what about States where the status of an out-of-state same-sex marriage is an unsettled question under local law? See Godfrey v. Spano, 13 N. Y. 3d 358, 920 N. E. 2d 328 (2009). DOMA avoided all of this uncertainty by specifying which marriages would be recognized for federal purposes. That is a classic purpose for a definitional provision.

Further, DOMA preserves the intended effects of prior legislation against then-unforeseen changes in circumstance. When Congress provided (for example) that a special estate-tax exemption would exist for spouses, this exemption reached only opposite-sex spouses—those being the only sort that were recognized in any State at the time of DOMA's passage. When it became clear that changes in state law might one day alter that balance, DOMA's definitional section was enacted to ensure that state-level experimentation did not automatically alter the basic operation of federal law, unless and until Congress made the further judgment to do so on its own. That is not animus—just stabilizing prudence. Congress has hardly demonstrated itself unwilling to make such further, revising judgments upon due deliberation. See, e.g., Don't Ask, Don't Tell Repeal Act of 2010, 124 Stat. 3515.

The Court mentions none of this. Instead, it accuses the Congress that enacted this law and the President who signed it of something much worse than, for example, having acted in excess of enumerated federal powers—or even having drawn distinctions that prove to be irrational. Those legal errors may be made in good faith, errors though they are. But the majority says that the supporters of this Act acted with malice—with the "purpose" (ante, at 25) "to disparage and to injure" same-sex couples. It says that the motivation for DOMA was to "demean," ibid.; to "impose inequality," ante, at 22; to "impose . . . a stigma," ante, at 21; to deny people "equal dignity," ibid.; to brand gay people as "unworthy," ante, at 23; and to "humiliat[e]" their children, ibid. (emphasis added).

I am sure these accusations are quite untrue. To be sure (as the majority points out), the legislation is called the Defense of Marriage Act. But to defend traditional marriage is not to condemn, demean, or humiliate those who would prefer other arrangements, any more than to defend the Constitution of the United States is to condemn, demean, or humiliate other constitutions. To hurl such accusations so casually demeans this institution. In the majority's judgment, any resistance to its holding is beyond the pale of reasoned disagreement. To question its high-handed invalidation of a presumptively valid statute is to act (the majority is sure) with the purpose to "disparage," "injure," "degrade," "demean," and "humiliate" our fellow human beings, our fellow citizens, who are homosexual. All that, simply for supporting an Act that did no more than codify an aspect of marriage that had been unquestioned in our society for most of its existence— indeed, had been unquestioned in virtually all societies for virtually all of human history. It is one thing for a society to elect change; it is another for a court of law to impose change by adjudging those who oppose it hostes humani generis, enemies of the human race.

* * *

The penultimate sentence of the majority's opinion is a naked declaration that "[t]his opinion and its holding are confined" to those couples "joined in same-sex marriages made lawful by the State." Ante, at 26, 25. I have heard such "bald, unreasoned disclaimer[s]" before. Lawrence, 539 U. S., at 604. When the Court declared a constitutional right to homosexual sodomy, we were assured that the case had nothing, nothing at all to do with "whether the government must give formal recognition to any relationship that homosexual persons seek to enter." Id., at 578. Now we are told that DOMA is invalid because it "demeans the couple, whose moral and sexual choices the Constitution protects," ante, at 23—with an accompanying citation of Lawrence. It takes real cheek for today's majority to assure us, as it is going out the door, that a constitutional requirement to give formal recognition to same-sex marriage is not at issue here—when what has preceded that assurance is a lecture on how superior the majority's moral judgment in favor of same-sex marriage is to the Congress's hateful moral judgment against it. I promise you this: The only thing that will "confine" the Court's holding is its sense of what it can get away with.

I do not mean to suggest disagreement with THE CHIEF JUSTICE's view, ante, p. 2-4 (dissenting opinion), that lower federal courts and state courts can distinguish today's case when the issue before them is state denial of marital status to same-sex couples—or even that this Court could theoretically do so. Lord, an opinion with such scatter-shot rationales as this one (federalism noises among them) can be distinguished in many ways. And deserves to be. State and lower federal courts should take the Court at its word and distinguish away.

In my opinion, however, the view that this Court will take of state prohibition of same-sex marriage is indicated beyond mistaking by today's opinion. As I have said, the real rationale of today's opinion, whatever disappearing trail of its legalistic argle-bargle one chooses to follow, is that DOMA is motivated by "`bare . . . desire to harm'" couples in same-sex marriages. Supra, at 18. How easy it is, indeed how inevitable, to reach the same conclusion with regard to state laws denying same-sex couples marital status. Consider how easy (inevitable) it is to make the following substitutions in a passage from today's opinion ante, at 22:

"DOMA's This state law's principal effect is to identify a subset of state-sanctioned marriages constitutionally protected sexual relationships, see Lawrence, and make them unequal. The principal purpose is to impose inequality, not for other reasons like governmental efficiency. Responsibilities, as well as rights, enhance the dignity and integrity of the person. And DOMA this state law contrives to deprive some couples married under the laws of their State enjoying constitutionally protected sexual relationships, but not other couples, of both rights and responsibilities."

Or try this passage, from ante, at 22-23:

"[DOMA] This state law tells those couples, and all the world, that their otherwise valid marriages relationships are unworthy of federal state recognition. This places same-sex couples in an unstable position of being in a second-tier marriage relationship. The differentiation demeans the couple, whose moral and sexual choices the Constitution protects, see Lawrence, . . . ."

Or this, from ante, at 23—which does not even require alteration, except as to the invented number:

"And it humiliates tens of thousands of children now being raised by same-sex couples. The law in question makes it even more difficult for the children to understand the integrity and closeness of their own family and its concord with other families in their community and in their daily lives."

Similarly transposable passages—deliberately transposable, I think—abound. In sum, that Court which finds it so horrific that Congress irrationally and hatefully robbed same-sex couples of the "personhood and dignity" which state legislatures conferred upon them, will of a certitude be similarly appalled by state legislatures' irrational and hateful failure to acknowledge that "personhood and dignity" in the first place. Ante, at 26. As far as this Court is concerned, no one should be fooled; it is just a matter of listening and waiting for the other shoe.

By formally declaring anyone opposed to same-sex marriage an enemy of human decency, the majority arms well every challenger to a state law restricting marriage to its traditional definition. Henceforth those challengers will lead with this Court's declaration that there is "no legitimate purpose" served by such a law, and will claim that the traditional definition has "the purpose and effect to disparage and to injure" the "personhood and dignity" of same-sex couples, see ante, at 25, 26. The majority's limiting assurance will be meaningless in the face of language like that, as the majority well knows. That is why the language is there. The result will be a judicial distortion of our society's debate over marriage—a debate that can seem in need of our clumsy "help" only to a member of this institution.

As to that debate: Few public controversies touch an institution so central to the lives of so many, and few inspire such attendant passion by good people on all sides. Few public controversies will ever demonstrate so vividly the beauty of what our Framers gave us, a gift the Court pawns today to buy its stolen moment in the spotlight: a system of government that permits us to rule ourselves. Since DOMA's passage, citizens on all sides of the question have seen victories and they have seen defeats. There have been plebiscites, legislation, persuasion, and loud voices—in other words, democracy. Victories in one place for some, see North Carolina Const., Amdt. 1 (providing that "[m]arriage between one man and one woman is the only domestic legal union that shall be valid or recognized in this State") (approved by a popular vote, 61% to 39% on May 8, 2012),^[6] are offset by victories in other places for others, see Maryland Question 6 (establishing "that Maryland's civil marriage laws allow gay and lesbian couples to obtain a civil marriage license") (approved by a popular vote, 52% to 48%, on November 6, 2012).^[7] Even in a single State, the question has come out differently on different occasions. Compare Maine Question 1 (permitting "the State of Maine to issue marriage licenses to same-sex couples") (approved by a popular vote, 53% to 47%, on November 6, 2012)^[8] with Maine Question 1 (rejecting "the new law that lets same-sex couples marry") (approved by a popular vote, 53% to 47%, on November 3, 2009).^[9]

In the majority's telling, this story is black-and-white: Hate your neighbor or come along with us. The truth is more complicated. It is hard to admit that one's political opponents are not monsters, especially in a struggle like this one, and the challenge in the end proves more than today's Court can handle. Too bad. A reminder that disagreement over something so fundamental as marriage can still be politically legitimate would have been a fit task for what in earlier times was called the judicial temperament. We might have covered ourselves with honor today, by promising all sides of this debate that it was theirs to settle and that we would respect their resolution. We might have let the People decide.

But that the majority will not do. Some will rejoice in today's decision, and some will despair at it; that is the nature of a controversy that matters so much to so many. But the Court has cheated both sides, robbing the winners of an honest victory, and the losers of the peace that comes from a fair defeat. We owed both of them better. I dissent.

JUSTICE ALITO, with whom JUSTICE THOMAS joins as to Parts II and III, dissenting.

Our Nation is engaged in a heated debate about samesex marriage. That debate is, at bottom, about the nature of the institution of marriage. Respondent Edith Windsor, supported by the United States, asks this Court to intervene in that debate, and although she couches her argument in different terms, what she seeks is a holding that enshrines in the Constitution a particular understanding of marriage under which the sex of the partners makes no difference. The Constitution, however, does not dictate that choice. It leaves the choice to the people, acting through their elected representatives at both the federal and state levels. I would therefore hold that Congress did not violate Windsor's constitutional rights by enacting §3 of the Defense of Marriage Act (DOMA), 110 Stat. 2419, which defines the meaning of marriage under federal statutes that either confer upon married persons certain federal benefits or impose upon them certain federal obligations.

I

I turn first to the question of standing. In my view, the United States clearly is not a proper petitioner in this case. The United States does not ask us to overturn the judgment of the court below or to alter that judgment in any way. Quite to the contrary, the United States argues emphatically in favor of the correctness of that judgment. We have never before reviewed a decision at the sole behest of a party that took such a position, and to do so would be to render an advisory opinion, in violation of Article III's dictates. For the reasons given in JUSTICE SCALIA's dissent, I do not find the Court's arguments to the contrary to be persuasive.

Whether the Bipartisan Legal Advisory Group of the House of Representatives (BLAG) has standing to petition is a much more difficult question. It is also a significantly closer question than whether the intervenors in Hollingsworth v. Perry, ante, p. ___ —which the Court also decides today—have standing to appeal. It is remarkable that the Court has simultaneously decided that the United States, which "receive[d] all that [it] ha[d] sought" below, Deposit Guaranty Nat. Bank v. Roper, 445 U. S. 326, 333 (1980), is a proper petitioner in this case but that the intervenors in Hollingsworth, who represent the party that lost in the lower court, are not. In my view, both the Hollingsworth intervenors and BLAG have standing.^[10]

A party invoking the Court's authority has a sufficient stake to permit it to appeal when it has "`suffered an injury in fact' that is caused by `the conduct complained of' and that `will be redressed by a favorable decision.'" Camreta v. Greene, 563 U. S. ___, ___ (2011) (slip op., at 5) (quoting Lujan v. Defenders of Wildlife, 504 U. S. 555, 560-561 (1992)). In the present case, the House of Representatives, which has authorized BLAG to represent its interests in this matter,^[11] suffered just such an injury.

In INS v. Chadha, 462 U. S. 919 (1983), the Court held that the two Houses of Congress were "proper parties" to file a petition in defense of the constitutionality of the one-house veto statute, id., at 930, n. 5 (internal quotation marks omitted). Accordingly, the Court granted and decided petitions by both the Senate and the House, in addition to the Executive's petition. Id., at 919, n. *. That the two Houses had standing to petition is not surprising: The Court of Appeals' decision in Chadha, by holding the one-house veto to be unconstitutional, had limited Congress' power to legislate. In discussing Article III standing, the Court suggested that Congress suffered a similar injury whenever federal legislation it had passed was struck down, noting that it had "long held that Congress is the proper party to defend the validity of a statute when an agency of government, as a defendant charged with enforcing the statute, agrees with plaintiffs that the statute is inapplicable or unconstitutional." Id., at 940.

The United States attempts to distinguish Chadha on the ground that it "involved an unusual statute that vested the House and the Senate themselves each with special procedural rights—namely, the right effectively to veto Executive action." Brief for United States (jurisdiction) 36. But that is a distinction without a difference: just as the Court of Appeals decision that the Chadha Court affirmed impaired Congress' power by striking down the one-house veto, so the Second Circuit's decision here impairs Congress' legislative power by striking down an Act of Congress. The United States has not explained why the fact that the impairment at issue in Chadha was "special" or "procedural" has any relevance to whether Congress suffered an injury. Indeed, because legislating is Congress' central function, any impairment of that function is a more grievous injury than the impairment of a procedural add-on.

The Court's decision in Coleman v. Miller, 307 U. S. 433 (1939), bolsters this conclusion. In Coleman, we held that a group of state senators had standing to challenge a lower court decision approving the procedures used to ratify an amendment to the Federal Constitution. We reasoned that the senators' votes—which would otherwise have carried the day—were nullified by that action. See id., at 438 ("Here, the plaintiffs include twenty senators, whose votes against ratification have been overridden and virtually held for naught although if they are right in their contentions their votes would have been sufficient to defeat ratification. We think that these senators have a plain, direct and adequate interest in maintaining the effectiveness of their votes"); id., at 446 ("[W]e find no departure from principle in recognizing in the instant case that at least the twenty senators whose votes, if their contention were sustained, would have been sufficient to defeat the resolution ratifying the proposed constitutional amendment, have an interest in the controversy which, treated by the state court as a basis for entertaining and deciding the federal questions, is sufficient to give the Court jurisdiction to review that decision"). By striking down §3 of DOMA as unconstitutional, the Second Circuit effectively "held for naught" an Act of Congress. Just as the state-senator-petitioners in Coleman were necessary parties to the amendment's ratification, the House of Representatives was a necessary party to DOMA's passage; indeed, the House's vote would have been sufficient to prevent DOMA's repeal if the Court had not chosen to execute that repeal judicially.

Both the United States and the Court-appointed amicus err in arguing that Raines v. Byrd, 521 U. S. 811 (1997), is to the contrary. In that case, the Court held that Members of Congress who had voted "nay" to the Line Item Veto Act did not have standing to challenge that statute in federal court. Raines is inapposite for two reasons. First, Raines dealt with individual Members of Congress and specifically pointed to the individual Members' lack of institutional endorsement as a sign of their standing problem: "We attach some importance to the fact that appellees have not been authorized to represent their respective Houses of Congress in this action, and indeed both Houses actively oppose their suit." Id., at 829; see also ibid., n. 10 (citing cases to the effect that "members of collegial bodies do not have standing to perfect an appeal the body itself has declined to take" (internal quotation marks omitted)).

Second, the Members in Raines—unlike the state senators in Coleman—were not the pivotal figures whose votes would have caused the Act to fail absent some challenged action. Indeed, it is telling that Raines characterized Coleman as standing "for the proposition that legislators whose votes would have been sufficient to defeat (or enact) a specific legislative Act have standing to sue if that legislative action goes into effect (or does not go into effect), on the ground that their votes have been completely nullified." 521 U. S., at 823. Here, by contrast, passage by the House was needed for DOMA to become law. U. S. Const., Art. I, §7 (bicameralism and presentment requirements for legislation).

I appreciate the argument that the Constitution confers on the President alone the authority to defend federal law in litigation, but in my view, as I have explained, that argument is contrary to the Court's holding in Chadha, and it is certainly contrary to the Chadha Court's endorsement of the principle that "Congress is the proper party to defend the validity of a statute" when the Executive refuses to do so on constitutional grounds. 462 U. S., at 940. See also 2 U. S. C. §288h(7) (Senate Legal Counsel shall defend the constitutionality of Acts of Congress when placed in issue).^[12] Accordingly, in the narrow category of cases in which a court strikes down an Act of Congress and the Executive declines to defend the Act, Congress both has standing to defend the undefended statute and is a proper party to do so.

II

Windsor and the United States argue that §3 of DOMA violates the equal protection principles that the Court has found in the Fifth Amendment's Due Process Clause. See Brief for Respondent Windsor (merits) 17-62; Brief for United States (merits) 16-54; cf. Bolling v. Sharpe, 347 U. S. 497 (1954). The Court rests its holding on related arguments. See ante, at 24-25.

Same-sex marriage presents a highly emotional and important question of public policy—but not a difficult question of constitutional law. The Constitution does not guarantee the right to enter into a same-sex marriage. Indeed, no provision of the Constitution speaks to the issue.

The Court has sometimes found the Due Process Clauses to have a substantive component that guarantees liberties beyond the absence of physical restraint. And the Court's holding that "DOMA is unconstitutional as a deprivation of the liberty of the person protected by the Fifth Amendment of the Constitution," ante, at 25, suggests that substantive due process may partially underlie the Court's decision today. But it is well established that any "substantive" component to the Due Process Clause protects only "those fundamental rights and liberties which are, objectively, `deeply rooted in this Nation's history and tradition,'" Washington v. Glucksberg, 521 U. S. 702, 720-721 (1997); Snyder v. Massachusetts, 291 U. S. 97, 105 (1934) (referring to fundamental rights as those that are so "rooted in the traditions and conscience of our people as to be ranked as fundamental"), as well as "`implicit in the concept of ordered liberty,' such that `neither liberty nor justice would exist if they were sacrificed.'" Glucksberg, supra, at 721 (quoting Palko v. Connecticut, 302 U. S. 319, 325-326 (1937)).

It is beyond dispute that the right to same-sex marriage is not deeply rooted in this Nation's history and tradition. In this country, no State permitted same-sex marriage until the Massachusetts Supreme Judicial Court held in 2003 that limiting marriage to opposite-sex couples violated the State Constitution. See Goodridge v. Department of Public Health, 440 Mass. 309, 798 N. E. 2d 941. Nor is the right to same-sex marriage deeply rooted in the traditions of other nations. No country allowed same-sex couples to marry until the Netherlands did so in 2000.^[13]

What Windsor and the United States seek, therefore, is not the protection of a deeply rooted right but the recognition of a very new right, and they seek this innovation not from a legislative body elected by the people, but from unelected judges. Faced with such a request, judges have cause for both caution and humility.

The family is an ancient and universal human institution. Family structure reflects the characteristics of a civilization, and changes in family structure and in the popular understanding of marriage and the family can have profound effects. Past changes in the understanding of marriage—for example, the gradual ascendance of the idea that romantic love is a prerequisite to marriage— have had far-reaching consequences. But the process by which such consequences come about is complex, involving the interaction of numerous factors, and tends to occur over an extended period of time.

We can expect something similar to take place if samesex marriage becomes widely accepted. The long-term consequences of this change are not now known and are unlikely to be ascertainable for some time to come.^[14] There are those who think that allowing same-sex marriage will seriously undermine the institution of marriage. See, e.g., S. Girgis, R. Anderson, & R. George, What is Marriage? Man and Woman: A Defense 53-58 (2012); Finnis, Marriage: A Basic and Exigent Good, 91 The Monist 388, 398 (2008).^[15] Others think that recognition of same-sex marriage will fortify a now-shaky institution. See, e.g., A. Sullivan, Virtually Normal: An Argument About Homosexuality 202-203 (1996); J. Rauch, Gay Marriage: Why It Is Good for Gays, Good for Straights, and Good for America 94 (2004).

At present, no one—including social scientists, philosophers, and historians—can predict with any certainty what the long-term ramifications of widespread acceptance of same-sex marriage will be. And judges are certainly not equipped to make such an assessment. The Members of this Court have the authority and the responsibility to interpret and apply the Constitution. Thus, if the Constitution contained a provision guaranteeing the right to marry a person of the same sex, it would be our duty to enforce that right. But the Constitution simply does not speak to the issue of same-sex marriage. In our system of government, ultimate sovereignty rests with the people, and the people have the right to control their own destiny. Any change on a question so fundamental should be made by the people through their elected officials.

III

Perhaps because they cannot show that same-sex marriage is a fundamental right under our Constitution, Windsor and the United States couch their arguments in equal protection terms. They argue that §3 of DOMA discriminates on the basis of sexual orientation, that classifications based on sexual orientation should trigger a form of "heightened" scrutiny, and that §3 cannot survive such scrutiny. They further maintain that the governmental interests that §3 purports to serve are not sufficiently important and that it has not been adequately shown that §3 serves those interests very well. The Court's holding, too, seems to rest on "the equal protection guarantee of the Fourteenth Amendment," ante, at 25— although the Court is careful not to adopt most of Windsor's and the United States' argument.

In my view, the approach that Windsor and the United States advocate is misguided. Our equal protection framework, upon which Windsor and the United States rely, is a judicial construct that provides a useful mechanism for analyzing a certain universe of equal protection cases. But that framework is ill suited for use in evaluating the constitutionality of laws based on the traditional understanding of marriage, which fundamentally turn on what marriage is.

Underlying our equal protection jurisprudence is the central notion that "[a] classification `must be reasonable, not arbitrary, and must rest upon some ground of difference having a fair and substantial relation to the object of the legislation, so that all persons similarly circumstanced shall be treated alike.'" Reed v. Reed, 404 U. S. 71, 76 (1971) (quoting F. S. Royter Guano Co. v. Virginia, 253 U. S. 412, 415 (1920)). The modern tiers of scrutiny—on which Windsor and the United States rely so heavily—are a heuristic to help judges determine when classifications have that "fair and substantial relation to the object of the legislation." Reed, supra, at 76.

So, for example, those classifications subject to strict scrutiny—i.e., classifications that must be "narrowly tailored" to achieve a "compelling" government interest, Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701, 720 (2007) (internal quotation marks omitted)—are those that are "so seldom relevant to the achievement of any legitimate state interest that laws grounded in such considerations are deemed to reflect prejudice and antipathy." Cleburne v. Cleburne Living Center, Inc., 473 U. S. 432, 440 (1985); cf. id., at 452-453 (Stevens, J., concurring) ("It would be utterly irrational to limit the franchise on the basis of height or weight; it is equally invalid to limit it on the basis of skin color. None of these attributes has any bearing at all on the citizen's willingness or ability to exercise that civil right").

In contrast, those characteristics subject to so-called intermediate scrutiny—i.e., those classifications that must be "`substantially related'" to the achievement of "important governmental objective[s]," United States v. Virginia, 518 U. S. 515, 524 (1996); id., at 567 (SCALIA, J., dissenting)—are those that are sometimes relevant considerations to be taken into account by legislators, but "generally provid[e] no sensible ground for different treatment," Cleburne, supra, at 440. For example, the Court has held that statutory rape laws that criminalize sexual intercourse with a woman under the age of 18 years, but place no similar liability on partners of underage men, are grounded in the very real distinction that "young men and young women are not similarly situated with respect to the problems and the risks of sexual intercourse." Michael M. v. Superior Court, Sonoma Cty., 450 U. S. 464, 471 (1981) (plurality opnion). The plurality reasoned that "[o]nly women may become pregnant, and they suffer disproportionately the profound physical, emotional, and psychological consequences of sexual activity." Ibid. In other contexts, however, the Court has found that classifications based on gender are "arbitrary," Reed, supra, at 76, and based on "outmoded notions of the relative capabilities of men and women," Cleburne, supra, at 441, as when a State provides that a man must always be preferred to an equally qualified woman when both seek to administer the estate of a deceased party, see Reed, supra, at 76-77.

Finally, so-called rational-basis review applies to classifications based on "distinguishing characteristics relevant to interests the State has the authority to implement." Cleburne, supra, at 441. We have long recognized that "the equal protection of the laws must coexist with the practical necessity that most legislation classifies for one purpose or another, with resulting disadvantages to various groups or persons." Romer v. Evans, 517 U. S. 620, 631 (1996). As a result, in rational-basis cases, where the court does not view the classification at issue as "inherently suspect," Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 218 (1995) (internal quotation marks omitted), "the courts have been very reluctant, as they should be in our federal system and with our respect for the separation of powers, to closely scrutinize legislative choices as to whether, how, and to what extent those interests should be pursued." Cleburne, supra, at 441-442.

In asking the Court to determine that §3 of DOMA is subject to and violates heightened scrutiny, Windsor and the United States thus ask us to rule that the presence of two members of the opposite sex is as rationally related to marriage as white skin is to voting or a Y-chromosome is to the ability to administer an estate. That is a striking request and one that unelected judges should pause before granting. Acceptance of the argument would cast all those who cling to traditional beliefs about the nature of marriage in the role of bigots or superstitious fools.

By asking the Court to strike down DOMA as not satisfying some form of heightened scrutiny, Windsor and the United States are really seeking to have the Court resolve a debate between two competing views of marriage.

The first and older view, which I will call the "traditional" or "conjugal" view, sees marriage as an intrinsically opposite-sex institution. BLAG notes that virtually every culture, including many not influenced by the Abrahamic religions, has limited marriage to people of the opposite sex. Brief for Respondent BLAG (merits) 2 (citing Hernandez v. Robles, 7 N. Y. 3d 338, 361, 855 N. E. 2d 1, 8 (2006) ("Until a few decades ago, it was an accepted truth for almost everyone who ever lived, in any society in which marriage existed, that there could be marriages only between participants of different sex")). And BLAG attempts to explain this phenomenon by arguing that the institution of marriage was created for the purpose of channeling heterosexual intercourse into a structure that supports child rearing. Brief for Respondent BLAG 44-46, 49. Others explain the basis for the institution in more philosophical terms. They argue that marriage is essentially the solemnizing of a comprehensive, exclusive, permanent union that is intrinsically ordered to producing new life, even if it does not always do so. See, e.g., Girgis, Anderson, & George, What is Marriage? Man and Woman: A Defense, at 23-28. While modern cultural changes have weakened the link between marriage and procreation in the popular mind, there is no doubt that, throughout human history and across many cultures, marriage has been viewed as an exclusively opposite-sex institution and as one inextricably linked to procreation and biological kinship.

The other, newer view is what I will call the "consentbased" vision of marriage, a vision that primarily defines marriage as the solemnization of mutual commitment— marked by strong emotional attachment and sexual attraction—between two persons. At least as it applies to heterosexual couples, this view of marriage now plays a very prominent role in the popular understanding of the institution. Indeed, our popular culture is infused with this understanding of marriage. Proponents of same-sex marriage argue that because gender differentiation is not relevant to this vision, the exclusion of same-sex couples from the institution of marriage is rank discrimination.

The Constitution does not codify either of these views of marriage (although I suspect it would have been hard at the time of the adoption of the Constitution or the Fifth Amendment to find Americans who did not take the traditional view for granted). The silence of the Constitution on this question should be enough to end the matter as far as the judiciary is concerned. Yet, Windsor and the United States implicitly ask us to endorse the consent-based view of marriage and to reject the traditional view, thereby arrogating to ourselves the power to decide a question that philosophers, historians, social scientists, and theologians are better qualified to explore.^[16] Because our constitutional order assigns the resolution of questions of this nature to the people, I would not presume to enshrine either vision of marriage in our constitutional jurisprudence.

Legislatures, however, have little choice but to decide between the two views. We have long made clear that neither the political branches of the Federal Government nor state governments are required to be neutral between competing visions of the good, provided that the vision of the good that they adopt is not countermanded by the Constitution. See, e.g., Rust v. Sullivan, 500 U. S. 173, 192 (1991) ("[T]he government `may make a value judgment favoring childbirth over abortion'" (quoting Maher v. Rue, 432 U. S. 464, 474 (1977))). Accordingly, both Congress and the States are entitled to enact laws recognizing either of the two understandings of marriage. And given the size of government and the degree to which it now regulates daily life, it seems unlikely that either Congress or the States could maintain complete neutrality even if they tried assiduously to do so.

Rather than fully embracing the arguments made by Windsor and the United States, the Court strikes down §3 of DOMA as a classification not properly supported by its objectives. The Court reaches this conclusion in part because it believes that §3 encroaches upon the States' sovereign prerogative to define marriage. See ante, at 21-22 ("As the title and dynamics of the bill indicate, its purpose is to discourage enactment of state same-sex marriage laws and to restrict the freedom and choice of couples married under those laws if they are enacted. The congressional goal was `to put a thumb on the scales and influence a state's decision as to how to shape its own marriage laws'" (quoting Massachusetts v. United States Dept. of Health and Human Servs., 682 F. 3d 1, 12-13 (CA1 2012))). Indeed, the Court's ultimate conclusion is that DOMA falls afoul of the Fifth Amendment because it "singles out a class of persons deemed by a State entitled to recognition and protection to enhance their own liberty" and "imposes a disability on the class by refusing to acknowledge a status the State finds to be dignified and proper." Ante, at 25 (emphasis added).

To the extent that the Court takes the position that the question of same-sex marriage should be resolved primarily at the state level, I wholeheartedly agree. I hope that the Court will ultimately permit the people of each State to decide this question for themselves. Unless the Court is willing to allow this to occur, the whiffs of federalism in the today's opinion of the Court will soon be scattered to the wind.

In any event, §3 of DOMA, in my view, does not encroach on the prerogatives of the States, assuming of course that the many federal statutes affected by DOMA have not already done so. Section 3 does not prevent any State from recognizing same-sex marriage or from extending to same-sex couples any right, privilege, benefit, or obligation stemming from state law. All that §3 does is to define a class of persons to whom federal law extends certain special benefits and upon whom federal law imposes certain special burdens. In these provisions, Congress used marital status as a way of defining this class—in part, I assume, because it viewed marriage as a valuable institution to be fostered and in part because it viewed married couples as comprising a unique type of economic unit that merits special regulatory treatment. Assuming that Congress has the power under the Constitution to enact the laws affected by §3, Congress has the power to define the category of persons to whom those laws apply.

* * *

For these reasons, I would hold that §3 of DOMA does not violate the Fifth Amendment. I respectfully dissent.

[1] For an even more advanced scavenger hunt, one might search the annals of Anglo-American law for another "Motion to Dismiss" like the one the United States filed in District Court: It argued that the court should agree "with Plaintiff and the United States" and "not dismiss" the complaint. (Emphasis mine.) Then, having gotten exactly what it asked for, the United States promptly appealed.

[2] There the Justice Department's refusal to defend the legislation was in accord with its longstanding (and entirely reasonable) practice of declining to defend legislation that in its view infringes upon Presidential powers. There is no justification for the Justice Department's abandoning the law in the present case. The majority opinion makes a point of scolding the President for his "failure to defend the constitutionality of an Act of Congress based on a constitutional theory not yet established in judicial decisions," ante, at 12. But the rebuke is tonguein-cheek, for the majority gladly gives the President what he wants. Contrary to all precedent, it decides this case (and even decides it the way the President wishes) despite his abandonment of the defense and the consequent absence of a case or controversy.

[3] JUSTICE ALITO attempts to limit his argument by claiming that Congress is injured (and can therefore appeal) when its statute is held unconstitutional without Presidential defense, but is not injured when its statute is held unconstitutional despite Presidential defense. I do not understand that line. The injury to Congress is the same whether the President has defended the statute or not. And if the injury is threatened, why should Congress not be able to participate in the suit from the beginning, just as the President can? And if having a statute declared unconstitutional (and therefore inoperative) by a court is an injury, why is it not an injury when a statute is declared unconstitutional by the President and rendered inoperative by his consequent failure to enforce it? Or when the President simply declines to enforce it without opining on its constitutionality? If it is the inoperativeness that constitutes the injury—the "impairment of [the legislative] function," as JUSTICE ALITO puts it, post, at 4—it should make no difference which of the other two branches inflicts it, and whether the Constitution is the pretext. A principled and predictable system of jurisprudence cannot rest upon a shifting concept of injury, designed to support standing when we would like it. If this Court agreed with JUSTICE ALITO's distinction, its opinion in Raines v. Byrd, 521 U. S. 811 (1997), which involved an original suit by Members of Congress challenging an assertedly unconstitutional law, would have been written quite differently; and JUSTICE ALITO's distinguishing of that case on grounds quite irrelevant to his theory of standing would have been unnecessary.

[4] Such a suggestion would be impossible, given the Federal Government's long history of making pronouncements regarding marriage—for example, conditioning Utah's entry into the Union upon its prohibition of polygamy. See Act of July 16, 1894, ch. 138, §3, 28 Stat. 108 ("The constitution [of Utah]" must provide "perfect toleration of religious sentiment," "Provided, That polygamous or plural marriages are forever prohibited").

[5] Since the Equal Protection Clause technically applies only against the States, see U. S. Const., Amdt. 14, Bolling and Moreno, dealing with federal action, relied upon "the equal protection component of the Due Process Clause of the Fifth Amendment," Moreno, 413 U. S., at 533.

[6] North Carolina State Board of Elections, Official Results: Primary Election of May 8, 2012, Constitutional Amendment.

[7] Maryland State Board of Elections, Official 2012 Presidential General Election Results for All State Questions, Question 06.

[8] Maine Bureau of Elections, Nov. 3, 2009, Referendum Tabulation (Question 1).

[9] Maine Bureau of Elections, Nov. 6, 2012, Referendum Election Tabulations (Question 1).

[10] Our precedents make clear that, in order to support our jurisdiction, BLAG must demonstrate that it had Article III standing in its own right, quite apart from its status as an intervenor. See Diamond v. Charles, 476 U. S. 54, 68 (1986) ("Although intervenors are considered parties entitled, among other things, to seek review by this Court, an intervenor's right to continue a suit in the absence of the party on whose side intervention was permitted is contingent upon a showing by the intervenor that he fulfills the requirements of Art. III" (citation omitted)); Arizonans for Official English v. Arizona, 520 U. S. 43, 64 (1997) ("Standing to defend on appeal in the place of an original defendant, no less than standing to sue, demands that the litigant possess a direct stake in the outcome" (internal quotation marks omitted)); id., at 65 ("An intervenor cannot step into the shoes of the original party unless the intervenor independently fulfills the requirements of Article III" (internal quotation marks omitted)).

[11] H. Res. 5, 113th Cong., 1st Sess., §4(a)(1)(B) (2013) ("[BLAG] continues to speak for, and articulates the institutional position of, the House in all litigation matters in which it appears, including in Windsor v. United States").

[12] Buckley v. Valeo, 424 U. S. 1 (1976), is not to the contrary. The Court's statements there concerned enforcement, not defense.

[13] Curry-Sumner, A Patchwork of Partnerships: Comparative Overview of Registration Schemes in Europe, in Legal Recognition of Same-Sex Partnerships 71, 72 (K. Boele-Woelki & A. Fuchs eds., rev. 2d ed., 2012).

[14] As sociologists have documented, it sometimes takes decades to document the effects of social changes—like the sharp rise in divorce rates following the advent of no-fault divorce—on children and society. See generally J. Wallerstein, J. Lewis, & S. Blakeslee, The Unexpected Legacy of Divorce: The 25 Year Landmark Study (2000).

[15] Among those holding that position, some deplore and some applaud this predicted development. Compare, e.g., Wardle, "Multiply and Replenish": Considering Same-Sex Marriage in Light of State Interests in Marital Procreation, 24 Harv. J. L. & Pub. Pol'y 771, 799 (2001) ("Culturally, the legalization of same-sex marriage would send a message that would undermine the social boundaries relating to marriage and family relations. The confusion of social roles linked with marriage and parenting would be tremendous, and the message of `anything goes' in the way of sexual behavior, procreation, and parenthood would wreak its greatest havoc among groups of vulnerable individuals who most need the encouragement of bright line laws and clear social mores concerning procreative responsibility") and Gallagher, (How) Will Gay Marriage Weaken Marriage as a Social Institution: A Reply to Andrew Koppelman, 2 U. St. Thomas L. J. 33, 58 (2005) ("If the idea of marriage really does matter—if society really does need a social institution that manages opposite-sex attractions in the interests of children and society—then taking an already weakened social institution, subjecting it to radical new redefinitions, and hoping that there are no consequences is probably neither a wise nor a compassionate idea"), with Brownworth, Something Borrowed, Something Blue: Is Marriage Right for Queers? in I Do/I Don't: Queers on Marriage 53, 58-59 (G. Wharton & I. Phillips eds. 2004) (Former President George W. "Bush is correct . . . when he states that allowing same-sex couples to marry will weaken the institution of marriage. It most certainly will do so, and that will make marriage a far better concept than it previously has been") and Willis, Can Marriage Be Saved? A Forum, The Nation, p. 16 (2004) (celebrating the fact that "conferring the legitimacy of marriage on homosexual relations will introduce an implicit revolt against the institution into its very heart").

[16] The degree to which this question is intractable to typical judicial processes of decisionmaking was highlighted by the trial in Hollingsworth v. Perry, ante, p. ___. In that case, the trial judge, after receiving testimony from some expert witnesses, purported to make "findings of fact" on such questions as why marriage came to be, Perry v. Schwarzenegger, 704 F. Supp. 2d 921, 958 (ND Cal. 2010) (finding of fact no. 27) ("Marriage between a man and a woman was traditionally organized based on presumptions of division of labor along gender lines. Men were seen as suited for certain types of work and women for others. Women were seen as suited to raise children and men were seen as suited to provide for the family"), what marriage is, id., at 961 (finding of fact no. 34) ("Marriage is the state recognition and approval of a couple's choice to live with each other, to remain committed to one another and to form a household based on their own feelings about one another and to join in an economic partnership and support one another and any dependents"), and the effect legalizing same-sex marriage would have on opposite-sex marriage, id.,at 972 (finding of fact no. 55) ("Permitting same-sex couples to marry will not affect the number of opposite-sex couples who marry, divorce, cohabit, have children outside of marriage or otherwise affect the stability of opposite-sex marriages").

At times, the trial reached the heights of parody, as when the trial judge questioned his ability to take into account the views of great thinkers of the past because they were unavailable to testify in person in his courtroom. See 13 Tr. in No. C 09-2292 VRW (ND Cal.), pp. 3038-3039.

And, if this spectacle were not enough, some professors of constitutional law have argued that we are bound to accept the trial judge's findings—including those on major philosophical questions and predictions about the future—unless they are "clearly erroneous." See Brief for Constitutional Law and Civil Procedure Professors as Amici Curiae in Hollingsworth v. Perry, O. T. 2012, No. 12-144, pp. 2-3 ("[T]he district court's factual findings are compelling and should be given significant weight"); id., at 25 ("Under any standard of review, this Court should credit and adopt the trial court's findings because they result from rigorous and exacting application of the Federal Rules of Evidence, and are supported by reliable research and by the unanimous consensus of mainstream social science experts"). Only an arrogant legal culture that has lost all appreciation of its own limitations could take such a suggestion seriously.

          Syllabus

          A married woman Air Force officer (hereafter appellant) sought increased benefits for her husband as a 'dependent' under 37 U.S.C. §§ 401, 403, and 10 U.S.C. §§ 1072, 1076. Those statutes provide, solely for administrative convenience, that spouses of male members of the uniformed services are dependents for purposes of obtaining increased quarters allowances and medical and dental benefits, but that spouses of female members are not dependents unless they are in fact dependent for over one-half of their support. When her application was denied for failure to satisfy the statutory dependency standard, appellant and her husband brought this suit in District Court, contending that the statutes deprived servicewomen of due process. From that Court's adverse ruling, they took a direct appeal. Held: The judgment is reversed. Pp. 682—691, 691—692, 341 F.Supp. 201, reversed.

Page 678

          Joseph J. Levin, Jr., Montgomery, Ala., for appellants.

          Ruth B. Ginsburg, New York City, for American Civil Liberties Union, amicus curiae, by special leave of Court.

          Samuel Huntington, Washington, D.C., for appellees.

           Mr. Justice BRENNAN announced the judgment of the Court in an opinion in which Mr. Justice DOUGLAS, Mr. Justice WHITE, and Mr. Justice MARSHALL join.

          The question before us concerns the right of a female member of the uniformed services¹ to claim her spouse as a 'dependent' for the purposes of obtaining increased quarters allowances and medical and dental benefits under 37 U.S.C. §§ 401, 403, and 10 U.S.C. §§ 1072, 1076, on an equal footing with male members. Under these statutes, a serviceman may claim his wife as a 'dependent' without regard to whether she is in fact dependent upon him for any part of her support. 37 U.S.C. § 401(1); 10 U.S.C. § 1072(2)(A). A servicewoman, on the other hand, may not claim her husband as a 'dependent' under these programs unless he is in fact dependent upon her for over one-half of his sup-

Page 679

port. 37 U.S.C. § 401; 10 U.S.C. § 1072(2)(C).² Thus, the question for decision is whether this difference in treatment constitutes an unconstitutional discrimination against servicewomen in violation of the Due Process Clause of the Fifth Amendment. A three-judge District Court for the Middle District of Alabama, one judge issenting, rejected this contention and sustained the constitutionality of the provisions of the statutes making this distinction. 341 F.Supp. 201 (1972). We noted probable jurisdiction. 409 U.S. 840, 93 S.Ct. 64, 34 L.Ed.2d 78 (1972). We reverse.

          In an effort to attract career personnel through reenlistment, Congress established, in 37 U.S.C. § 401 et seq., and 10 U.S.C. § 1071 et seq., a scheme for the provision of fringe benefits to members of the uniformed services on a competitive basis with business and industry.³ Thus, under 37 U.S.C. § 403, a member of the uniformed services with dependents is entitled to an

Page 680

increased 'basic allowence for quarters' and, under 10 U.S.C. § 1076, a member's dependents are provided comprehensive medical and dental care.

          Appellant Sharron Frontiero, a lieutenant in the United States Air Force, sought increased quarters allowances, and housing and medical benefits for her husband, appellant Joseph Frontiero, on the ground that he was her 'dependent.' Although such benefits would automatically have been granted with respect to the wife of a male member of the uniformed services, appellant's application was denied because she failed to demonstrate that her husband was dependent on her for more than one-half of his support.⁴ Appellants then commenced this suit, contendingthat, by making this distinction, the statutes unreasonably discriminate on the basis of sex in violation of the Due Process Clause of the Fifth Amendment.⁵ In essence, appellants asserted that the discriminatory impact of the statutes is twofold: first, as a procedural matter, a female member is required to demonstrate her spouse's dependency, while no such burden is imposed upon male members; and, second, as a substantive matter, a male member who does not provide more than one-half of his wife's support receives benefits, while a similarly situated female member is denied such benefits. Appellants therefore sought a permanent in-

Page 681

junction against the continued enforcement of these statutes and an order directing the appellees to provide Lieutenant Frontiero with the same housing and medical benefits that a similarly situated male member would receive.

          Although the legislative history of these statutes sheds virtually no light on the purposes underlying the differential treatment accorded male and female members,⁶ a majority of the three-judge District Court surmised that Congress might reasonably have concluded that, since the husband in our society is generally the 'breadwinner' in the family—and the wife typically the 'dependent' partner—'it would be more economical to require married female members claiming husbands to prove actual dependency than to extend the presumption of dependency to such members.' 341 F.Supp., at 207. Indeed, given the fact that approximately 99% of all members of the uniformed services are male, the District

Page 682

Court speculated that such differential treatment might conceivably lead to a 'considerable saving of administrative expense and manpower.' Ibid.

          At the outset, appellants contend that classifications based upon sex, like classifications based upon race,⁷ alienage,⁸ and national origin,⁹ are inherently suspect and must therefore be subjected to close judicial scrutiny. We agree and, indeed, find at least implicit support for such an approach in our unanimous decision only last Term in Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971).

          In Reed, the Court considered the constitutionality of an Idaho statute providing that, when two individuals are otherwise equally entitled to appointment as administrator of an estate, the male applicant must be preferred to the female. Appellant, the mother of the deceased, and appellee, the father, filed competing petitions for appointment as administrator of their son's estate. Since the parties, as parents of the deceased, were members of the same entitlement class the statutory preference was invoked and the father's petition was therefore granted. Appellant claimed that this statute, by giving a mandatory preference to males over females without regard to their individual qualifications, violated the Equal Protection Clause of the Fourteenth Amendment.

          The Court noted that the Idaho statute 'provides that different treatment be accorded to the applicants on the basis of their sex; it thus establishes a classification sub-

Page 683

ject to scrutiny under the Equal Protection Clause.' 404 U.S., at 75, 92 S.Ct. at 253. Under 'traditional' equal protection analysis, a legislative classification must be sustained unless it is 'patently arbitrary' and bears no rational relationship to a legitimate governmental interest. See Jefferson v. Hackney, 406 U.S. 535, 546, 92 S.Ct. 1724, 1731, 32 L.Ed.2d 285 (1972); Richardson v. Belcher, 404 U.S. 78, 81, 92 S.Ct. 254, 257, 30 L.Ed.2d 231 (1971); Flemming v. Nestor, 363 U.S. 603, 611, 80 S.Ct. 1367, 1372, 4 L.Ed.2d 1435 (1960); McGowan v. Maryland, 366 U.S. 420, 426, 81 S.Ct. 1101, 1105, 6 L.Ed.2d 393 (1961); Dandridge v. Williams, 397 U.S. 471, 485, 90 S.Ct. 115o, 1161, 25 L.Ed.2d 491 (1970).

          In an effort to meet this standard, appellee contended that the statutory scheme was a reasonable measure designed to reduce the workload on probate courts by eliminating one class of contests. Moreover, appellee argued that the mandatory preference for male applicants was in itself reasonable since 'men (are) as a rule more conversant with business affairs than . . . women.' ¹⁰ Indeed, appellee maintained that 'it is a matter of common knowledge, that women still are not engaged in politics, the professions, business or industry to the extent that men are.'¹¹ And the Idaho Supreme Court, in upholding the constitutionality of this statute, suggested that the Idaho Legislature might reasonably have 'concluded that in general men are better qualified to act as an administrator than are women.'¹²

          Despite these contentions, however, the Court held the statutory preference for male applicants unconstitutional. In reaching this result, the Court implicitly rejected appellee's apparently rational explanation of the statutory scheme, and concluded that, by ignoring the individual qualifications of particular applicants, the challenged statute provide 'dissimilar treatment for men and women who are . . . similarly situated.' 404 U.S.,

Page 684

at 77, 92 S.Ct., at 254. The Court therefore held that, even though the State's interest in achieving administrative efficiency 'is not without some legitimacy,' '(t)o give a mandatory preference to members of either sex over members of the other, merely to accomplish the elimination of hearings on the merits, is to make the very kind of arbitrary legislative choice forbidden by the (Constitution) . . ..' Id., at 76, 92 S.Ct. at 254. This departure from 'traditional' rational-basis analysis with respect to sex-based classifications is clearly justified.

          There can be no doubt that our Nation has had a long and unfortunate history of sex discrimination.¹³ Traditionally, such discrimination was rationalized by an attitude of 'romantic paternalism' which, in practical effect, put women, not on a pedestal, but in a cage. Indeed, this paternalistic attitude became so firmly rooted in our national consciousness that, 100 years ago, a distinguished Member of this Court was able to proclaim:

          'Man is, or should be, women's protector and defender. The natural and proper timidity and delicacy which belongs to the female sex evidently unfits it for many of the occupations of civil life. The constitution of the family organization, which is founded in the divine ordinance, as well as in the nature of things, indicates the domestic sphere as that which properly belongs to the domain and functions of womanhood. The harmony, not to say identity, of interests and views which belong, or should belong, to the family institution is repugnant to the idea of a woman adopting a distinct and

Page 685

          independent career from that of her husband. . . .

          '. . . The paramount destiny and mission of woman are to fulfil the noble and benign offices of wife and mother. This is the law of the Creator.' Bradwell v. State of Illinois, 16 Wall. 130, 141, 21 L.Ed.2d 442 (1873) (Bradley, J., concurring).

          As a result of notions such as these, our statute books gradually became laden with gross, stereotyped distinctions between the sexes and, indeed, throughout much of the 19th century the position of women in our society was, in many respects, comparable to that of blacks under the pre-Civil War slave codes. Neither slaves nor women could hold office, serve on juries, or bring suit in their own names, and married women traditionally were denied the legal capacity to hold or convey property or to serve as legal guardians of their own children. See generally L. Kanowitz, Women and the Law: The Unfinished Revolution 5—6 (1969); G. Myrdal, An American Dilemma 1073 (20th Anniversary ed. 1962). And although blacks were guaranteed the right to vote in 1870, women were denied even that right—which is itself 'preservative of other basic civil and political rights'¹⁴—until adoption of the Nineteenth Amendment half a century later.

          It is true, of course, that the position of women in America has improved markedly in recent decades.¹⁵

Page 686

Nevertheless, it can hardly be doubted that, in part because of the high visibility of the sex characteristic,¹⁶ women still face pervasive, although at times more subtle, discrimination in our educational institutions, in the job market and, perhaps most conspicuously, in the political arena.¹⁷ See generally K. Amundsen, The Silenced Majority: Women and American Democracy (1971); The President's Task Force on Women's Rights and Responsibilities, A Matter of Simple Act of 1964, 84 Harv.L.Rev. 1109 (1971).

          Moreover, since sex, like race and national origin, is an immutable characteristic determined solely by the accident of birth, the imposition of special disabilities upon the members of a particular sex because of their sex would seem to violate 'the basic concept of our system that legal burdens should bear some relationship to individual responsibility . . ..' Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 175, 92 S.Ct. 1400, 1407, 31 L.Ed.2d 768 (1972). And what differentiates sex from such non-suspect statuses as intelligence or physical disability, and aligns it with the recognized suspect criteria, is that the sex characteristic frequently bears no relation to ability to perform or contribute to society.¹⁸ As a result, statutory distinc-

Page 687

tions between the sexes often have the effect of invidiously relegating the entire class of females to inferior legal status without regard to the actual capabilities of its individual members.

          We might also note that, over the past decade, Congress has itself manifested an increasing sensitivity to sex-based classifications. In Tit. VII of the Civil Rights Act of 1964, for example, Congress expressly declared that no employer, labor union, or other organization subject to the provisions of the Act shall discriminate against any individual on the basis of 'race, color, religion, sex, or national origin.'¹⁹ Similarly, the Equal Pay Act of 1963 provides that no employer covered by the Act 'shall discriminate . . . between employees on the basis of sex.'²⁰ And § 1 of the Equal Rights Amendment, passed by Congress on March 22, 1972, and submitted to the legislatures of the States for ratification, declares that '(e)quality of rights under the law shall not be denied or abridged by the United States or by any State on account of sex.'²¹ Thus, Congress itself has concluded that classifications based upon sex are inherently invidious, and this conclusion of a coequal

Page 688

branch of Government is not without significance to the question presently under consideration. Cf. Oregon v. Mitchell, 400 U.S. 112, 240, 248—249, 91 S.Ct. 260, 322, 327, 27 L.Ed.2d 272 (1970) (opinion of Brennan, White, and Marshall, JJ.); Katzenbach v. Morgan, 384 U.S. 641, 648—649, 86 S.Ct. 1717, 1722, 16 L.Ed.2d 828 (1966).

          With these considerations in mind, we can only conclude that classifications based upon sex, like classifications based upon race, alienage, or national origin, are inherently suspect, and must therefore be subjected to strict judicial scrutiny. Applying the analysis mandated by that stricter standard of review, it is clear that the statutory scheme now before us is constitutionally invalid.

          The sole basis of the classification established in the challenged statutes is the sex of the individuals involved. Thus, under 37 U.S.C. §§ 401, 403, and 10 U.S.C. §§ 2072, 2076, a female member of the uniformed services seeking to obtain housing and medical benefits for her spouse must prove his dependency in fact, whereas no such burden is imposed upon male members. In addition, the statutes operate so as to deny benefits to a female member, such as appellant Sharron Frontiero, who provides less than one-half of her spouse's support, while at the same time granting such benefits to a male member who likewise provides less than one-half of his spouse's support. Thus, to this extent at least, it may fairly be said that these statutes command 'dissimilar treatment for men and women who are . . . similarly situated.' Reed v. Reed, 404 U.S., at 77, 92 S.Ct., at 254.

          Moreover, the Government concedes that the differential treatment accorded men and women under these statutes serves no purpose other than mere 'administrative convenience.' In essence, the Government maintains that, as an empirical matter, wives in our society frequently are dependent upon their husbands, while hus-

Page 689

bands rarely are dependent upon their wives. Thus, the Government argues that Congress might reasonably have concluded that it would be both cheaper and easier simply conclusively to presume that wives of male members are financially dependent upon their husbands, while burdening female members with the task of establishing dependency in fact.²²

          The Government offers no concrete evidence, however, tending to support its view that such differential treatment in fact saves the Government any money. In order to satisfy the demands of strict judicial scrutiny, the Government must demonstrate, for example, that it is actually cheaper to grant increased benefits with respect to all male members, than it is to determine which male members are in fact entitled to such benefits and to grant increased benefits only to those members whose wives actually meet the dependency requirement. Here, however, there is substantial evidence that, if put to the test, many of the wives of male members would fail to qualify for benefits.²³ And in light of the fact that the

Page 690

dependency determination with respect to the husbands of female members is presently made solely on the basis of affidavits rather than through the more costly hearing process,²⁴ the Government's explanation of the statutory scheme is, to say the least, questionable.

          In any case, our prior decisions make clear that, although efficacious administration of governmental programs is not without some importance, 'the Constitution recognizes higher values than speed and efficiency.' Stanley v. Illinois, 405 U.S. 645, 656, 92 S.Ct. 1208, 1215, 31 L.Ed.2d 551 (1972). And when we enter the realm of 'strict judicial scrutiny,' there can be no doubt that 'administrative convenience' is not a shibboleth, the mere recitation of which dictates constitutionality. See Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969); Carrington v. Rash, 380 U.S. 89, 85 S.Ct. 775, 13 L.Ed.2d 675 (1965). On the contrary, any statutory scheme which draws a sharp line between the sexes, Solely for the purpose of achieving administrative convenience, necessarily commands 'dissimilar treatment for men and women who are . . . similarly situated,' and therefore involves the 'very kind of arbitrary legislative choice forbidden by the (Constitution) . . ..' Reed v. Reed, 404 U.S., at 77, 76, 92 S.Ct., at 254. We therefore conclude that, by according differential treatment to male and female members of the uniformed services for the sole purpose of achieving administrative

Page 691

convenience, the challenged statutes violate the Due Process Clause of the Fifth Amendment insofar as they require a female member to prove the dependency of her husband.²⁵

          Reversed.

          Mr. Justice STEWART concurs in the judgment, agreeing that the statutes before us work an invidious discrimination in violation of the Constitution. Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225.

          Mr. Justice REHNQUIST dissents for the reasons stated by Judge Rives in his opinion for the District Court, Frontiero v. Laird, 341 F.Supp. 201 (1972).

           Mr. Justice POWELL, with whom THE CHIEF JUSTICE and Mr. Justice BLACKMUN join, concurring in the judgment.

          I agree that the challenged statutes constitute an unconstitutional discrimination against servicewomen in violation of the Due Process Clause of the Fifth Amendment, but I cannot join the opinion of Mr. Justice BRENNAN, which would hold that all classifications based upon sex, 'like classifications based upon race, alienage, and national origin,' are 'inherently suspect and must therefore be subjected to close judicial scrutiny.' Ante, at 682. It is unnecessary for the Court in this case to

Page 692

characterize sex as a suspect classification, with all of the far-reaching implications of such a holding. Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971), which abundantly supports our decision today, did not add sex to the narrowly limited group of classifications which are inherently suspect. In my view, we can and should decide this case on the authority of Reed and reserve for the future any expansion of its rationale.

          There is another, and I find compelling, reason for deferring a general categorizing of sex classifications as invoking the strictest test of judicial scrutiny. The Equal Rights Amendment, which if adopted will resolve the substance of this precise question, has been approved by the Congress and submitted for ratification by the States. If this Amendment is duly adopted, it will represent the will of the people accomplished in the manner prescribed by the Constitution. By acting prematurely and unnecessarily, as I view it, the Court has assumed a decisional responsibility at the very time when state legislatures, functioning within the traditional democractic process, are debating the proposed Amendment. It seems to me that this reaching out to pre-empt by judicial action a major political decision which is currently in process of resolution does not reflect appropriate respect for duly prescribed legislative processes.

          There are times when this Court, under our system, cannot avoid a constitutional decision on issues which normally should be resolved by the elected representatives of the people. But democratic institutions are weakened, and confidence in the restraint of the Court is impaired, when we appear unnecessarily to decide sensitive issues of broad social and political importance at the very time they are under consideration within the prescribed constitutional processes.

^{1. The 'uniformed services' include the Army, Navy, Air Force, Marine Corps, Coast Guard, Environmental Science Services Administration, and Public Health Service. 37 U.S.C. § 101(3); 10 U.S.C. § 1072(1).}

^2. Title 37 U.S.C. § 401 provides in pertinent part:

'In this chapter, 'dependent,' with respect to a member of a uniformed service, means—

'(1) his spouse;

'However, a person is not a dependent of a female member unless he is in fact dependent on her for over one-half of his support . . ..'

10 U.S.C. § 1072(2) provides in pertinent part:

"Dependent,' with respect to a member . . . of a uniformed service, means—

'(A) the wife;

'(C) the husband, if he is in fact dependent on the member . . . for over one-half of his support. . . .'

^3. See 102 Cong.Rec. 3849—3850 (Cong. Kilday), 8043 (Sen. Saltonstall); 95 Cong.Rec. 7662 (Cong. Kilday), 7664 (Cong. Short), 7666 (Cong. Havenner), 7667 (Cong. Bates), 7671 (Cong. Price). See also 10 U.S.C. § 1071.

^4. Appellant Joseph Frontiero is a full-time student at Huntingdon College in Montgomery, Alabama. According to the agreed stipulation of facts, his living expenses, including his share of the house-hold expenses, total approximately $354 per month. Since he receives $205 per month in veterans' benefits, it is clear that he is not dependent upon appellant Sharron Frontiero for more than one-half of his support.

^5. '(W)hile the Fifth Amendment contains no equal protection clause, it does forbid discrimination that is 'so unjustifiable as to be violative of due process." Schneider v. Rusk, 377 U.S. 163, 168, 84 S.Ct. 1187, 1190, 12 L.Ed.2d 218 (1964); see Shapiro v. Thompson, 394 U.S. 618, 641—642, 89 S.Ct. 1322, 1335, 22 L.Ed.2d 600 (1969); Bolling v. Sharpe, 347 U.S. 497, 74 S.Ct. 693, 98 L.Ed. 884 (1954).

^6. The housing provisions, set forth in 37 U.S.C. § 401 et seq., were enacted as part of the Career Compensation Act of 1949, which established a uniform pattern of military pay and allowances, consolidating and revising the piecemeal legislation that had been developed over the previous 40 years. See H.R.Rep. No. 779, 81st Cong., 1st Sess.; S.Rep. No. 733, 81st Cong., 1st Sess. The Act apparently retained in substance the dependency definitions of § 4 of the Pay Readjustment Act of 1942 (56 Stat. 361), as amended by § 6 of the Act of September 7, 1944 (58 Stat. 730), which required a female member of the service to demonstrate her spouse's dependency. It appears that this provision was itself derived from unspecified earlier enactments. See S.Rep. No. 917, 78th Cong., 2d Sess., 4.

The medical benefits legislation, 10 U.S.C. § 1071 et seq., was enacted as the Dependents' Medical Care Act of 1956. As such, it was designed to revise and make uniform the existing law relating to medical services for military personnel. It, too, appears to have carried forward, without explanation, the dependency provisions found in other military pay and allowance legislation. See H.R.Rep. No. 1805, 84th Cong., 2d Sess.; S.Rep. No. 1878, 84th Cong., 2d Sess.

^7. See Loving v. Virginia, 388 U.S. 1, 11, 87 S.Ct. 1817, 1823, 18 L.Ed.2d 1010 (1967); McLaughlin v. Florida, 379 U.S. 184, 191—192, 85 S.Ct. 283, 287—288, 13 L.Ed.2d 222 (1964); Bolling v. Sharpe, supra, 347 U.S., at 499, 74 S.Ct., at 694.

^8. See Graham v. Richardson, 403 U.S. 365, 372, 91 S.Ct. 1848, 1852, 29 L.Ed.2d 534 (1971).

^9. See Oyama v. California, 332 U.S. 633, 644—646, 68 S.Ct. 269, 274—275, 92 L.Ed. 249 (1948); Korematsu v. United States, 323 U.S. 214, 216, 65 S.Ct. 193, 194, 89 L.Ed. 194 (1944); Hirabayashi v. United States, 320 U.S. 81, 100, 63 S.Ct. 1375, 1385, 87 L.Ed. 1774 (1943).

^10. Brief for Appellee in No. 70—4, O.T. 1971, Reed v. Reed, p. 12.

^11. Id., at 12—13.

^12. Reed v. Reed, 93 Idaho 511, 514, 465 P.2d 635, 638 (1970).

^13. Indeed, the position of women in this country at its inception is reflected in the view expressed by Thomas Jefferson that women should be neither seen nor heard in society's decisionmaking councils. See M. Gruberg, Women in American Politics 4 (1968). See also 2 A. de Tocqueville, Democracy in America (Reeves trans. 1948).

^14. Reynolds v. Sims, 377 U.S. 533, 562, 84 S.Ct. 1362, 1381, 12 L.Ed.2d 506 (1964); see Dunn v. Blumstein, 405 U.S. 330, 336, 92 S.Ct. 995, 999, 31 L.Ed.2d 274 (1972); Kramer v. Union Free School District, 395 U.S. 621, 626, 89 S.Ct. 1886, 1889, 23 L.Ed.2d 583 (1969); Yick Wo v. Hopkins, 118 U.S. 356, 370, 6 S.Ct. 1064, 1071, 30 L.Ed. 220 (1886).

^15. See generally The President's Task Force on Women's Rights and Responsibilities, A Matter of Simple Justice (1970); L. Kanowitz, Women and the Law: The Unfinished Revolution (1969); A. Montagu, Man's Most Dangerous Myth (4th ed. 1964); The President's Commission on the Status of Women, American Women (1963).

^16. See, e.g., Note, Sex Discrimination and Equal Protection: Do We Need a Constitutional Amendment?, 84 Harv.L.Rev. 1499, 1507 (1971).

^17. It is true, of course, that when viewed in the abstract, women do not constitute a small and powerless minority. Nevertheless, in part because of past discrimination, women are vastly underrepresented in this Nation's decisionmaking councils. There has never been a female President, nor a female member of this Court. Not a single woman presently sits in the United States Senate, and only 14 women hold seats in the House of Representatives. And, as appellants point out, this underrepresentation is present throughout all levels of our State and Federal Government. See Joint Reply Brief of Appellants and American Civil Liberties Union (Amicus Curiae) 9.

^18. See, e.g., Developments in the Law—Equal Protection, 82 Harv.L.Rev. 1065, 1173—1174 (1969).

^19. 42 U.S.C. § 2000e—2(a), (b), (c) (emphasis added). See generally, Sape & Hart, Title VII Reconsidered: The Equal Employment Opportunity Act of 1972, 40 Geo.Wash.L.Rev. 824 (1972); Developments in the Law—Employment Discrimination and Title VII of the Civil Rights Act of 1964, 8j Harv.L.Rev. 1109 (1971).

^20. 29 U.S.C. § 206(d) (emphasis added). See generally Murphy, Female Wage Discrimination: A Study of the Equal Pay Act 1963—1970, 39 U.Cin.L.Rev. 615 (1970).

^21. H.R.J.Res. No. 208, 92d Cong., 2d Sess. (1972). In conformity with these principles, Congress in recent years has amended various statutory schemes similar to those presently under consideration so as to eliminate the differential treatment of men and women. See 5 U.S.C. § 2108, as amended, 85 Stat. 644; 5 U.S.C. § 7152, as amended, 85 Stat. 644; 5 U.S.C. § 8341, as amended, 84 Stat. 1961; 38 U.S.C. § 102(b), as amended, 86 Stat. 1092.

^22. It should be noted that these statutes are not in any sense designed to rectify the effects of past discrimination against women. See Gruenwald v. Gardner, 390 F.2d 591 (CA2), cert. denied, 393 U.S. 982, 89 S.Ct. 456, 21 L.Ed.2d 445 (1968); cf. Jones v. Alfred H. Mayer Co., 392 U.S. 409, 88 S.Ct. 2186, 20 L.Ed.2d 1189 (1968); South Carolina v. Katzenbach, 383 U.S. 301, 86 S.Ct. 803, 15 L.Ed.2d 769 (1966). On the contrary, these statutes seize upon a group—women—who have historically suffered discrimination in employment, and rely on the effects of this past discrimination as a justification for heaping on additional economic disadvantages. Cf. Gaston County v. United States, 395 U.S. 285, 296—297, 89 S.Ct. 1720, 1725—1726, 23 L.Ed.2d 309 (1969).

^23. In 1971, 43% of all women over the age of 16 were in the labor force, and 18% of all women worked full time 12 months per year. See U.S. Women's Bureau, Dept. of Labor, Highlights of Women's Employment & Education 1 (W.B.Pub. No. 72—191, Mar. 1972). Moreover, 41.5% of all married women are employed. See U.S. Bureau of Labor Statistics, Dept. of Labor, Work Experience of the Population in 1971, p. 4 (Summary Special Labor Force Report, Aug. 1972). It is also noteworthy that, while the median income of a male member of the armed forces is approximately $3,686, see The Report of the President's Commission on an All-Volunteer Armed Force 51, 181 (1970), the median income for all women over the age of 14, including those who are not employed, is approximately $2,237. See Statistical Abstract of the United States Table No. 535 (1972), Source: U.S. Bureau of the Census, Current Population Reports, Series P—60, No. 80. Applying the statutory definition of 'dependency' to these statistics, it appears that in the 'median' family, the wife of a male member must have personal expenses of approximately $4,474, or about 75% of the total family income, in order to qualify as a 'dependent.'

^24. Tr. of Oral Arg. 27—28.

^25. As noted earlier, the basic purpose of these statutes was to provide fringe benefits to members of the uniformed services in order to establish a compensation pattern which would attract career personnel through re- enlistment. See n. 3, supra, and accompanying text. Our conclusion in no wise invalidates the statutory schemes except insofar as they require a female member to prove the dependency of her spouse. See Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 92 S.Ct. 1400, 31 L.Ed.2d 768 (1972); Levy v. Louisiana, 391 U.S. 68, 88 S.Ct. 1509, 20 L.Ed.2d 436 (1968); Moritz v. Commissioner of Internal Revenue, 469 F.2d 466 (CA10 1972). See also 1 U.S.C. § 1.

518 U.S. 515 (1996)

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT

[517] [517] [518] Ginsburg, J.,delivered the opinion of the Court, in which Stevens, O'Connor, Kennedy, Souter, and Breyer, JJ., joined. Rehnquist, C. J., filed an opinion concurring in the judgment, post, p. 558. Scalia, J., filed a dissenting opinion, post, p. 566. Thomas, J., took no part in the consideration or decision of the case.

Paul Bender argued the cause for the United States in both cases. With him on the briefs were Solicitor General Days, Assistant Attorney General Patrick, Cornelia T. L. Pillard, Jessica Dunsay Silver, and Thomas E. Chandler.

Theodore B. Olson argued the cause and filed briefs for respondents in No. 94-1941 and petitioners in No. 94-2107. With him on the briefs were James S. Gilmore III, Attorney General of Virginia, William H. Hurd, Deputy Attorney General, Thomas G. Hungar, D. Jarrett Arp, Robert H. Patterson, Jr., Anne Marie Whittemore, William G. Broaddus, J. William Boland, Griffin B. Bell, and William A. Clineburg, Jr.^[2]

[519] Justice Ginsburg, delivered the opinion of the Court.

Virginia's public institutions of higher learning include an incomparable military college, Virginia Military Institute (VMI). The United States maintains that the Constitution's equal protection guarantee precludes Virginia from reserving exclusively to men the unique educational opportunities VMI affords. We agree.

[520] I

Founded in 1839, VMI is today the sole single-sex school among Virginia's 15 public institutions of higher learning. VMI's distinctive mission is to produce "citizen-soldiers," men prepared for leadership in civilian life and in military service. VMI pursues this mission through pervasive training of a kind not available anywhere else in Virginia. Assigning prime place to character development, VMI uses an "adversative method" modeled on English public schools and once characteristic of military instruction. VMI constantly endeavors to instill physical and mental discipline in its cadets and impart to them a strong moral code. The school's graduates leave VMI with heightened comprehension of their capacity to deal with duress and stress, and a large sense of accomplishment for completing the hazardous course.

VMI has notably succeeded in its mission to produce leaders; among its alumni are military generals, Members of Congress, and business executives. The school's alumni overwhelmingly perceive that their VMI training helped them to realize their personal goals. VMI's endowment reflects the loyalty of its graduates; VMI has the largest per-student endowment of all public undergraduate institutions in the Nation.

Neither the goal of producing citizen-soldiers nor VMI's implementing methodology is inherently unsuitable to women. And the school's impressive record in producing leaders has made admission desirable to some women. Nevertheless, Virginia has elected to preserve exclusively for men the advantages and opportunities a VMI education affords.

II

A

From its establishment in 1839 as one of the Nation's first state military colleges, see 1839 Va. Acts, ch. 20, VMI has remained financially supported by Virginia and "subject to [521] the control of the [Virginia] General Assembly," Va. Code Ann. § 23-92 (1993). First southern college to teach engineering and industrial chemistry, see H. Wise, Drawing Out the Man: The VMI Story 13 (1978) (The VMI Story), VMI once provided teachers for the Commonwealth's schools, see 1842 Va. Acts, ch. 24, § 2 (requiring every cadet to teach in one of the Commonwealth's schools for a 2-year period).^[3] Civil War strife threatened the school's vitality, but a resourceful superintendent regained legislative support by highlighting "VMI's great potential[,] through its technical know-how," to advance Virginia's postwar recovery. The VMI Story 47.

VMI today enrolls about 1,300 men as cadets.^[4] Its academic offerings in the liberal arts, sciences, and engineering are also available at other public colleges and universities in Virginia. But VMI's mission is special. It is the mission of the school

"`to produce educated and honorable men, prepared for the varied work of civil life, imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready as citizen-soldiers to defend their country in [522] time of national peril.' " 766 F. Supp. 1407, 1425 (WD Va. 1991) (quoting Mission Study Committee of the VMI Board of Visitors, Report, May 16, 1986).

In contrast to the federal service academies, institutions maintained "to prepare cadets for career service in the armed forces," VMI's program "is directed at preparation for both military and civilian life"; "[o]nly about 15% of VMI cadets enter career military service." 766 F. Supp., at 1432.

VMI produces its "citizen-soldiers" through "an adversative, or doubting, model of education" which features "[p]hysical rigor, mental stress, absolute equality of treatment, absence of privacy, minute regulation of behavior, and indoctrination in desirable values." Id., at 1421. As one Commandant of Cadets described it, the adversative method "`dissects the young student,' " and makes him aware of his "`limits and capabilities,' " so that he knows "`how far he can go with his anger, . . . how much he can take under stress, . . . exactly what he can do when he is physically exhausted.' " Id. , at 1421-1422 (quoting Col. N. Bissell).

VMI cadets live in spartan barracks where surveillance is constant and privacy nonexistent; they wear uniforms, eat together in the mess hall, and regularly participate in drills. Id. , at 1424, 1432. Entering students are incessantly exposed to the rat line, "an extreme form of the adversative model," comparable in intensity to Marine Corps boot camp. Id. , at 1422. Tormenting and punishing, the rat line bonds new cadets to their fellow sufferers and, when they have completed the 7-month experience, to their former tormentors. Ibid.

VMI's "adversative model" is further characterized by a hierarchical "class system" of privileges and responsibilities, a "dyke system" for assigning a senior class mentor to each entering class "rat," and a stringently enforced "honor code," which prescribes that a cadet "`does not lie, cheat, steal nor tolerate those who do.' " Id. , at 1422-1423.

[523] VMI attracts some applicants because of its reputation as an extraordinarily challenging military school, and "because its alumni are exceptionally close to the school." Id. , at 1421. "[W]omen have no opportunity anywhere to gain the benefits of [the system of education at VMI]." Ibid.

B

In 1990, prompted by a complaint filed with the Attorney General by a female high-school student seeking admission to VMI, the United States sued the Commonwealth of Virginia and VMI, alleging that VMI's exclusively male admission policy violated the Equal Protection Clause of the Fourteenth Amendment. Id. , at 1408.^[5] Trial of the action consumed six days and involved an array of expert witnesses on each side. Ibid.

In the two years preceding the lawsuit, the District Court noted, VMI had received inquiries from 347 women, but had responded to none of them. Id. , at 1436. "[S]ome women, at least," the court said, "would want to attend the school if they had the opportunity." Id. , at 1414. The court further recognized that, with recruitment, VMI could "achieve at least 10% female enrollment"—"a sufficient `critical mass' to provide the female cadets with a positive educational experience." Id. , at 1437-1438. And it was also established that "some women are capable of all of the individual activities required of VMI cadets." Id. , at 1412. In addition, experts agreed that if VMI admitted women, "the VMI ROTC experience would become a better training program from the perspective of the armed forces, because it would provide training in dealing with a mixed-gender army." Id. , at 1441.

The District Court ruled in favor of VMI, however, and rejected the equal protection challenge pressed by the United States. That court correctly recognized that Mississippi Univ. for Women v. Hogan, 458 U. S. 718 (1982), was [524] the closest guide. 766 F. Supp., at 1410. There, this Court underscored that a party seeking to uphold government action based on sex must establish an "exceedingly persuasive justification" for the classification. Mississippi Univ. for Women, 458 U. S., at 724 (internal quotation marks omitted). To succeed, the defender of the challenged action must show "at least that the classification serves important governmental objectives and that the discriminatory means employed are substantially related to the achievement of those objectives." Ibid. (internal quotation marks omitted).

The District Court reasoned that education in "a singlegender environment, be it male or female," yields substantial benefits. 766 F. Supp., at 1415. VMI's school for men brought diversity to an otherwise coeducational Virginia system, and that diversity was "enhanced by VMI's unique method of instruction." Ibid. If single-gender education for males ranks as an important governmental objective, it becomes obvious, the District Court concluded, that the only means of achieving the objective "is to exclude women from the all-male institution—VMI." Ibid.

"Women are [indeed] denied a unique educational opportunity that is available only at VMI," the District Court acknowledged. Id. , at 1432. But "[VMI's] single-sex status would be lost, and some aspects of the [school's] distinctive method would be altered," if women were admitted, id. , at 1413: "Allowance for personal privacy would have to be made," id. , at 1412; "[p]hysical education requirements would have to be altered, at least for the women," id. , at 1413; the adversative environment could not survive unmodified, id. , at 1412-1413. Thus, "sufficient constitutional justification" had been shown, the District Court held, "for continuing [VMI's] single-sex policy." Id. , at 1413.

The Court of Appeals for the Fourth Circuit disagreed and vacated the District Court's judgment. The appellate court held: "The Commonwealth of Virginia has not . . . advanced any state policy by which it can justify its determination, [525] under an announced policy of diversity, to afford VMI's unique type of program to men and not to women." 976 F. 2d 890, 892 (1992).

The appeals court greeted with skepticism Virginia's assertion that it offers single-sex education at VMI as a facet of the Commonwealth's overarching and undisputed policy to advance "autonomy and diversity." The court underscored Virginia's nondiscrimination commitment: "`[I]t is extremely important that [colleges and universities] deal with faculty, staff, and students without regard to sex, race, or ethnic origin. ` " Id. , at 899 (quoting 1990 Report of the Virginia Commission on the University of the 21st Century). "That statement," the Court of Appeals said, "is the only explicit one that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions." 976 F. 2d, at 899. Furthermore, the appeals court observed, in urging "diversity" to justify an all-male VMI, the Commonwealth had supplied "no explanation for the movement away from [single-sex education] in Virginia by public colleges and universities." Ibid. In short, the court concluded, "[a] policy of diversity which aims to provide an array of educational opportunities, including single-gender institutions, must do more than favor one gender." Ibid.

The parties agreed that "some women can meet the physical standards now imposed on men," id. , at 896, and the court was satisfied that "neither the goal of producing citizen soldiers nor VMI's implementing methodology is inherently unsuitable to women," id. , at 899. The Court of Appeals, however, accepted the District Court's finding that "at least these three aspects of VMI's program—physical training, the absence of privacy, and the adversative approach—would be materially affected by coeducation." Id. , at 896-897. Remanding the case, the appeals court assigned to Virginia, in the first instance, responsibility for selecting a remedial course. The court suggested these options for the Commonwealth: Admit women to VMI; establish parallel institutions [526] or programs; or abandon state support, leaving VMI free to pursue its policies as a private institution. Id. , at 900. In May 1993, this Court denied certiorari. See 508 U. S. 946; see also ibid. (opinion of Scalia, J., noting the interlocutory posture of the litigation).

C

In response to the Fourth Circuit's ruling, Virginia proposed a parallel program for women: Virginia Women's Institute for Leadership (VWIL). The 4-year, state-sponsored undergraduate program would be located at Mary Baldwin College, a private liberal arts school for women, and would be open, initially, to about 25 to 30 students. Although VWIL would share VMI's mission—to produce "citizensoldiers"—the VWIL program would differ, as does Mary Baldwin College, from VMI in academic offerings, methods of education, and financial resources. See 852 F. Supp. 471, 476-477 (WD Va. 1994).

The average combined SAT score of entrants at Mary Baldwin is about 100 points lower than the score for VMI freshmen. See id., at 501. Mary Baldwin's faculty holds "significantly fewer Ph. D.'s than the faculty at VMI," id. , at 502, and receives significantly lower salaries, see Tr. 158 (testimony of James Lott, Dean of Mary Baldwin College), reprinted in 2 App. in Nos. 94-1667 and 94-1717 (CA4) (hereinafter Tr.). While VMI offers degrees in liberal arts, the sciences, and engineering, Mary Baldwin, at the time of trial, offered only bachelor of arts degrees. See 852 F. Supp., at 503. A VWIL student seeking to earn an engineering degree could gain one, without public support, by attending Washington University in St. Louis, Missouri, for two years, paying the required private tuition. See ibid.

Experts in educating women at the college level composed the Task Force charged with designing the VWIL program; Task Force members were drawn from Mary Baldwin's own faculty and staff. Id. , at 476. Training its attention on methods of instruction appropriate for "most women," the [527] Task Force determined that a military model would be "wholly inappropriate" for VWIL. Ibid.; see 44 F. 3d 1229, 1233 (CA4 1995).

VWIL students would participate in ROTC programs and a newly established, "largely ceremonial" Virginia Corps of Cadets, id. , at 1234, but the VWIL House would not have a military format, 852 F. Supp., at 477, and VWIL would not require its students to eat meals together or to wear uniforms during the schoolday, id., at 495. In lieu of VMI's adversative method, the VWIL Task Force favored "a cooperative method which reinforces self-esteem." Id., at 476. In addition to the standard bachelor of arts program offered at Mary Baldwin, VWIL students would take courses in leadership, complete an off-campus leadership externship, participate in community service projects, and assist in arranging a speaker series. See 44 F. 3d, at 1234.

Virginia represented that it will provide equal financial support for in-state VWIL students and VMI cadets, 852 F. Supp., at 483, and the VMI Foundation agreed to supply a $5.4625 million endowment for the VWIL program, id., at 499. Mary Baldwin's own endowment is about $19 million; VMI's is $131 million. Id., at 503. Mary Baldwin will add $35 million to its endowment based on future commitments; VMI will add $220 million. Ibid. The VMI Alumni Association has developed a network of employers interested in hiring VMI graduates. The Association has agreed to open its network to VWIL graduates, id., at 499, but those graduates will not have the advantage afforded by a VMI degree.

D

Virginia returned to the District Court seeking approval of its proposed remedial plan, and the court decided the plan met the requirements of the Equal Protection Clause. Id. , at 473. The District Court again acknowledged evidentiary support for these determinations: "[T]he VMI methodology could be used to educate women and, in fact, some [528] women . . . may prefer the VMI methodology to the VWIL methodology." Id. , at 481. But the "controlling legal principles," the District Court decided, "do not require the Commonwealth to provide a mirror image VMI for women." Ibid. The court anticipated that the two schools would "achieve substantially similar outcomes." Ibid. It concluded: "If VMI marches to the beat of a drum, then Mary Baldwin marches to the melody of a fife and when the march is over, both will have arrived at the same destination." Id. , at 484.

A divided Court of Appeals affirmed the District Court's judgment. 44 F. 3d 1229 (CA4 1995). This time, the appellate court determined to give "greater scrutiny to the selection of means than to the [Commonwealth's] proffered objective." Id. , at 1236. The official objective or purpose, the court said, should be reviewed deferentially. Ibid. Respect for the "legislative will," the court reasoned, meant that the judiciary should take a "cautious approach," inquiring into the "legitima[cy]" of the governmental objective and refusing approval for any purpose revealed to be "pernicious." Ibid.

"[P]roviding the option of a single-gender college education may be considered a legitimate and important aspect of a public system of higher education," the appeals court observed, id., at 1238; that objective, the court added, is "not pernicious," id., at 1239. Moreover, the court continued, the adversative method vital to a VMI education "has never been tolerated in a sexually heterogeneous environment." Ibid. The method itself "was not designed to exclude women," the court noted, but women could not be accommodated in the VMI program, the court believed, for female participation in VMI's adversative training "would destroy. . . any sense of decency that still permeates the relationship between the sexes." Ibid.

Having determined, deferentially, the legitimacy of Virginia's purpose, the court considered the question of means. [529] Exclusion of "men at Mary Baldwin College and women at VMI," the court said, was essential to Virginia's purpose, for without such exclusion, the Commonwealth could not "accomplish [its] objective of providing single-gender education." Ibid.

The court recognized that, as it analyzed the case, means merged into end, and the merger risked "bypass[ing] any equal protection scrutiny." Id., at 1237. The court therefore added another inquiry, a decisive test it called "substantive comparability." Ibid. The key question, the court said, was whether men at VMI and women at VWIL would obtain "substantively comparable benefits at their institution or through other means offered by the [S]tate." Ibid. Although the appeals court recognized that the VWIL degree "lacks the historical benefit and prestige" of a VMI degree, it nevertheless found the educational opportunities at the two schools "sufficiently comparable." Id., at 1241.

Senior Circuit Judge Phillips dissented. The court, in his judgment, had not held Virginia to the burden of showing an "`exceedingly persuasive [justification]' " for the Commonwealth's action. Id., at 1247 (quoting Mississippi Univ. for Women, 458 U. S., at 724). In Judge Phillips' view, the court had accepted "rationalizations compelled by the exigencies of this litigation," and had not confronted the Commonwealth's "actual overriding purpose." 44 F. 3d, at 1247. That purpose, Judge Phillips said, was clear from the historical record; it was "not to create a new type of educational opportunity for women, . . . nor to further diversify the Commonwealth's higher education system[,] . . . but [was] simply . . . to allow VMI to continue to exclude women in order to preserve its historic character and mission." Ibid.

Judge Phillips suggested that the Commonwealth would satisfy the Constitution's equal protection requirement if it "simultaneously opened single-gender undergraduate institutions having substantially comparable curricular and extra-curricular programs, funding, physical plant, administration [530] and support services, and faculty and library resources." Id., at 1250. But he thought it evident that the proposed VWIL program, in comparison to VMI, fell "far short . . . from providing substantially equal tangible and intangible educational benefits to men and women." Ibid.

The Fourth Circuit denied rehearing en banc. 52 F. 3d 90 (1995). Circuit Judge Motz, joined by Circuit Judges Hall, Murnaghan, and Michael, filed a dissenting opinion.^[6] Judge Motz agreed with Judge Phillips that Virginia had not shown an "`exceedingly persuasive justification' " for the disparate opportunities the Commonwealth supported. Id., at 92 (quoting Mississippi Univ. for Women, 458 U. S., at 724). She asked: "[H]ow can a degree from a yet to be implemented supplemental program at Mary Baldwin be held `substantively comparable' to a degree from a venerable Virginia military institution that was established more than 150 years ago?" 52 F. 3d, at 93. "Women need not be guaranteed equal `results,' " Judge Motz said, "but the Equal Protection Clause does require equal opportunity . . . [and] that opportunity is being denied here." Ibid.

III

The cross-petitions in this suit present two ultimate issues. First, does Virginia's exclusion of women from the educational opportunities provided by VMI—extraordinary opportunities for military training and civilian leadership development—deny to women "capable of all of the individual activities required of VMI cadets," 766 F. Supp., at 1412, the equal protection of the laws guaranteed by the Fourteenth Amendment? Second, if VMI's "unique" situation, id. , at 1413—as Virginia's sole single-sex public institution of [531] higher education—offends the Constitution's equal protection principle, what is the remedial requirement?

IV

We note, once again, the core instruction of this Court's pathmarking decisions in J. E. B. v. Alabama ex rel. T. B., 511 U. S. 127, 136-137, and n. 6 (1994), and Mississippi Univ. for Women, 458 U. S., at 724 (internal quotation marks omitted): Parties who seek to defend gender-based government action must demonstrate an "exceedingly persuasive justification" for that action.

Today's skeptical scrutiny of official action denying rights or opportunities based on sex responds to volumes of history. As a plurality of this Court acknowledged a generation ago, "our Nation has had a long and unfortunate history of sex discrimination." Frontiero v. Richardson, 411 U. S. 677, 684 (1973). Through a century plus three decades and more of that history, women did not count among voters composing "We the People";^[7] not until 1920 did women gain a constitutional right to the franchise. Id., at 685. And for a half century thereafter, it remained the prevailing doctrine that government, both federal and state, could withhold from women opportunities accorded men so long as any "basis in reason" could be conceived for the discrimination. See, e. g., Goesaert v. Cleary, 335 U. S. 464, 467 (1948) (rejecting challenge of female tavern owner and her daughter to Michigan law denying bartender licenses to females—except for wives and daughters of male tavern owners; Court would not "give ear" to the contention that "an unchivalrous desire of male [532] bartenders to . . . monopolize the calling" prompted the legislation).

In 1971, for the first time in our Nation's history, this Court ruled in favor of a woman who complained that her State had denied her the equal protection of its laws. Reed v. Reed, 404 U. S. 71, 73 (holding unconstitutional Idaho Code prescription that, among "`several persons claiming and equally entitled to administer [a decedent's estate], males must be preferred to females' "). Since Reed, the Court has repeatedly recognized that neither federal nor state government acts compatibly with the equal protection principle when a law or official policy denies to women, simply because they are women, full citizenship stature—equal opportunity to aspire, achieve, participate in and contribute to society based on their individual talents and capacities. See, e. g., Kirchberg v. Feenstra, 450 U. S. 455, 462-463 (1981) (affirming invalidity of Louisiana law that made husband "head and master" of property jointly owned with his wife, giving him unilateral right to dispose of such property without his wife's consent); Stanton v. Stanton, 421 U. S. 7 (1975) (invalidating Utah requirement that parents support boys until age 21, girls only until age 18).

Without equating gender classifications, for all purposes, to classifications based on race or national origin,^[8] the Court, in post-Reed decisions, has carefully inspected official action that closes a door or denies opportunity to women (or to men). See J. E. B. , 511 U. S., at 152 (Kennedy, J., concurring in judgment) (case law evolving since 1971 "reveal[s] a strong presumption that gender classifications are invalid"). To summarize the Court's current directions for cases of official classification based on gender: Focusing on the differential [533] treatment or denial of opportunity for which relief is sought, the reviewing court must determine whether the proffered justification is "exceedingly persuasive." The burden of justification is demanding and it rests entirely on the State. See Mississippi Univ. for Women, 458 U. S., at 724. The State must show "at least that the [challenged] classification serves `important governmental objectives and that the discriminatory means employed' are `substantially related to the achievement of those objectives.' " Ibid. (quoting Wengler v. Druggists Mut. Ins. Co. , 446 U. S. 142, 150 (1980)). The justification must be genuine, not hypothesized or invented post hoc in response to litigation. And it must not rely on overbroad generalizations about the different talents, capacities, or preferences of males and females. See Weinberger v. Wiesenfeld, 420 U. S. 636, 643, 648 (1975); Califano v. Goldfarb, 430 U. S. 199, 223-224 (1977) (Stevens, J., concurring in judgment).

The heightened review standard our precedent establishes does not make sex a proscribed classification. Supposed "inherent differences" are no longer accepted as a ground for race or national origin classifications. See Loving v. Virginia, 388 U. S. 1 (1967). Physical differences between men and women, however, are enduring: "[T]he two sexes are not fungible; a community made up exclusively of one [sex] is different from a community composed of both." Ballard v. United States, 329 U. S. 187, 193 (1946).

"Inherent differences" between men and women, we have come to appreciate, remain cause for celebration, but not for denigration of the members of either sex or for artificial constraints on an individual's opportunity. Sex classifications may be used to compensate women "for particular economic disabilities [they have] suffered," Califano v. Webster, 430 U. S. 313, 320 (1977) (per curiam), to "promot[e] equal employment opportunity," see California Fed. Sav. & Loan Assn. v. Guerra, 479 U. S. 272, 289 (1987), to advance full development of the talent and capacities of our Nation's people.^[9] [534] But such classifications may not be used, as they once were, see Goesaert, 335 U. S., at 467, to create or perpetuate the legal, social, and economic inferiority of women.

Measuring the record in this case against the review standard just described, we conclude that Virginia has shown no "exceedingly persuasive justification" for excluding all women from the citizen-soldier training afforded by VMI. We therefore affirm the Fourth Circuit's initial judgment, which held that Virginia had violated the Fourteenth Amendment's Equal Protection Clause. Because the remedy proffered by Virginia—the Mary Baldwin VWIL program—does not cure the constitutional violation, i. e. , it does not provide equal opportunity, we reverse the Fourth Circuit's final judgment in this case.

V

The Fourth Circuit initially held that Virginia had advanced no state policy by which it could justify, under equal protection principles, its determination "to afford VMI's unique type of program to men and not to women." 976 F. 2d, at 892. Virginia challenges that "liability" ruling and asserts two justifications in defense of VMI's exclusion of [535] women. First, the Commonwealth contends, "single-sex education provides important educational benefits," Brief for Cross-Petitioners 20, and the option of single-sex education contributes to "diversity in educational approaches," id. , at 25. Second, the Commonwealth argues, "the unique VMI method of character development and leadership training," the school's adversative approach, would have to be modified were VMI to admit women. Id. , at 33-36 (internal quotation marks omitted). We consider these two justifications in turn.

A

Single-sex education affords pedagogical benefits to at least some students, Virginia emphasizes, and that reality is uncontested in this litigation.^[10] Similarly, it is not disputed that diversity among public educational institutions can serve the public good. But Virginia has not shown that VMI was established, or has been maintained, with a view to diversifying, by its categorical exclusion of women, educational opportunities within the Commonwealth. In cases of this genre, our precedent instructs that "benign" justifications proffered in defense of categorical exclusions will not be accepted automatically; a tenable justification must describe actual state purposes, not rationalizations for actions [536] in fact differently grounded. See Wiesenfeld, 420 U. S., at 648, and n. 16 ("mere recitation of a benign [or] compensatory purpose" does not block "inquiry into the actual purposes" of government-maintained gender-based classifications); Goldfarb, 430 U. S., at 212-213 (rejecting government-proffered purposes after "inquiry into the actual purposes" (internal quotation marks omitted)).

Mississippi Univ. for Women is immediately in point. There the State asserted, in justification of its exclusion of men from a nursing school, that it was engaging in "educational affirmative action" by "compensat[ing] for discrimination against women." 458 U. S., at 727. Undertaking a "searching analysis," id. , at 728, the Court found no close resemblance between "the alleged objective" and "the actual purpose underlying the discriminatory classification," id. , at 730. Pursuing a similar inquiry here, we reach the same conclusion.

Neither recent nor distant history bears out Virginia's alleged pursuit of diversity through single-sex educational options. In 1839, when the Commonwealth established VMI, a range of educational opportunities for men and women was scarcely contemplated. Higher education at the time was considered dangerous for women;^[11] reflecting [537] widely held views about women's proper place, the Nation's first universities and colleges—for example, Harvard in Massachusetts, William and Mary in Virginia—admitted only men. See E. Farello, A History of the Education of Women in the United States 163 (1970). VMI was not at all novel in this respect: In admitting no women, VMI followed the lead of the Commonwealth's flagship school, the University of Virginia, founded in 1819.

"[N]o struggle for the admission of women to a state university," a historian has recounted, "was longer drawn out, or developed more bitterness, than that at the University of Virginia." 2 T. Woody, A History of Women's Education in the United States 254 (1929) (History of Women's Education). In 1879, the State Senate resolved to look into the possibility of higher education for women, recognizing that Virginia "`has never, at any period of her history,' " provided for the higher education of her daughters, though she "`has liberally provided for the higher education of her sons.' " Ibid. (quoting 10 Educ. J. Va. 212 (1879)). Despite this recognition, no new opportunities were instantly open to women.^[12]

Virginia eventually provided for several women's seminaries and colleges. Farmville Female Seminary became a public institution in 1884. See supra, at 521, n. 2. Two women's schools, Mary Washington College and James Madison University, were founded in 1908; another, Radford University, was founded in 1910. 766 F. Supp., at 1418-1419. By the mid-1970's, all four schools had become coeducational. Ibid.

Debate concerning women's admission as undergraduates at the main university continued well past the century's midpoint. Familiar arguments were rehearsed. If women [538] were admitted, it was feared, they "would encroach on the rights of men; there would be new problems of government, perhaps scandals; the old honor system would have to be changed; standards would be lowered to those of other coeducational schools; and the glorious reputation of the university, as a school for men, would be trailed in the dust." 2 History of Women's Education 255.

Ultimately, in 1970, "the most prestigious institution of higher education in Virginia," the University of Virginia, introduced coeducation and, in 1972, began to admit women on an equal basis with men. See Kirstein v. Rector and Visitors of Univ. of Virginia, 309 F. Supp. 184, 186 (ED Va. 1970). A three-judge Federal District Court confirmed: "Virginia may not now deny to women, on the basis of sex, educational opportunities at the Charlottesville campus that are not afforded in other institutions operated by the [S]tate." Id. , at 187.

Virginia describes the current absence of public single-sex higher education for women as "an historical anomaly." Brief for Cross-Petitioners 30. But the historical record indicates action more deliberate than anomalous: First, protection of women against higher education; next, schools for women far from equal in resources and stature to schools for men; finally, conversion of the separate schools to coeducation. The state legislature, prior to the advent of this controversy, had repealed "[a]ll Virginia statutes requiring individual institutions to admit only men or women." 766 F. Supp., at 1419. And in 1990, an official commission, "legislatively established to chart the future goals of higher education in Virginia," reaffirmed the policy "`of affording broad access" while maintaining "autonomy and diversity.' " 976 F. 2d, at 898-899 (quoting Report of the Virginia Commission on the University of the 21st Century). Significantly, the commission reported:

"`Because colleges and universities provide opportunities for students to develop values and learn from role [539] models, it is extremely important that they deal with faculty, staff, and students without regard to sex, race, or ethnic origin.' " Id. , at 899 (emphasis supplied by Court of Appeals deleted).

This statement, the Court of Appeals observed, "is the only explicit one that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions." Ibid.

Our 1982 decision in Mississippi Univ. for Women prompted VMI to reexamine its male-only admission policy. See 766 F. Supp., at 1427-1428. Virginia relies on that reexamination as a legitimate basis for maintaining VMI's single-sex character. See Reply Brief for Cross-Petitioners 6. A Mission Study Committee, appointed by the VMI Board of Visitors, studied the problem from October 1983 until May 1986, and in that month counseled against "change of VMI status as a single-sex college." See 766 F. Supp., at 1429 (internal quotation marks omitted). Whatever internal purpose the Mission Study Committee served—and however well meaning the framers of the report—we can hardly extract from that effort any commonwealth policy evenhandedly to advance diverse educational options. As the District Court observed, the Committee's analysis "primarily focuse[d] on anticipated difficulties in attracting females to VMI," and the report, overall, supplied "very little indication of how th[e] conclusion was reached." Ibid.

In sum, we find no persuasive evidence in this record that VMI's male-only admission policy "is in furtherance of a state policy of `diversity.' " See 976 F. 2d, at 899. No such policy, the Fourth Circuit observed, can be discerned from the movement of all other public colleges and universities in Virginia away from single-sex education. See ibid. That court also questioned "how one institution with autonomy, but with no authority over any other state institution, can give effect to a state policy of diversity among institutions." Ibid. A purpose genuinely to advance an array of educational [540] options, as the Court of Appeals recognized, is not served by VMI's historic and constant plan—a plan to "affor[d] a unique educational benefit only to males." Ibid. However "liberally" this plan serves the Commonwealth's sons, it makes no provision whatever for her daughters. That is not equal protection.

B

Virginia next argues that VMI's adversative method of training provides educational benefits that cannot be made available, unmodified, to women. Alterations to accommodate women would necessarily be "radical," so "drastic," Virginia asserts, as to transform, indeed "destroy," VMI's program. See Brief for Cross-Petitioners 34-36. Neither sex would be favored by the transformation, Virginia maintains: Men would be deprived of the unique opportunity currently available to them; women would not gain that opportunity because their participation would "eliminat[e] the very aspects of [the] program that distinguish [VMI] from . . . other institutions of higher education in Virginia." Id., at 34.

The District Court forecast from expert witness testimony, and the Court of Appeals accepted, that coeducation would materially affect "at least these three aspects of VMI's program—physical training, the absence of privacy, and the adversative approach." 976 F. 2d, at 896-897. And it is uncontested that women's admission would require accommodations, primarily in arranging housing assignments and physical training programs for female cadets. See Brief for Cross-Respondent 11, 29-30. It is also undisputed, however, that "the VMI methodology could be used to educate women." 852 F. Supp., at 481. The District Court even allowed that some women may prefer it to the methodology a women's college might pursue. See ibid. "[S]ome women, at least, would want to attend [VMI] if they had the opportunity," the District Court recognized, 766 F. Supp., at 1414, and "some women," the expert testimony established, "are [541] capable of all of the individual activities required of VMI cadets," id., at 1412. The parties, furthermore, agree that "some women can meet the physical standards [VMI] now impose[s] on men." 976 F. 2d, at 896. In sum, as the Court of Appeals stated, "neither the goal of producing citizen soldiers," VMI's raison d'être, "nor VMI's implementing methodology is inherently unsuitable to women." Id., at 899.

In support of its initial judgment for Virginia, a judgment rejecting all equal protection objections presented by the United States, the District Court made "findings" on "gender-based developmental differences." 766 F. Supp., at 1434-1435. These "findings" restate the opinions of Virginia's expert witnesses, opinions about typically male or typically female "tendencies." Id., at 1434. For example, "[m]ales tend to need an atmosphere of adversativeness," while "[f]emales tend to thrive in a cooperative atmosphere." Ibid. "I'm not saying that some women don't do well under [the] adversative model," VMI's expert on educational institutions testified, "undoubtedly there are some [women] who do"; but educational experiences must be designed "around the rule," this expert maintained, and not "around the exception." Ibid. (internal quotation marks omitted).

The United States does not challenge any expert witness estimation on average capacities or preferences of men and women. Instead, the United States emphasizes that time and again since this Court's turning point decision in Reed v. Reed, 404 U. S. 71 (1971), we have cautioned reviewing courts to take a "hard look" at generalizations or "tendencies" of the kind pressed by Virginia, and relied upon by the District Court. See O'Connor, Portia's Progress, 66 N. Y. U. L. Rev. 1546, 1551 (1991). State actors controlling gates to opportunity, we have instructed, may not exclude qualified individuals based on "fixed notions concerning the roles and abilities of males and females." Mississippi Univ. for Women, 458 U. S., at 725; see J. E. B., 511 U. S., at 139, n. 11 (equal protection principles, as applied to gender classifications, mean [542] state actors may not rely on "overbroad" generalizations to make "judgments about people that are likely to . . . perpetuate historical patterns of discrimination").

It may be assumed, for purposes of this decision, that most women would not choose VMI's adversative method. As Fourth Circuit Judge Motz observed, however, in her dissent from the Court of Appeals' denial of rehearing en banc, it is also probable that "many men would not want to be educated in such an environment." 52 F. 3d, at 93. (On that point, even our dissenting colleague might agree.) Education, to be sure, is not a "one size fits all" business. The issue, however, is not whether "women—or men—should be forced to attend VMI"; rather, the question is whether the Commonwealth can constitutionally deny to women who have the will and capacity, the training and attendant opportunities that VMI uniquely affords. Ibid.

The notion that admission of women would downgrade VMI's stature, destroy the adversative system and, with it, even the school,^[13] is a judgment hardly proved,^[14] a prediction [543] hardly different from other "self-fulfilling prophec[ies]," see Mississippi Univ. for Women, 458 U. S., at 730, once routinely used to deny rights or opportunities. When women first sought admission to the bar and access to legal education, concerns of the same order were expressed. For example, in 1876, the Court of Common Pleas of Hennepin County, Minnesota, explained why women were thought ineligible for the practice of law. Women train and educate the young, the court said, which

"forbids that they shall bestow that time (early and late) and labor, so essential in attaining to the eminence to which the true lawyer should ever aspire. It cannot therefore be said that the opposition of courts to the admission of females to practice . . . is to any extent the outgrowth of . . . `old fogyism[.]' . . . [I]t arises rather from a comprehension of the magnitude of the responsibilities connected with the successful practice of law, and a desire to grade up the profession." In re Application of Martha Angle Dorsett to Be Admitted to Practice as Attorney and Counselor at Law (Minn. C. P. Hennepin Cty., 1876), in The Syllabi, Oct. 21, 1876, pp. 5, 6 (emphasis added).

A like fear, according to a 1925 report, accounted for Columbia Law School's resistance to women's admission, although

"[t]he faculty . . . never maintained that women could not master legal learning . . . . No, its argument has been . . . more practical. If women were admitted to [544] the Columbia Law School, [the faculty] said, then the choicer, more manly and red-blooded graduates of our great universities would go to the Harvard Law School!" The Nation, Feb. 18, 1925, p. 173.

Medical faculties similarly resisted men and women as partners in the study of medicine. See R. Morantz-Sanchez, Sympathy and Science: Women Physicians in American Medicine 51-54, 250 (1985); see also M. Walsh, "Doctors Wanted: No Women Need Apply" 121-122 (1977) (quoting E. Clarke, Medical Education of Women, 4 Boston Med. & Surg. J. 345, 346 (1869) ("`God forbid that I should ever see men and women aiding each other to display with the scalpel the secrets of the reproductive system . . . .' ")); cf. supra, at 536-537, n. 9. More recently, women seeking careers in policing encountered resistance based on fears that their presence would "undermine male solidarity," see F. Heidensohn, Women in Control? 201 (1992); deprive male partners of adequate assistance, see id. , at 184-185; and lead to sexual misconduct, see C. Milton et al., Women in Policing 32-33 (1974). Field studies did not confirm these fears. See Heidensohn, supra, at 92-93; P. Bloch & D. Anderson, Policewomen on Patrol: Final Report (1974).

Women's successful entry into the federal military academies,^[15] and their participation in the Nation's military forces,^[16] indicate that Virginia's fears for the future of VMI [545] may not be solidly grounded.^[17] The Commonwealth's justification for excluding all women from "citizen-soldier" training for which some are qualified, in any event, cannot rank as "exceedingly persuasive," as we have explained and applied that standard.

Virginia and VMI trained their argument on "means" rather than "end," and thus misperceived our precedent. Single-sex education at VMI serves an "important governmental objective," they maintained, and exclusion of women is not only "substantially related," it is essential to that objective. By this notably circular argument, the "straightforward" test Mississippi Univ. for Women described, see 458 U. S., at 724-725, was bent and bowed.

The Commonwealth's misunderstanding and, in turn, the District Court's, is apparent from VMI's mission: to produce "citizen-soldiers," individuals

"`imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready . . . to defend their country in time of national peril.' " 766 F. Supp., at 1425 (quoting Mission Study Committee of the VMI Board of Visitors, Report, May 16, 1986).

Surely that goal is great enough to accommodate women, who today count as citizens in our American democracy equal in stature to men. Just as surely, the Commonwealth's [546] great goal is not substantially advanced by women's categorical exclusion, in total disregard of their individual merit, from the Commonwealth's premier "citizen-soldier" corps.^[18] Virginia, in sum, "has fallen far short of establishing the `exceedingly persuasive justification,' " Mississippi Univ. for Women, 458 U. S., at 731, that must be the solid base for any gender-defined classification.

VI

In the second phase of the litigation, Virginia presented its remedial plan—maintain VMI as a male-only college and create VWIL as a separate program for women. The plan met District Court approval. The Fourth Circuit, in turn, deferentially reviewed the Commonwealth's proposal and decided that the two single-sex programs directly served Virginia's reasserted purposes: single-gender education, and "achieving the results of an adversative method in a military environment." See 44 F. 3d, at 1236, 1239. Inspecting the VMI and VWIL educational programs to determine whether they "afford[ed] to both genders benefits comparable in substance, [if] not in form and detail," id., at 1240, the Court of Appeals concluded that Virginia had arranged for men and women opportunities "sufficiently comparable" to survive equal protection evaluation, id., at 1240-1241. The United States challenges this "remedial" ruling as pervasively misguided.

[547] A

A remedial decree, this Court has said, must closely fit the constitutional violation; it must be shaped to place persons unconstitutionally denied an opportunity or advantage in "the position they would have occupied in the absence of [discrimination]." See Milliken v. Bradley, 433 U. S. 267, 280 (1977) (internal quotation marks omitted). The constitutional violation in this suit is the categorical exclusion of women from an extraordinary educational opportunity afforded men. A proper remedy for an unconstitutional exclusion, we have explained, aims to "eliminate [so far as possible] the discriminatory effects of the past" and to "bar like discrimination in the future." Louisiana v. United States, 380 U. S. 145, 154 (1965).

Virginia chose not to eliminate, but to leave untouched, VMI's exclusionary policy. For women only, however, Virginia proposed a separate program, different in kind from VMI and unequal in tangible and intangible facilities.^[19] Having violated the Constitution's equal protection requirement, Virginia was obliged to show that its remedial proposal "directly address[ed] and relate[d] to" the violation, see Milliken, 433 U. S., at 282, i. e., the equal protection denied to women ready, willing, and able to benefit from educational [548] opportunities of the kind VMI offers. Virginia described VWIL as a "parallel program," and asserted that VWIL shares VMI's mission of producing "citizen-soldiers" and VMI's goals of providing "education, military training, mental and physical discipline, character . . . and leadership development." Brief for Respondents 24 (internal quotation marks omitted). If the VWIL program could not "eliminate the discriminatory effects of the past," could it at least "bar like discrimination in the future"? See Louisiana, 380 U. S., at 154. A comparison of the programs said to be "parallel" informs our answer. In exposing the character of, and differences in, the VMI and VWIL programs, we recapitulate facts earlier presented. See supra, at 520-523, 526-527.

VWIL affords women no opportunity to experience the rigorous military training for which VMI is famed. See 766 F. Supp., at 1413-1414 ("No other school in Virginia or in the United States, public or private, offers the same kind of rigorous military training as is available at VMI."); id. , at 1421 (VMI "is known to be the most challenging military school in the United States"). Instead, the VWIL program "deemphasize[s]" military education, 44 F. 3d, at 1234, and uses a "cooperative method" of education "which reinforces self-esteem," 852 F. Supp., at 476.

VWIL students participate in ROTC and a "largely ceremonial" Virginia Corps of Cadets, see 44 F. 3d, at 1234, but Virginia deliberately did not make VWIL a military institute. The VWIL House is not a military-style residence and VWIL students need not live together throughout the 4-year program, eat meals together, or wear uniforms during the schoolday. See 852 F. Supp., at 477, 495. VWIL students thus do not experience the "barracks" life "crucial to the VMI experience," the spartan living arrangements designed to foster an "egalitarian ethic." See 766 F. Supp., at 1423-1424. "[T]he most important aspects of the VMI educational experience occur in the barracks," the District Court [549] found, id., at 1423, yet Virginia deemed that core experience nonessential, indeed inappropriate, for training its female citizen-soldiers.

VWIL students receive their "leadership training" in seminars, externships, and speaker series, see 852 F. Supp., at 477, episodes and encounters lacking the "[p]hysical rigor, mental stress, . .. minute regulation of behavior, and indoctrination in desirable values" made hallmarks of VMI's citizen-soldier training, see 766 F. Supp., at 1421.^[20] Kept away from the pressures, hazards, and psychological bonding characteristic of VMI's adversative training, see id., at 1422, VWIL students will not know the "feeling of tremendous accomplishment" commonly experienced by VMI's successful cadets, id. , at 1426.

Virginia maintains that these methodological differences are "justified pedagogically," based on "important differences between men and women in learning and developmental needs," "psychological and sociological differences" Virginia describes as "real" and "not stereotypes." Brief for Respondents 28 (internal quotation marks omitted). The Task Force charged with developing the leadership program for women, drawn from the staff and faculty at Mary Baldwin College, "determined that a military model and, especially VMI's adversative method, would be wholly inappropriate for educating and training most women. " 852 F. Supp., at 476 (emphasis added). See also 44 F. 3d, at 1233-1234 (noting Task Force conclusion that, while "some women would be suited to and interested in [a VMI-style experience]," VMI's adversative method "would not be effective for women as a group " (emphasis added)). The Commonwealth [550] embraced the Task Force view, as did expert witnesses who testified for Virginia. See 852 F. Supp., at 480-481.

As earlier stated, see supra, at 541-542, generalizations about "the way women are," estimates of what is appropriate for most women, no longer justify denying opportunity to women whose talent and capacity place them outside the average description. Notably, Virginia never asserted that VMI's method of education suits most men. It is also revealing that Virginia accounted for its failure to make the VWIL experience "the entirely militaristic experience of VMI" on the ground that VWIL "is planned for women who do not necessarily expect to pursue military careers." 852 F. Supp., at 478. By that reasoning, VMI's "entirely militaristic" program would be inappropriate for men in general or as a group, for "[o]nly about 15% of VMI cadets enter career military service." See 766 F. Supp., at 1432.

In contrast to the generalizations about women on which Virginia rests, we note again these dispositive realities: VMI's "implementing methodology" is not "inherently unsuitable to women," 976 F. 2d, at 899; "some women . . . do well under [the] adversative model," 766 F. Supp., at 1434 (internal quotation marks omitted); "some women, at least, would want to attend [VMI] if they had the opportunity," id., at 1414; "some women are capable of all of the individual activities required of VMI cadets," id., at 1412, and "can meet the physical standards [VMI] now impose[s] on men," 976 F. 2d, at 896. It is on behalf of these women that the United States has instituted this suit, and it is for them that a remedy must be crafted,^[21] a remedy that will end their [551] exclusion from a state-supplied educational opportunity for which they are fit, a decree that will "bar like discrimination in the future." Louisiana, 380 U. S., at 154.

B

In myriad respects other than military training, VWIL does not qualify as VMI's equal. VWIL's student body, faculty, course offerings, and facilities hardly match VMI's. Nor can the VWIL graduate anticipate the benefits associated with VMI's 157-year history, the school's prestige, and its influential alumni network.

Mary Baldwin College, whose degree VWIL students will gain, enrolls first-year women with an average combined SAT score about 100 points lower than the average score for VMI freshmen. 852 F. Supp., at 501. The Mary Baldwin faculty holds "significantly fewer Ph. D.'s," id. , at 502, and receives substantially lower salaries, see Tr. 158 (testimony of James Lott, Dean of Mary Baldwin College), than the faculty at VMI.

Mary Baldwin does not offer a VWIL student the range of curricular choices available to a VMI cadet. VMI awards baccalaureate degrees in liberal arts, biology, chemistry, civil engineering, electrical and computer engineering, and mechanical engineering. See 852 F. Supp., at 503; Virginia Military Institute: More than an Education 11 (Govt. exh. 75, [552] lodged with Clerk of this Court). VWIL students attend a school that "does not have a math and science focus," 852 F. Supp., at 503; they cannot take at Mary Baldwin any courses in engineering or the advanced math and physics courses VMI offers, see id., at 477.

For physical training, Mary Baldwin has "two multipurpose fields" and "[o]ne gymnasium." Id. , at 503. VMI has "an NCAA competition level indoor track and field facility; a number of multi-purpose fields; baseball, soccer and lacrosse fields; an obstacle course; large boxing, wrestling and martial arts facilities; an 11-laps-to-the-mile indoor running course; an indoor pool; indoor and outdoor rifle ranges; and a football stadium that also contains a practice field and outdoor track." Ibid.

Although Virginia has represented that it will provide equal financial support for in-state VWIL students and VMI cadets, id. , at 483, and the VMI Foundation has agreed to endow VWIL with $5.4625 million, id. , at 499, the difference between the two schools' financial reserves is pronounced. Mary Baldwin's endowment, currently about $19 million, will gain an additional $35 million based on future commitments; VMI's current endowment, $131 million—the largest public college per-student endowment in the Nation—will gain $220 million. Id. , at 503.

The VWIL student does not graduate with the advantage of a VMI degree. Her diploma does not unite her with the legions of VMI "graduates [who] have distinguished themselves" in military and civilian life. See 976 F. 2d, at 892-893. "[VMI] alumni are exceptionally close to the school," and that closeness accounts, in part, for VMI's success in attracting applicants. See 766 F. Supp., at 1421. A VWIL graduate cannot assume that the "network of business owners, corporations, VMI graduates and non-graduate employers . . . interested in hiring VMI graduates," 852 F. Supp., at 499, will be equally responsive to her search for employment, [553] see 44 F. 3d, at 1250 (Phillips, J., dissenting) ("the powerful political and economic ties of the VMI alumni network cannot be expected to open" for graduates of the fledgling VWIL program).

Virginia, in sum, while maintaining VMI for men only, has failed to provide any "comparable single-gender women's institution." Id. , at 1241. Instead, the Commonwealth has created a VWIL program fairly appraised as a "pale shadow" of VMI in terms of the range of curricular choices and faculty stature, funding, prestige, alumni support and influence. See id., at 1250 (Phillips, J., dissenting).

Virginia's VWIL solution is reminiscent of the remedy Texas proposed 50 years ago, in response to a state trial court's 1946 ruling that, given the equal protection guarantee, African-Americans could not be denied a legal education at a state facility. See Sweatt v. Painter, 339 U. S. 629 (1950). Reluctant to admit African-Americans to its flagship University of Texas Law School, the State set up a separate school for Heman Sweatt and other black law students. Id. , at 632. As originally opened, the new school had no independent faculty or library, and it lacked accreditation. Id. , at 633. Nevertheless, the state trial and appellate courts were satisfied that the new school offered Sweatt opportunities for the study of law "substantially equivalent to those offered by the State to white students at the University of Texas." Id. , at 632 (internal quotation marks omitted).

Before this Court considered the case, the new school had gained "a faculty of five full-time professors; a student body of 23; a library of some 16,500 volumes serviced by a full-time staff; a practice court and legal aid association; and one alumnus who ha[d] become a member of the Texas Bar." Id. , at 633. This Court contrasted resources at the new school with those at the school from which Sweatt had been excluded. The University of Texas Law School had a full-time faculty of 16, a student body of 850, a library containing over [554] 65,000 volumes, scholarship funds, a law review, and moot court facilities. Id. , at 632-633.

More important than the tangible features, the Court emphasized, are "those qualities which are incapable of objective measurement but which make for greatness" in a school, including "reputation of the faculty, experience of the administration, position and influence of the alumni, standing in the community, traditions and prestige." Id. , at 634. Facing the marked differences reported in the Sweatt opinion, the Court unanimously ruled that Texas had not shown "substantial equality in the [separate] educational opportunities" the State offered. Id. , at 633. Accordingly, the Court held, the Equal Protection Clause required Texas to admit AfricanAmericans to the University of Texas Law School. Id. , at 636. In line with Sweatt, we rule here that Virginia has not shown substantial equality in the separate educational opportunities the Commonwealth supports at VWIL and VMI.

C

When Virginia tendered its VWIL plan, the Fourth Circuit did not inquire whether the proposed remedy, approved by the District Court, placed women denied the VMI advantage in "the position they would have occupied in the absence of [discrimination]." Milliken, 433 U. S., at 280 (internal quotation marks omitted). Instead, the Court of Appeals considered whether the Commonwealth could provide, with fidelity to the equal protection principle, separate and unequal educational programs for men and women.

The Fourth Circuit acknowledged that "the VWIL degree from Mary Baldwin College lacks the historical benefit and prestige of a degree from VMI." 44 F. 3d, at 1241. The Court of Appeals further observed that VMI is "an ongoing and successful institution with a long history," and there remains no "comparable single-gender women's institution." Ibid. Nevertheless, the appeals court declared the substantially different and significantly unequal VWIL program satisfactory. [555] The court reached that result by revising the applicable standard of review. The Fourth Circuit displaced the standard developed in our precedent, see supra, at 532-534, and substituted a standard of its own invention.

We have earlier described the deferential review in which the Court of Appeals engaged, see supra, at 528-529, a brand of review inconsistent with the more exacting standard our precedent requires, see supra, at 532-534. Quoting in part from Mississippi Univ. for Women, the Court of Appeals candidly described its own analysis as one capable of checking a legislative purpose ranked as "pernicious," but generally according "deference to [the] legislative will." 44 F. 3d, at 1235, 1236. Recognizing that it had extracted from our decisions a test yielding "little or no scrutiny of the effect of a classification directed at [single-gender education]," the Court of Appeals devised another test, a "substantive comparability" inquiry, id., at 1237, and proceeded to find that new test satisfied, id. , at 1241.

The Fourth Circuit plainly erred in exposing Virginia's VWIL plan to a deferential analysis, for "all gender-based classifications today" warrant "heightened scrutiny." See J. E. B., 511 U. S., at 136. Valuable as VWIL may prove for students who seek the program offered, Virginia's remedy affords no cure at all for the opportunities and advantages withheld from women who want a VMI education and can make the grade. See supra, at 549-554.^[22] In sum, Virginia's [556] remedy does not match the constitutional violation; the Commonwealth has shown no "exceedingly persuasive justification" for withholding from women qualified for the experience premier training of the kind VMI affords.

VII

A generation ago, "the authorities controlling Virginia higher education," despite long established tradition, agreed "to innovate and favorably entertain[ed] the [then] relatively new idea that there must be no discrimination by sex in offering educational opportunity." Kirstein, 309 F. Supp., at 186. Commencing in 1970, Virginia opened to women "educational opportunities at the Charlottesville campus that [were] not afforded in other [state-operated] institutions." Id., at 187; see supra, at 538. A federal court approved the Commonwealth's innovation, emphasizing that the University of Virginia "offer[ed] courses of instruction . . . not available elsewhere." 309 F. Supp., at 187. The court further noted: "[T]here exists at Charlottesville a `prestige' factor [557] [not paralleled in] other Virginia educational institutions." Ibid.

VMI, too, offers an educational opportunity no other Virginia institution provides, and the school's "prestige"—associated with its success in developing "citizen-soldiers"—is unequaled. Virginia has closed this facility to its daughters and, instead, has devised for them a "parallel program," with a faculty less impressively credentialed and less well paid, more limited course offerings, fewer opportunities for military training and for scientific specialization. Cf. Sweatt, 339 U. S., at 633. VMI, beyond question, "possesses to a far greater degree" than the VWIL program "those qualities which are incapable of objective measurement but which make for greatness in a . . . school," including "position and influence of the alumni, standing in the community, traditions and prestige." Id., at 634. Women seeking and fit for a VMI-quality education cannot be offered anything less, under the Commonwealth's obligation to afford them genuinely equal protection.

A prime part of the history of our Constitution, historian Richard Morris recounted, is the story of the extension of constitutional rights and protections to people once ignored or excluded.^[23] VMI's story continued as our comprehension of "We the People" expanded. See supra, at 532, n. 6. [558] There is no reason to believe that the admission of women capable of all the activities required of VMI cadets would destroy the Institute rather than enhance its capacity to serve the "more perfect Union."

* * *

For the reasons stated, the initial judgment of the Court of Appeals, 976 F. 2d 890 (CA4 1992), is affirmed, the final judgment of the Court of Appeals, 44 F. 3d 1229 (CA4 1995), is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Justice Thomas took no part in the consideration or decision of these cases.

Chief Justice Rehnquist, concurring in the judgment.

The Court holds first that Virginia violates the Equal Protection Clause by maintaining the Virginia Military Institute's (VMI's) all-male admissions policy, and second that establishing the Virginia Women's Institute for Leadership (VWIL) program does not remedy that violation. While I agree with these conclusions, I disagree with the Court's analysis and so I write separately.

I

Two decades ago in Craig v. Boren, 429 U. S. 190, 197 (1976), we announced that "[t]o withstand constitutional challenge, . . . classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives." We have adhered to that standard of scrutiny ever since. See Califano v. Goldfarb, 430 U. S. 199, 210-211 (1977); Califano v. Webster, 430 U. S. 313, 316-317 (1977); Orr v. Orr, 440 U. S. 268, 279 (1979); Caban v. Mohammed, 441 U. S. 380, 388 (1979); Davis v. Passman, 442 U. S. 228, 234-235, 235, n. 9 (1979); Personnel Administrator of Mass. v. Feeney, 442 U. S. 256, 273 (1979); [559] Califano v. Westcott, 443 U. S. 76, 85 (1979); Wengler v. Druggists Mut. Ins. Co., 446 U. S. 142, 150 (1980); Kirchberg v. Feenstra, 450 U. S. 455, 459-460 (1981); Michael M. v. Superior Court, Sonoma Cty., 450 U. S. 464, 469 (1981); Mississippi Univ. for Women v. Hogan, 458 U. S. 718, 724 (1982); Heckler v. Mathews, 465 U. S. 728, 744 (1984); J. E. B. v. Alabama ex rel. T. B., 511 U. S. 127, 137, n. 6 (1994). While the majority adheres to this test today, ante, at 524, 533, it also says that the Commonwealth must demonstrate an "`exceedingly persuasive justification' " to support a gender-based classification. See ante, at 524, 529, 530, 531, 533, 534, 545, 546, 556. It is unfortunate that the Court thereby introduces an element of uncertainty respecting the appropriate test.

While terms like "important governmental objective" and "substantially related" are hardly models of precision, they have more content and specificity than does the phrase "exceedingly persuasive justification." That phrase is best confined, as it was first used, as an observation on the difficulty of meeting the applicable test, not as a formulation of the test itself. See, e. g., Feeney, supra, at 273 ("[T]hese precedents dictate that any state law overtly or covertly designed to prefer males over females in public employment require an exceedingly persuasive justification"). To avoid introducing potential confusion, I would have adhered more closely to our traditional, "firmly established," Hogan, supra, at 723; Heckler, supra, at 744, standard that a gender-based classification "must bear a close and substantial relationship to important governmental objectives." Feeney, supra, at 273.

Our cases dealing with gender discrimination also require that the proffered purpose for the challenged law be the actual purpose. See ante, at 533, 535-536. It is on this ground that the Court rejects the first of two justifications Virginia offers for VMI's single-sex admissions policy, namely, the goal of diversity among its public educational institutions. While I ultimately agree that the Commonwealth [560] has not carried the day with this justification, I disagree with the Court's method of analyzing the issue.

VMI was founded in 1839, and, as the Court notes, ante, at 536-537, admission was limited to men because under the then-prevailing view men, not women, were destined for higher education. However misguided this point of view may be by present-day standards, it surely was not unconstitutional in 1839. The adoption of the Fourteenth Amendment, with its Equal Protection Clause, was nearly 30 years in the future. The interpretation of the Equal Protection Clause to require heightened scrutiny for gender discrimination was yet another century away.

Long after the adoption of the Fourteenth Amendment, and well into this century, legal distinctions between men and women were thought to raise no question under the Equal Protection Clause. The Court refers to our decision in Goesaert v. Cleary, 335 U. S. 464 (1948). Likewise representing that now abandoned view was Hoyt v. Florida, 368 U. S. 57 (1961), where the Court upheld a Florida system of jury selection in which men were automatically placed on jury lists, but women were placed there only if they expressed an affirmative desire to serve. The Court noted that despite advances in women's opportunities, the "woman is still regarded as the center of home and family life." Id. , at 62.

Then, in 1971, we decided Reed v. Reed, 404 U. S. 71, which the Court correctly refers to as a seminal case. But its facts have nothing to do with admissions to any sort of educational institution. An Idaho statute governing the administration of estates and probate preferred men to women if the other statutory qualifications were equal. The statute's purpose, according to the Idaho Supreme Court, was to avoid hearings to determine who was better qualified as between a man and a woman both applying for letters of administration. This Court held that such a rule violated the Fourteenth Amendment because "a mandatory preference to members of either [561] sex over members of the other, merely to accomplish the elimination of hearings," was an "arbitrary legislative choice forbidden by the Equal Protection Clause." Id. , at 76. The brief opinion in Reed made no mention of either Goesaert or Hoyt.

Even at the time of our decision in Reed v. Reed, therefore, Virginia and VMI were scarcely on notice that its holding would be extended across the constitutional board. They were entitled to believe that "one swallow doesn't make a summer" and await further developments. Those developments were 11 years in coming. In Mississippi Univ. for Women v. Hogan, supra, a case actually involving a singlesex admissions policy in higher education, the Court held that the exclusion of men from a nursing program violated the Equal Protection Clause. This holding did place Virginia on notice that VMI's men-only admissions policy was open to serious question.

The VMI Board of Visitors, in response, appointed a Mission Study Committee to examine "the legality and wisdom of VMI's single-sex policy in light of" Hogan. 766 F. Supp. 1407, 1427 (WD Va. 1991). But the committee ended up cryptically recommending against changing VMI's status as a single-sex college. After three years of study, the committee found "`no information' " that would warrant a change in VMI's status. Id. , at 1429. Even the District Court, ultimately sympathetic to VMI's position, found that "[t]he Report provided very little indication of how [its] conclusion was reached" and that "[t]he one and one-half pages in the committee's final report devoted to analyzing the information it obtained primarily focuses on anticipated difficulties in attracting females to VMI." Ibid. The reasons given in the report for not changing the policy were the changes that admission of women to VMI would require, and the likely effect of those changes on the institution. That VMI would have to change is simply not helpful in addressing the constitutionality of the status after Hogan.

[562] Before this Court, Virginia has sought to justify VMI's single-sex admissions policy primarily on the basis that diversity in education is desirable, and that while most of the public institutions of higher learning in the Commonwealth are coeducational, there should also be room for single-sex institutions. I agree with the Court that there is scant evidence in the record that this was the real reason that Virginia decided to maintain VMI as men only.^[24] But, unlike the majority, I would consider only evidence that postdates our decision in Hogan, and would draw no negative inferences from the Commonwealth's actions before that time. I think that after Hogan, the Commonwealth was entitled to reconsider its policy with respect to VMI, and not to have earlier justifications, or lack thereof, held against it.

Even if diversity in educational opportunity were the Commonwealth's actual objective, the Commonwealth's position would still be problematic. The difficulty with its position is that the diversity benefited only one sex; there was single-sex public education available for men at VMI, but no corresponding single-sex public education available for women. When Hogan placed Virginia on notice that [563] VMI's admissions policy possibly was unconstitutional, VMI could have dealt with the problem by admitting women; but its governing body felt strongly that the admission of women would have seriously harmed the institution's educational approach. Was there something else the Commonwealth could have done to avoid an equal protection violation? Since the Commonwealth did nothing, we do not have to definitively answer that question.

I do not think, however, that the Commonwealth's options were as limited as the majority may imply. The Court cites, without expressly approving it, a statement from the opinion of the dissenting judge in the Court of Appeals, to the effect that the Commonwealth could have "simultaneously opened single-gender undergraduate institutions having substantially comparable curricular and extra-curricular programs, funding, physical plant, administration and support services, and faculty and library resources." Ante, at 529-530 (internal quotation marks omitted). If this statement is thought to exclude other possibilities, it is too stringent a requirement. VMI had been in operation for over a century and a half, and had an established, successful, and devoted group of alumni. No legislative wand could instantly call into existence a similar institution for women; and it would be a tremendous loss to scrap VMI's history and tradition. In the words of Grover Cleveland's second inaugural address, the Commonwealth faced a condition, not a theory. And it was a condition that had been brought about, not through defiance of decisions construing gender bias under the Equal Protection Clause, but, until the decision in Hogan, a condition that had not appeared to offend the Constitution. Had Virginia made a genuine effort to devote comparable public resources to a facility for women, and followed through on such a plan, it might well have avoided an equal protection violation. I do not believe the Commonwealth was faced with the stark choice of either admitting women to VMI, on the [564] one hand, or abandoning VMI and starting from scratch for both men and women, on the other.

But, as I have noted, neither the governing board of VMI nor the Commonwealth took any action after 1982. If diversity in the form of single-sex, as well as coeducational, institutions of higher learning were to be available to Virginians, that diversity had to be available to women as well as to men.

The dissent criticizes me for "disregarding the four allwomen's private colleges in Virginia (generously assisted by public funds)." Post, at 595. The private women's colleges are treated by the Commonwealth exactly as all other private schools are treated, which includes the provision of tuition-assistance grants to Virginia residents. Virginia gives no special support to the women's single-sex education. But obviously, the same is not true for men's education. Had the Commonwealth provided the kind of support for the private women's schools that it provides for VMI, this may have been a very different case. For in so doing, the Commonwealth would have demonstrated that its interest in providing a single-sex education for men was to some measure matched by an interest in providing the same opportunity for women.

Virginia offers a second justification for the single-sex admissions policy: maintenance of the adversative method. I agree with the Court that this justification does not serve an important governmental objective. A State does not have substantial interest in the adversative methodology unless it is pedagogically beneficial. While considerable evidence shows that a single-sex education is pedagogically beneficial for some students, see 766 F. Supp., at 1414, and hence a State may have a valid interest in promoting that methodology, there is no similar evidence in the record that an adversative method is pedagogically beneficial or is any more likely to produce character traits than other methodologies.

[565] II

The Court defines the constitutional violation in these cases as "the categorical exclusion of women from an extraordinary educational opportunity afforded to men." Ante, at 547. By defining the violation in this way, and by emphasizing that a remedy for a constitutional violation must place the victims ofdiscrimination in "`the position they would have occupied in the absence of [discrimination],' " ibid. , the Court necessarily implies that the only adequate remedy would be the admission of women to the allmale institution. As the foregoing discussion suggests, I would not define the violation in this way; it is not the "exclusion of women" that violates the Equal Protection Clause, but the maintenance of an all-men school without providing any—much less a comparable—institution for women.

Accordingly, the remedy should not necessarily require either the admission of women to VMI or the creation of a VMI clone for women. An adequate remedy in my opinion might be a demonstration by Virginia that its interest in educating men in a single-sex environment is matched by its interest in educating women in a single-sex institution. To demonstrate such, the Commonwealth does not need to create two institutions with the same number of faculty Ph. D.'s, similar SAT scores, or comparable athletic fields. See ante, at 551-552. Nor would it necessarily require that the women's institution offer the same curriculum as the men's; one could be strong in computer science, the other could be strong in liberal arts. It would be a sufficient remedy, I think, if the two institutions offered the same quality of education and were of the same overall caliber.

If a State decides to create single-sex programs, the State would, I expect, consider the public's interest and demand in designing curricula. And rightfully so. But the State should avoid assuming demand based on stereotypes; it must not assume a priori, without evidence, that there would be [566] no interest in a women's school of civil engineering, or in a men's school of nursing.

In the end, the women's institution Virginia proposes, VWIL, fails as a remedy, because it is distinctly inferior to the existing men's institution and will continue to be for the foreseeable future. VWIL simply is not, in any sense, the institution that VMI is. In particular, VWIL is a program appended to a private college, not a self-standing institution; and VWIL is substantially underfunded as compared to VMI. I therefore ultimately agree with the Court that Virginia has not provided an adequate remedy.

Justice Scalia, dissenting.

Today the Court shuts down an institution that has served the people of the Commonwealth of Virginia with pride and distinction for over a century and a half. To achieve that desired result, it rejects (contrary to our established practice) the factual findings of two courts below, sweeps aside the precedents of this Court, and ignores the history of our people. As to facts: It explicitly rejects the finding that there exist "gender-based developmental differences" supporting Virginia's restriction of the "adversative" method to only a men's institution, and the finding that the all-male composition of the Virginia Military Institute (VMI) is essential to that institution's character. As to precedent: It drastically revises our established standards for reviewing sex-based classifications. And as to history: It counts for nothing the long tradition, enduring down to the present, of men's military colleges supported by both States and the Federal Government.

Much of the Court's opinion is devoted to deprecating the closed-mindedness of our forebears with regard to women's education, and even with regard to the treatment of women in areas that have nothing to do with education. Closedminded they were—as every age is, including our own, with regard to matters it cannot guess, because it simply does not [567] consider them debatable. The virtue of a democratic system with a First Amendment is that it readily enables the people, over time, to be persuaded that what they took for granted is not so, and to change their laws accordingly. That system is destroyed if the smug assurances of each age are removed from the democratic process and written into the Constitution. So to counterbalance the Court's criticism of our ancestors, let me say a word in their praise: They left us free to change. The same cannot be said of this most illiberal Court, which has embarked on a course of inscribing one after another of the current preferences of the society (and in some cases only the countermajoritarian preferences of the society's law-trained elite) into our Basic Law. Today it enshrines the notion that no substantial educational value is to be served by an all-men's military academy—so that the decision by the people of Virginia to maintain such an institution denies equal protection to women who cannot attend that institution but can attend others. Since it is entirely clear that the Constitution of the United States—the old one—takes no sides in this educational debate, I dissent.

I

I shall devote most of my analysis to evaluating the Court's opinion on the basis of our current equal protection jurisprudence, which regards this Court as free to evaluate everything under the sun by applying one of three tests: "rational basis" scrutiny, intermediate scrutiny, or strict scrutiny. These tests are no more scientific than their names suggest, and a further element of randomness is added by the fact that it is largely up to us which test will be applied in each case. Strict scrutiny, we have said, is reserved for state "classifications based on race or national origin and classifications affecting fundamental rights," Clark v. Jeter, 486 U. S. 456, 461 (1988) (citation omitted). It is my position that the term "fundamental rights" should be limited to "interest[s] traditionally protected by our society," Michael H. [568] v. Gerald D., 491 U. S. 110, 122 (1989) (plurality opinion of Scalia, J.); but the Court has not accepted that view, so that strict scrutiny will be applied to the deprivation of whatever sort of right we consider "fundamental." We have no established criterion for "intermediate scrutiny" either, but essentially apply it when it seems like a good idea to load the dice. So far it has been applied to content-neutral restrictions that place an incidental burden on speech, to disabilities attendant to illegitimacy, and to discrimination on the basis of sex. See, e. g., Turner Broadcasting System, Inc. v. FCC, 512 U. S. 622, 662 (1994); Mills v. Habluetzel, 456 U. S. 91, 98-99 (1982); Craig v. Boren, 429 U. S. 190, 197 (1976).

I have no problem with a system of abstract tests such as rational basis, intermediate, and strict scrutiny (though I think we can do better than applying strict scrutiny and intermediate scrutiny whenever we feel like it). Such formulas are essential to evaluating whether the new restrictions that a changing society constantly imposes upon private conduct comport with that "equal protection" our society has always accorded in the past. But in my view the function of this Court is to preserve our society's values regarding (among other things) equal protection, not to revise them; to prevent backsliding from the degree of restriction the Constitution imposed upon democratic government, not to prescribe, on our own authority, progressively higher degrees. For that reason it is my view that, whatever abstract tests we may choose to devise, they cannot supersede—and indeed ought to be crafted so as to reflect —those constant and unbroken national traditions that embody the people's understanding of ambiguous constitutional texts. More specifically, it is my view that "when a practice not expressly prohibited by the text of the Bill of Rights bears the endorsement of a long tradition of open, widespread, and unchallenged use that dates back to the beginning of the Republic, we have no proper basis for striking it down." Rutan v. Republican Party of Ill., 497 U. S. 62, 95 (1990) (Scalia, J., [569] dissenting). The same applies, mutatis mutandis, to a practice asserted to be in violation of the post-Civil War Fourteenth Amendment. See, e. g., Burnham v. Superior Court of Cal., County of Marin, 495 U. S. 604 (1990) (plurality opinion of Scalia, J.) (Due Process Clause); J. E. B. v. Alabama ex rel. T. B., 511 U. S. 127, 156-163 (1994) (Scalia, J., dissenting) (Equal Protection Clause); Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 979-984, 1000-1001 (1992) (Scalia, J., dissenting) (various alleged "penumbras").

The all-male constitution of VMI comes squarely within such a governing tradition. Founded by the Commonwealth of Virginia in 1839 and continuously maintained by it since, VMI has always admitted only men. And in that regard it has not been unusual. For almost all of VMI's more than a century and a half of existence, its single-sex status reflected the uniform practice for government-supported military colleges. Another famous Southern institution, The Citadel, has existed as a state-funded school of South Carolina since 1842. And all the federal military colleges—West Point, the Naval Academy at Annapolis, and even the Air Force Academy, which was not established until 1954—admitted only males for most of their history. Their admission of women in 1976 (upon which the Court today relies, see ante, at 544-545, nn. 13, 15) came not by court decree, but because the people, through their elected representatives, decreed a change. See, e. g., § 803(a), 89 Stat. 537, note following 10 U. S. C. § 4342. In other words, the tradition of having government-funded military schools for men is as well rooted in the traditions of this country as the tradition of sending only men into military combat. The people may decide to change the one tradition, like the other, through democratic processes; but the assertion that either tradition has been unconstitutional through the centuries is not law, but politics-smuggled-into-law.

And the same applies, more broadly, to single-sex education in general, which, as I shall discuss, is threatened by [570] today's decision with the cutoff of all state and federal support. Government-run non military educational institutions for the two sexes have until very recently also been part of our national tradition. "[It is] [c]oeducation, historically, [that] is a novel educational theory. From grade school through high school, college, and graduate and professional training, much of the Nation's population during much of our history has been educated in sexually segregated classrooms." Mississippi Univ. for Women v. Hogan, 458 U. S. 718, 736 (1982) (Powell, J.,dissenting); see id., at 736-739. These traditions may of course be changed by the democratic decisions of the people, as they largely have been.

Today, however, change is forced upon Virginia, and reversion to single-sex education is prohibited nationwide, not by democratic processes but by order of this Court. Even while bemoaning the sorry, bygone days of "fixed notions" concerning women's education, see ante, at 536-537, and n. 10, 537-539, 542-544, the Court favors current notions so fixedly that it is willing to write them into the Constitution of the United States by application of custom-built "tests." This is not the interpretation of a Constitution, but the creation of one.

II

To reject the Court's disposition today, however, it is not necessary to accept my view that the Court's made-up tests cannot displace longstanding national traditions as the primary determinant of what the Constitution means. It is only necessary to apply honestly the test the Court has been applying to sex-based classifications for the past two decades. It is well settled, as Justice O'Connor stated some time ago for a unanimous Court, that we evaluate a statutory classification based on sex under a standard that lies "[b]etween th[e] extremes of rational basis review and strict scrutiny." Clark v. Jeter, 486 U. S., at 461. We have denominated this standard "intermediate scrutiny" and under it have inquired whether the statutory classification is "substantially [571] related to an important governmental objective." Ibid. See, e. g., Heckler v. Mathews, 465 U. S. 728, 744 (1984); Wengler v. Druggists Mut. Ins. Co., 446 U. S. 142, 150 (1980); Craig v. Boren, 429 U. S., at 197.

Before I proceed to apply this standard to VMI, I must comment upon the manner in which the Court avoids doing so. Notwithstanding our above-described precedents and their "`firmly established principles,' " Heckler, supra, at 744 (quoting Hogan, supra, at 723), the United States urged us to hold in this litigation "that strict scrutiny is the correct constitutional standard for evaluating classifications that deny opportunities to individuals based on their sex." Brief for United States in No. 94-2107, p. 16. (This was in flat contradiction of the Government's position below, which was, in its own words, to "stat[e] unequivocally that the appropriate standard in this case is `intermediate scrutiny.' " 2 Record, Doc. No. 88, p. 3 (emphasis added).) The Court, while making no reference to the Government's argument, effectively accepts it.

Although the Court in two places recites the test as stated in Hogan, see ante, at 524, 532-533, which asks whether the State has demonstrated "that the classification serves important governmental objectives and that the discriminatory means employed are substantially related to the achievement of those objectives," 458 U. S., at 724 (internal quotation marks omitted), the Court never answers the question presented in anything resembling that form. When it engages in analysis, the Court instead prefers the phrase "exceedingly persuasive justification" from Hogan. The Court's nine invocations of that phrase, see ante, at 524, 529, 530, 531, 533, 534, 545, 546, 556, and even its fanciful description of that imponderable as "the core instruction" of the Court's decisions in J. E. B. v. Alabama ex rel. T. B., supra, and Hogan, supra, see ante, at 531, would be unobjectionable if the Court acknowledged that whether a "justification" is "exceedingly persuasive" must be assessed by asking [572] "[whether] the classification serves important governmental objectives and [whether] the discriminatory means employed are substantially related to the achievement of those objectives." Instead, however, the Court proceeds to interpret "exceedingly persuasive justification" in a fashion that contradicts the reasoning of Hogan and our other precedents.

That is essential to the Court's result, which can only be achieved by establishing that intermediate scrutiny is not survived if there are some women interested in attending VMI, capable of undertaking its activities, and able to meet its physical demands. Thus, the Court summarizes its holding as follows:

"In contrast to the generalizations about women on which Virginia rests, we note again these dispositive realities: VMI's implementing methodology is not inher- ently unsuitable to women; some women do well under the adversative model; some women, at least, would want to attend VMI if they had the opportunity; some women are capable of all of the individual activities required of VMI cadets and can meet the physical standards VMI now imposes on men." Ante, at 550 (internal quotation marks, citations, and punctuation omitted; emphasis added).

Similarly, the Court states that "[t]he Commonwealth's justification for excluding all women from `citizen-soldier' training for which some are qualified . . . cannot rank as `exceedingly persuasive' . . . ." Ante, at 545.^[25]

[573] Only the amorphous "exceedingly persuasive justification" phrase, and not the standard elaboration of intermediate scrutiny, can be made to yield this conclusion that VMI's single-sex composition is unconstitutional because there exist several women (or, one would have to conclude under the Court's reasoning, a single woman) willing and able to undertake VMI's program. Intermediate scrutiny has never required a least-restrictive-means analysis, but only a "substantial relation" between the classification and the state interests that it serves. Thus, in Califano v. Webster, 430 U. S. 313 (1977) (per curiam), we upheld a congressional statute that provided higher Social Security benefits for women than for men. We reasoned that "women . . . as such have been unfairly hindered from earning as much as men," but we did not require proof that each woman so benefited had suffered discrimination or that each disadvantaged man had not; it was sufficient that even under the former congressional scheme "women on the average received lower retirement benefits than men." Id., at 318, and n. 5 (emphasis added). The reasoning in our other intermediate-scrutiny cases has similarly required only a substantial relation between end and means, not a perfect fit. In Rostker v. Goldberg, 453 U. S. 57 (1981), we held that selective-service registration could constitutionally exclude women, because even "assuming that a small number of women could be drafted for noncombat roles, Congress simply did not consider it worth the added burdens of including women in draft and registration plans." Id., at 81. In Metro Broadcasting, Inc. v. FCC, 497 U. S. 547, 579, 582-583 (1990), overruled on other grounds, Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 227 (1995), we held that a classification need not be accurate "in every case" to survive intermediate scrutiny so long as, "in the aggregate," it advances the underlying [574] objective. There issimply no support in our cases for the notion that a sex-based classification is invalid unless it relates to characteristics that hold true in every instance.

Not content to execute a de facto abandonment of the intermediate scrutiny that has been our standard for sex-based classifications for some two decades, the Court purports to reserve the question whether, even in principle, a higher standard (i. e., strict scrutiny) should apply. "The Court has," it says, "thus far reserved most stringent judicial scrutiny for classifications based on race or national origin . . . ," ante, at 532, n. 6 (emphasis added); and it describes our earlier cases as having done no more than decline to "equat[e] gender classifications, for all purposes, to classifications based on race or national origin," ante, at 532 (emphasis added). The wonderful thing about these statements is that they are not actually false—just as it would not be actually false to say that "our cases have thus far reserved the `beyond a reasonable doubt' standard of proof for criminal cases," or that "we have not equated tort actions, for all purposes, to criminal prosecutions." But the statements are misleading, insofar as they suggest that we have not already categorically held strict scrutiny to be inapplicable to sexbased classifications. See, e. g., Heckler v. Mathews, 465 U. S. 728 (1984) (upholding state action after applying only intermediate scrutiny); Michael M. v. Superior Court, Somoma Cty., 450 U. S. 464 (1981) (plurality and both concurring opinions) (same); Califano v. Webster, supra (per curiam) (same). And the statements are irresponsible, insofar as they are calculated to destabilize current law. Our task is to clarify the law—not to muddy the waters, and not to exact overcompliance by intimidation. The States and the Federal Government are entitled to know before they act the standard to which they will be held, rather than be compelled to guess about the outcome of Supreme Court peek-a-boo.

The Court's intimations are particularly out of place because it is perfectly clear that, if the question of the applicable [575] standard of review for sex-based classifications were to be regarded as an appropriate subject for reconsideration, the stronger argument would be not for elevating the standard to strict scrutiny, but for reducing it to rational-basis review. The latter certainly has a firmer foundation in our past jurisprudence: Whereas no majority of the Court has ever applied strict scrutiny in a case involving sex-based classifications, we routinely applied rational-basis review until the 1970's, see, e. g., Hoyt v. Florida, 368 U. S. 57 (1961); Goesaert v. Cleary, 335 U. S. 464 (1948). And of course normal, rational-basis review of sex-based classifications would be much more in accord with the genesis of heightened standards of judicial review, the famous footnote in United States v. Carolene Products Co., 304 U. S. 144 (1938), which said (intimatingly) that we did not have to inquire in the case at hand

"whether prejudice against discrete and insular minorities may be a special condition, which tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching judicial inquiry." Id., at 152-153, n. 4.

It is hard to consider women a "discrete and insular minorit[y]" unable to employ the "political processes ordinarily to be relied upon," when they constitute a majority of the electorate. And the suggestion that they are incapable of exerting that political power smacks of the same paternalism that the Court so roundly condemns. See, e. g., ante, at 536-537, 542-546 (and accompanying notes). Moreover, a long list of legislation proves the proposition false. See, e. g., Equal Pay Act of 1963, 29 U. S. C. § 206(d); Title VII of the Civil Rights Act of 1964, 42 U. S. C. § 2000e-2; Title IX of the Education Amendments of 1972, 20 U. S. C. § 1681; Women's Business Ownership Act of 1988, Pub. L. 100-533, 102 Stat. 2689; [576] Violence Against Women Act of 1994, Pub. L. 103-322, Title IV, 108 Stat. 1902.

III

With this explanation of how the Court has succeeded in making its analysis seem orthodox—and indeed, if intimations are to be believed, even overly generous to VMI—I now proceed to describe how the analysis should have been conducted. The question to be answered, I repeat, is whether the exclusion of women from VMI is "substantially related to an important governmental objective."

A

It is beyond question that Virginia has an important state interest in providing effective college education for itscitizens. That single-sex instruction is an approach substantially related to that interest should be evident enough from the long and continuing history in this country of men's and women's colleges. But beyond that, as the Court of Appeals here stated: "That single-gender education at the college level is beneficial to both sexes is a fact established in this case. " 44 F. 3d 1229, 1238 (CA4 1995) (emphasis added).

The evidence establishing that fact was overwhelming— indeed, "virtually uncontradicted" in the words of the court that received the evidence, 766 F. Supp. 1407, 1415 (WD Va. 1991). As an initial matter, Virginia demonstrated at trial that "[a] substantial body of contemporary scholarship and research supports the proposition that, although males and females have significant areas of developmental overlap, they also have differing developmental needs that are deepseated." Id., at 1434. While no one questioned that for many students a coeducational environment was nonetheless not inappropriate, that could not obscure the demonstrated benefits of single-sex colleges. For example, the District Court stated as follows:

"One empirical study in evidence, not questioned by any expert, demonstrates that single-sex colleges provide [577] better educational experiences than coeducational institutions. Students of both sexes become more academically involved, interact with faculty frequently, show larger increases in intellectual self-esteem and are more satisfied with practically all aspects of college experience (the sole exception is social life) compared with their counterparts in coeducational institutions. Attendance at an all-male college substantially increases the likelihood that a student will carry out career plans in law, business and college teaching, and also has a substantial positive effect on starting salaries in business. Women's colleges increase the chances that those who attend will obtain positions of leadership, complete the baccalaureate degree, and aspire to higher degrees." Id., at 1412.

See also id., at 1434-1435 (factual findings). "[I]n the light of this very substantial authority favoring single-sex education," the District Court concluded that "the VMI Board's decision to maintain an all-male institution is fully justified even without taking into consideration the other unique features of VMI's teaching and training." Id., at 1412. This finding alone, which even this Court cannot dispute, see ante, at 535, should be sufficient to demonstrate the constitutionality of VMI's all-male composition.

But besides its single-sex constitution, VMI is different from other colleges in another way. It employs a "distinctive educational method," sometimes referred to as the "adversative, or doubting, model of education." 766 F. Supp., at 1413, 1421. "Physical rigor, mental stress, absolute equality of treatment, absence of privacy, minute regulation of behavior, and indoctrination in desirable values are the salient attributes of the VMI educational experience." Id., at 1421. No one contends that this method is appropriate for all individuals; education is not a "one size fits all" business. Just as a State may wish to support junior colleges, vocational institutes, or a law school that emphasizes case [578] practice instead of classroom study, so too a State's decision to maintain within its system one school that provides the adversative method is "substantially related" to its goal of good education. Moreover, it was uncontested that "if the state were to establish a women's VMI-type [i. e., adversative] program, the program would attract an insufficient number of participants to make the program work," 44 F. 3d, at 1241; and it was found by the District Court that if Virginia were to include women in VMI, the school "would eventually find it necessary to drop the adversative system altogether," 766 F. Supp., at 1413. Thus, Virginia's options were an adversative method that excludes women or no adversative method at all.

There can be no serious dispute that, as the District Court found, single-sex education and a distinctive educational method "represent legitimate contributions to diversity in the Virginia higher education system." Ibid. As a theoretical matter, Virginia's educational interest would have been best served (insofar as the two factors we have mentioned are concerned) by six different types of public colleges—an all-men's, an all-women's, and a coeducational college run in the "adversative method," and an all-men's, an all-women's, and a coeducational college run in the "traditional method." But as a practical matter, of course, Virginia's financial resources, like any State's, are not limitless, and the Commonwealth must select among the available options. Virginia thus has decided to fund, in addition to some 14 coeducational 4-year colleges, one college that is run as an all-male school on the adversative model: the Virginia Military Institute.

Virginia did not make this determination regarding the make-up of its public college system on the unrealistic assumption that no other colleges exist. Substantial evidence in the District Court demonstrated that the Commonwealth has long proceeded on the principle that "`[h]igher education resources should be viewed as a whole—public and private' [579] "—because such an approach enhances diversity and because "`it is academic and economic waste to permit unwarranted duplication.' " Id., at 1420-1421 (quoting 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia). It is thus significant that, whereas there are "four all-female private [colleges] in Virginia," there is only "one private all-male college," which "indicates that the private sector is providing for th[e] [former] form of education to a much greater extent that it provides for all-male education." 766 F. Supp., at 1420-1421. In these circumstances, Virginia's election to fund one public all-male institution and one on the adversative model—and to concentrate its resources in a single entity that serves both these interests in diversity—is substantially related to the Commonwealth's important educational interests.

B

The Court today has no adequate response to this clear demonstration of the conclusion produced by application of intermediate scrutiny. Rather, it relies on a series of contentions that are irrelevant or erroneous as a matter of law, foreclosed by the record in this litigation, or both.

1. I have already pointed out the Court's most fundamental error, which is its reasoning that VMI's all-male composition is unconstitutional because "some women are capable of all of the individual activities required of VMI cadets," 766 F. Supp., at 1412, and would prefer military training on the adversative model. See supra, at 571-574. This unacknowledged adoption of what amounts to (at least) strict scrutiny is without antecedent in our sex-discrimination cases and by itself discredits the Court's decision.

2. The Court suggests that Virginia's claimed purpose in maintaining VMI as an all-male institution—its asserted interest in promoting diversity of educational options—is not "genuin[e]," but is a pretext for discriminating against women. Ante, at 539; see ante, at 535-540. To support this [580] charge, the Court would have to impute that base motive to VMI's Mission Study Committee, which conducted a 3-year study from 1983 to 1986 and recommended to VMI's Board of Visitors that the school remain all male. The committee, a majority of whose members consisted of non-VMI graduates, "read materials on education and on women in the military," "made site visits to single-sex and newly coeducational institutions" including West Point and the Naval Academy, and "considered the reasons that other institutions had changed from single-sex to coeducational status"; its work was praised as "thorough" in the accreditation review of VMI conducted by the Southern Association of Colleges and Schools. See 766 F. Supp., at 1413, 1428; see also id., at 1427-1430 (detailed findings of fact concerning the Mission Study Committee). The Court states that "[w]hatever internal purpose the Mission Study Committee served— and however well meaning the framers of the report—we can hardly extract from that effort any commonwealth policy evenhandedly to advance diverse educational options." Ante, at 539. But whether it is part of the evidence to prove that diversity was the Commonwealth's objective (its short report said nothing on that particular subject) is quite separate from whether it is part of the evidence to prove that antifeminism was not. The relevance of the Mission Study Committee is that its very creation, its sober 3-year study, and the analysis it produced utterly refute the claim that VMI has elected to maintain its all-male student-body composition for some misogynistic reason.

The Court also supports its analysis of Virginia's "actual state purposes" in maintaining VMI's student body as all male by stating that there is no explicit statement in the record "`in which the Commonwealth has expressed itself' " concerning those purposes. Ante, at 535, 539 (quoting 976 F. 2d 890, 899 (CA4 1992)); see also ante, at 525. That is wrong on numerous grounds. First and foremost, in its implication that such an explicit statement of "actual purposes" [581] is needed. The Court adopts, in effect, the argument of the United States that since the exclusion of women from VMI in 1839 was based on the "assumptions" of the time "that men alone were fit for military and leadership roles," and since "[b]efore this litigation was initiated, Virginia never sought to supply a valid, contemporary rationale for VMI's exclusionary policy," "[t]hat failure itself renders the VMI policy invalid." Brief for United States in No. 94-2107, at 10. This is an unheard-of doctrine. Each state decision to adopt or maintain a governmental policy need not be accompanied—in anticipation of litigation and on pain of being found to lack a relevant state interest—by a lawyer's contemporaneous recitation of the State's purposes. The Constitution is not some giant Administrative Procedure Act, which imposes upon the States the obligation to set forth a "statement of basis and purpose" for their sovereign Acts, see 5 U. S. C. § 553(c). The situation would be different if what the Court assumes to have been the 1839 policy had been enshrined and remained enshrined in legislation—a VMI charter, perhaps, pronouncing that the institution's purpose is to keep women in their place. But since the 1839 policy was no more explicitly recorded than the Court contends the present one is, the mere fact that today's Commonwealth continues to fund VMI "is enough to answer [the United States'] contention that the [classification] was the `accidental by-product of a traditional way of thinking about females.' " Michael M., 450 U. S., at 471, n. 6 (plurality opinion) (quoting Califano v. Webster, 430 U. S., at 320) (internal quotation marks omitted).

It is, moreover, not true that Virginia's contemporary reasons for maintaining VMI are not explicitly recorded. It is hard to imagine a more authoritative source on this subject than the 1990 Report of the Virginia Commission on the University of the 21st Century (1990 Report). As the parties stipulated, that report "notes that the hallmarks of Virginia's educational policy are `diversity and autonomy.' " Stipulations [582] of Fact 37, reprinted in Lodged Materials from the Record 64 (Lodged Materials). It said: "The formal system of higher education in Virginia includes a great array of institutions: state-supported and independent, two-year and senior, research and highly specialized, traditionally black and single-sex. " 1990 Report, quoted in relevant part at Lodged Materials 64-65 (emphasis added).<sup>[26]</sup> The Court's only response to this is repeated reliance on the Court of Appeals' assertion that "`the only explicit [statement] that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions' " (namely, the statement in the 1990 Report that the Commonwealth's institutions must "deal with faculty, staff, and students without regard to sex") had nothing to do with the purpose of diversity. Ante, at 525, 539 (quoting 976 F. 2d, at 899). This proves, I suppose, that the Court of Appeals did not find a statement dealing with sex and diversity in the record; but the pertinent question (accepting the need for such a statement) is whether it was there. And the plain fact, which the Court does not deny, is that it was.

[583] The Court contends that "[a] purpose genuinely to advance an array of educational options . . . is not served" by VMI. Ante, at 539-540. It relies on the fact that all of Virginia's other public colleges have become coeducational. Ibid.; see also ante, at 521, n. 2. The apparent theory of this argument is that unless Virginia pursues a great deal of diversity, its pursuit of some diversity must be a sham. This fails to take account of the fact that Virginia's resources cannot support all possible permutations of schools, see supra, at 578, and of the fact that Virginia coordinates its public educational offerings with the offerings of in-state private educational institutions that the Commonwealth provides money for its residents to attend and otherwise assists—which include four women's colleges.^[27]

Finally, the Court unreasonably suggests that there is some pretext in Virginia's reliance upon decentralized decisionmaking [584] to achieve diversity—its granting of substantial autonomy to each institution with regard to student-body composition and other matters, see 766 F. Supp., at 1419. The Court adopts the suggestion of the Court of Appeals that it is not possible for "one institution with autonomy, but with no authority over any other state institution, [to] give effect to a state policy of diversity among institutions." Ante, at 539 (internal quotation marks omitted). If it were impossible for individual human beings (or groups of human beings) to act autonomously in effective pursuit of a common goal, the game of soccer would not exist. And where the goal is diversity in a free market for services, that tends to be achieved even by autonomous actors who act out of entirely selfish interests and make no effort to cooperate. Each Virginia institution, that is to say, has a natural incentive to make itself distinctive in order to attract a particular segment of student applicants. And of course none of the institutions is entirely autonomous; if and when the legislature decides that a particular school is not well serving the interest of diversity—if it decides, for example, that a men's school is not much needed—funding will cease.^[28]

[585] 3. In addition to disparaging Virginia's claim that VMI's single-sex status serves a state interest in diversity, the Court finds fault with Virginia's failure to offer education based on the adversative training method to women. It dismisses the District Court's "`findings' on `gender-based developmental differences' " on the ground that "[t]hese `findings' restate the opinions of Virginia's expert witnesses, opinions about typically male or typically female `tendencies.' " Ante, at 541 (quoting 766 F. Supp., at 1434-1435). How remarkable to criticize the District Court on the ground that its findings rest on the evidence (i. e., the testimony of Virginia's witnesses)! That is what findings are supposed to do. It is indefensible to tell the Commonwealth that "[t]he burden of justification is demanding and it rests entirely on [you]," ante, at 533, and then to ignore the District Court's findings because they rest on the evidence put forward by the Commonwealth—particularly when, as the District Court said, "[t]he evidence in the case . . . is virtually uncontradicted, " 766 F. Supp., at 1415 (emphasis added).

Ultimately, in fact, the Court does not deny the evidence supporting these findings. See ante, at 541-546. It instead makes evident that the parties to this litigation could have saved themselves a great deal of time, trouble, and expense by omitting a trial. The Court simply dispenses with the evidence submitted at trial—it never says that a single finding of the District Court is clearly erroneous—in favor of the Justices' own view of the world, which the Court proceeds to support with (1) references to observations of someone [586] who is not a witness, nor even an educational expert, nor even a judge who reviewed the record or participated in the judgment below, but rather a judge who merely dissented from the Court of Appeals' decision not to rehear this litigation en banc, see ante, at 542, (2) citations of nonevidentiary materials such as amicus curiae briefs filed in this Court, see ante, at 544-545, nn. 13, 14, and (3) various historical anecdotes designed to demonstrate that Virginia's support for VMI as currently constituted reminds the Justices of the "bad old days," see ante, at 542-544.

It is not too much to say that this approach to the litigation has rendered the trial a sham. But treating the evidence as irrelevant is absolutely necessary for the Court to reach its conclusion. Not a single witness contested, for example, Virginia's "substantial body of `exceedingly persuasive' evidence . . . that some students, both male and female, benefit from attending a single-sex college" and "[that] [f]or those students, the opportunity to attend a single-sex college is a valuable one, likely to lead to better academic and professional achievement." 766 F. Supp., at 1411-1412. Even the United States' expert witness "called himself a `believer in single-sex education,' " although it was his "personal, philosophical preference," not one "born of educational-benefit considerations," "that single-sex education should be provided only by the private sector." Id., at 1412.

4. The Court contends that Virginia, and the District Court, erred, and "misperceived our precedent," by "train[ing] their argument on `means' rather than `end,' " ante, at 545. The Court focuses on "VMI's mission," which is to produce individuals "imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready . . . to defend their country in time of national peril." 766 F. Supp., at 1425 (quoting Mission Study Committee of the VMI Board of [587] Visitors, Report, May 16, 1986). "Surely," the Court says, "that goal is great enough to accommodate women." Ante, at 545.

This is lawmaking by indirection. What the Court describes as "VMI's mission" is no less the mission of all Virginia colleges. Which of them would the Old Dominion continue to fund if they did not aim to create individuals "imbued with love of learning, etc.," right down to being ready "to defend their country in time of national peril"? It can be summed up as "learning, leadership, and patriotism." To be sure, those general educational values are described in a particularly martial fashion in VMI's mission statement, in accordance with the military, adversative, and all-male character of the institution. But imparting those values in that fashion —i. e., in a military, adversative, all-male environment—is the distinctive mission of VMI. And as I have discussed (and both courts below found), that mission is not "great enough to accommodate women."

The Court's analysis at least has the benefit of producing foreseeable results. Applied generally, it means that whenever a State's ultimate objective is "great enough to accommodate women" (as it always will be), then the State will be held to have violated the Equal Protection Clause if it restricts to men even one means by which it pursues that objective—no matter how few women are interested in pursuing the objective by that means, no matter how much the single-sex program will have to be changed if both sexes are admitted, and no matter how beneficial that program has theretofore been to its participants.

5. The Court argues that VMI would not have to change very much if it were to admit women. See, e. g., ante, at 540-542. The principal response to that argument is that it is irrelevant: If VMI's single-sex status is substantially related to the government's important educational objectives, as I have demonstrated above and as the Court refuses to discuss, [588] that concludes the inquiry. There should be no debate in the federal judiciary over "how much" VMI would be required to change if it admitted women and whether that would constitute "too much" change.

But if such a debate were relevant, the Court would certainly be on the losing side. The District Court found as follows: "[T]he evidence establishes that key elements of the adversative VMI educational system, with its focus on barracks life, would be fundamentally altered, and the distinctive ends of the system would be thwarted, if VMI were forced to admit females and to make changes necessary to accommodate their needs and interests." 766 F. Supp., at 1411. Changes that the District Court's detailed analysis found would be required include new allowances for personal privacy in the barracks, such as locked doors and coverings on windows, which would detract from VMI's approach of regulating minute details of student behavior, "contradict the principle that everyone is constantly subject to scrutiny by everyone else," and impair VMI's "total egalitarian approach" under which every student must be "treated alike"; changes in the physical training program, which would reduce "[t]he intensity and aggressiveness of the current program"; and various modifications in other respects of the adversative training program that permeates student life. See id., at 1412-1413, 1435-1443. As the Court of Appeals summarized it, "the record supports the district court's findings that at least these three aspects of VMI's program— physical training, the absence of privacy, and the adversative approach—would be materially affected by coeducation, leading to a substantial change in the egalitarian ethos that is a critical aspect of VMI's training." 976 F. 2d, at 896-897.

In the face of these findings by two courts below, amply supported by the evidence, and resulting in the conclusion that VMI would be fundamentally altered if it admitted women, this Court simply pronounces that "[t]he notion that [589] admission of women would downgrade VMI's stature, destroy the adversative system and, with it, even the school, is a judgment hardly proved." Ante, at 542 (footnote omitted). The point about "downgrad[ing] VMI's stature" is a straw man; no one has made any such claim. The point about "destroy[ing] the adversative system" is simply false; the District Court not only stated that "[e]vidence supports this theory," but specifically concluded that while "[w]ithout a doubt" VMI could assimilate women, "it is equally without a doubt that VMI's present methods of training and education would have to be changed" by a "move away from its adversative new cadet system." 766 F. Supp., at 1413, and n. 8, 1440. And the point about "destroy[ing] the school," depending upon what that ambiguous phrase is intended to mean, is either false or else sets a standard much higher than VMI had to meet. It sufficed to establish, as the District Court stated, that VMI would be "significantly different" upon the admission of women, 766 F. Supp., at 1412, and "would eventually find it necessary to drop the adversative system altogether," id., at 1413.^[29]

[590] 6. Finally, the absence of a precise "all-women's analogue" to VMI is irrelevant. In Mississippi Univ. for Women v. Hogan, 458 U. S. 718 (1982), we attached no constitutional significance to the absence of an all-male nursing school. As Virginia notes, if a program restricted to one sex is necessarily unconstitutional unless there is a parallel program restricted to the other sex, "the opinion in Hogan could have ended with its first footnote, which observed that `Mississippi maintains no other single-sex public university or college.' " Brief for Cross-Petitioners in No. 94-2107, p. 38 (quoting Mississippi Univ. for Women v. Hogan, supra, at 720, n. 1).

Although there is no precise female-only analogue to VMI, Virginia has created during this litigation the Virginia Women's Institute for Leadership (VWIL), a state-funded all-women's program run by Mary Baldwin College. I have thus far said nothing about VWIL because it is, under our established test, irrelevant, so long as VMI `s all-male character is "substantially related" to an important state goal. But VWIL now exists, and the Court's treatment of it shows how far reaching today's decision is.

VWIL was carefully designed by professional educators who have long experience in educating young women. The program rejects the proposition that there is a "difference in the respective spheres and destinies of man and woman," Bradwell v. State, 16 Wall. 130, 141 (1873), and is designed to "provide an all-female program that will achieve substantially similar outcomes [to VMI's] in an all-female environment," 852 F. Supp. 471, 481 (WD Va. 1994). After holding a trial where voluminous evidence was submitted and making detailed findings of fact, the District Court concluded that "there is a legitimate pedagogical basis for the different means employed [by VMI and VWIL] to achieve the substantially [591] similar ends." Ibid. The Court of Appeals undertook a detailed review of the record and affirmed. 44 F. 3d 1229 (CA4 1995).^[30] But it is Mary Baldwin College, which runs VWIL, that has made the point most succinctly:

"It would have been possible to develop the VWIL program to more closely resemble VMI, with adversative techniques associated with the rat line and barracks-like living quarters. Simply replicating an existing program would have required far less thought, research, and educational expertise. But such a facile approach would have produced a paper program with no real prospect of successful implementation." Brief for Mary Baldwin College as Amicus Curiae 5.

It is worth noting that none of the United States' own experts in the remedial phase of this litigation was willing to testify that VMI's adversative method was an appropriate methodology for educating women. This Court, however, does not care. Even though VWIL was carefully designed by professional educators who have tremendous experience in the area, and survived the test of adversarial litigation, the Court simply declares, with no basis in the evidence, that [592] these professionals acted on "`overbroad' generalizations," ante, at 542, 550.

C

A few words are appropriate in response to the concurrence, which finds VMI unconstitutional on a basis that is more moderate than the Court's but only at the expense of being even more implausible. The concurrence offers three reasons: First, that there is "scant evidence in the record," ante, at 562, that diversity of educational offering was the real reason for Virginia's maintaining VMI. "Scant" has the advantage of being an imprecise term. I have cited the clearest statements of diversity as a goal for higher education in the 1990 Report, the 1989 Virginia Plan for Higher Education, the Budget Initiatives prepared in 1989 by the State Council of Higher Education for Virginia, the 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia, and the 1969 Report of the Virginia Commission on Constitutional Revision. See supra, at 579, 581-582, and n. 2, 583, n. 3. There is no evidence to the contrary, once one rejects (as the concurrence rightly does) the relevance of VMI's founding in days when attitudes toward the education of women were different. Is this conceivably not enough to foreclose rejecting as clearly erroneous the District Court's determination regarding "the Commonwealth's objective of educational diversity"? 766 F. Supp., at 1413. Especially since it is absurd on its face even to demand "evidence" to prove that the Commonwealth's reason for maintaining a men's military academy is that a men's military academy provides a distinctive type of educational experience (i. e., fosters diversity). What other purpose would the Commonwealth have? One may argue, as the Court does, that this type of diversity is designed only to indulge hostility toward women—but that is a separate point, explicitly rejected by the concurrence, and amply refuted by the evidence I have mentioned in discussing [593] the Court's opinion.^[31] What is now under discussion—the concurrence's making central to the disposition of this litigation the supposedly "scant" evidence that Virginia maintained VMI in order to offer a diverse educational experience—is rather like making crucial to the lawfulness of the United States Army record "evidence" that its purpose is to do battle. A legal culture that has forgotten the concept of res ipsa loquitur deserves the fate that it today decrees for VMI.

Second, the concurrence dismisses out of hand what it calls Virginia's "second justification for the single-sex admissions policy: maintenance of the adversative method." Ante, at 564. The concurrence reasons that "this justification does not serve an important governmental objective" because, whatever the record may show about the pedagogical benefits of single-sex education, "there is no similar evidence in the record that an adversative method is pedagogically beneficial or is any more likely to produce character traits than other methodologies." Ibid. That is simply wrong. See, e. g., 766 F. Supp., at 1426 (factual findings concerning character traits produced by VMI's adversative methodology); id. , at 1434 (factual findings concerning benefits for many college-age men of an adversative approach in general). In reality, the pedagogical benefits of VMI's adversative approach were not only proved, but were a given in this litigation. The reason the woman applicant who prompted this suit wanted to enter VMI was assuredly not that she wanted to go to an all-male school; it would cease being all-male as [594] soon as she entered. She wanted the distinctive adversative education that VMI provided, and the battle was joined (in the main) over whether VMI had a basis for excluding women from that approach. The Court's opinion recognizes this, and devotes much of its opinion to demonstrating that "`some women . . . do well under [the] adversative model' " and that "[i]t is on behalf of these women that the United States has instituted this suit." Ante, at 550 (quoting 766 F. Supp., at 1434). Of course, in the last analysis it does not matter whether there are any benefits to the adversative method. The concurrence does not contest that there are benefits to single-sex education, and that alone suffices to make Virginia's case, since admission of a woman will even more surely put an end to VMI's single-sex education than it will to VMI's adversative methodology.

A third reason the concurrence offers in support of the judgment is that the Commonwealth and VMI were not quick enough to react to the "further developments" in this Court's evolving jurisprudence. Ante, at 561. Specifically, the concurrence believes it should have been clear after Hogan that "[t]he difficulty with [Virginia's] position is that the diversity benefited only one sex; there was single-sex public education available for men at VMI, but no corresponding single-sex public education available for women." Ante, at 562. If only, the concurrence asserts, Virginia had "made a genuine effort to devote comparable public resources to a facility for women, and followed through on such a plan, it might well have avoided an equal protection violation." Ante, at 563. That is to say, the concurrence believes that after our decision in Hogan (which held a program of the Mississippi University for Women to be unconstitutional—without any reliance on the fact that there was no corresponding Mississippi all-men's program), the Commonwealth should have known that what this Court expected of it was . . . yes!, the creation of a state all-women's program. Any lawyer who gave that advice to the Commonwealth [595] ought to have been either disbarred or committed. (The proof of that pudding is today's 6-Justice majority opinion.) And any Virginia politician who proposed such a step when there were already four 4-year women's colleges in Virginia (assisted by state support that may well exceed, in the aggregate, what VMI costs, see n. 3, supra ) ought to have been recalled.

In any event, "diversity in the form of single-sex, as well as coeducational, institutions of higher learning" is "available to women as well as to men" in Virginia. Ante, at 564. The concurrence is able to assert the contrary only by disregarding the four all-women's private colleges in Virginia (generously assisted by public funds) and the Commonwealth's longstanding policy of coordinating public with private educational offerings, see supra, at 579, 581-582, and n. 2, 583-584, and n. 3. According to the concurrence, the

2, Virginia's reason assistance to its four all-women's private colleges does not count is that "[t]he private women's colleges are treated by the State exactly as all other private schools are treated." Ante, at 564. But if Virginia cannot get credit for assisting women's education if it only treats women's private schools as it does all other private schools, then why should it get blame for assisting men's education if it only treats VMI as it does all other public schools? This is a great puzzlement.

IV

As is frequently true, the Court's decision today will have consequences that extend far beyond the parties to the litigation. What I take to be the Court's unease with these consequences, and its resulting unwillingness to acknowledge them, cannot alter the reality.

A

Under the constitutional principles announced and applied today, single-sex public education is unconstitutional. By going through the motions of applying a balancing test—asking [596] whether the State has adduced an "exceedingly persuasive justification" for its sex-based classification—the Court creates the illusion that government officials in some future case will have a clear shot at justifying some sort of singlesex public education. Indeed, the Court seeks to create even a greater illusion than that: It purports to have said nothing of relevance to other public schools at all. "We address specifically and only an educational opportunity recognized . . . as `unique.' " Ante, at 534, n. 7.

The Supreme Court of the United States does not sit to announce "unique" dispositions. Its principal function is to establish precedent —that is, to set forth principles of law that every court in America must follow. As we said only this Term, we expect both ourselves and lower courts to adhere to the "rationale upon which the Court based the results of its earlier decisions." Seminole Tribe of Fla. v. Florida, 517 U. S. 44, 66-67 (1996) (emphasis added). That is the principal reason we publish our opinions.

And the rationale of today's decision is sweeping: for sexbased classifications, a redefinition of intermediate scrutiny that makes it indistinguishable from strict scrutiny. See supra, at 571-574. Indeed, the Court indicates that if any program restricted to one sex is "uniqu[e]," it must be opened to members of the opposite sex "who have the will and capacity" to participate in it. Ante, at 542. I suggest that the single-sex program that will not be capable of being characterized as "unique" is not only unique but nonexistent.^[32]

In any event, regardless of whether the Court's rationale leaves some small amount of room for lawyers to argue, it ensures that single-sex public education is functionally dead. [597] The costs of litigating the constitutionality of a single-sex education program, and the risks of ultimately losing that litigation, are simply too high to be embraced by public officials. Any person with standing to challenge any sex-based classification can haul the State into federal court and compel it to establish by evidence (presumably in the form of expert testimony) that there is an "exceedingly persuasive justification" for the classification. Should the courts happen to interpret that vacuous phrase as establishing a standard that is not utterly impossible of achievement, there is considerable risk that whether the standard has been met will not be determined on the basis of the record evidence—indeed, that will necessarily be the approach of any court that seeks to walk the path the Court has trod today. No state official in his right mind will buy such a high-cost, high-risk lawsuit by commencing a single-sex program. The enemies of singlesex education have won; by persuading only seven Justices (five would have been enough) that their view of the world is enshrined in the Constitution, they have effectively imposed that view on all 50 States.

This is especially regrettable because, as the District Court here determined, educational experts in recent years have increasingly come to "suppor[t] [the] view that substantial educational benefits flow from a single-gender environment, be it male or female, that cannot be replicated in a coeducational setting. " 766 F. Supp., at 1415 (emphasis added). "The evidence in th[is] case," for example, "is virtually uncontradicted" to that effect. Ibid. Until quite recently, some public officials have attempted to institute new single-sex programs, at least as experiments. In 1991, for example, the Detroit Board of Education announced a program to establish three boys-only schools for inner-city youth; it was met with a lawsuit, a preliminary injunction was swiftly entered by a District Court that purported to rely on Hogan, see Garrett v. Board of Ed. of School Dist. of Detroit, 775 F. Supp. 1004, 1006 (ED Mich. 1991), and the [598] Detroit Board of Education voted to abandon the litigation and thus abandon the plan, see Detroit Plan to Aid Blacks with All-Boy Schools Abandoned, Los Angeles Times, Nov. 8, 1991, p. A4, col. 1. Today's opinion assures that no such experiment will be tried again.

B

There are few extant single-sex public educational programs. The potential of today's decision for widespread disruption of existing institutions lies in its application to private single-sex education. Government support is immensely important to private educational institutions. Mary Baldwin College—which designed and runs VWIL— notes that private institutions of higher education in the 1990-1991 school year derived approximately 19 percent of their budgets from federal, state, and local government funds, not including financial aid to students. See Brief for Mary Baldwin College as Amicus Curiae 22, n. 13 (citing U. S. Dept. of Education, National Center for Education Statistics, Digest of Education Statistics, p. 38 and Note (1993)). Charitable status under the tax laws is also highly significant for private educational institutions, and it is certainly not beyond the Court that rendered today's decision to hold that a donation to a single-sex college should be deemed contrary to public policy and therefore not deductible if the college discriminates on the basis of sex. See Note, The Independent Sector and the Tax Laws: Defining Charity in an Ideal Democracy, 64 S. Cal. L. Rev. 461, 476 (1991). See also Bob Jones Univ. v. United States, 461 U. S. 574 (1983).

The Court adverts to private single-sex education only briefly, and only to make the assertion (mentioned above) that "[w]e address specifically and only an educational opportunity recognized by the District Court and the Court of Appeals as `unique.' " Ante, at 534, n. 7. As I have already remarked, see supra, at 596, that assurance assures nothing, unless it is to be taken as a promise that in the future [599] the Court will disclaim the reasoning it has used today to destroy VMI. The Government, in its briefs to this Court, at least purports to address the consequences of its attack on VMI for public support of private single-sex education. It contends that private colleges that are the direct or indirect beneficiaries of government funding are not thereby necessarily converted into state actors to which the Equal Protection Clause is then applicable. See Brief for United States in No. 94-2107, at 35-37 (discussing Rendell-Baker v. Kohn, 457 U. S. 830 (1982), and Blum v. Yaretsky, 457 U. S. 991 (1982)). That is true. It is also virtually meaningless.

The issue will be not whether government assistance turns private colleges into state actors, but whether the government itself would be violating the Constitution by providing state support to single-sex colleges. For example, in Norwood v. Harrison, 413 U. S. 455 (1973), we saw no room to distinguish between state operation of racially segregated schools and state support of privately run segregated schools. "Racial discrimination in state-operated schools is barred by the Constitution and `[i]t is also axiomatic that a state may not induce, encourage or promote private persons to accomplish what it is constitutionally forbidden to accomplish.' " Id., at 465 (quoting Lee v. Macon County Bd. of Ed., 267 F. Supp. 458, 475-476 (MD Ala. 1967)); see also Cooper v. Aaron, 358 U. S. 1, 19 (1958) ("State support of segregated schools through any arrangement, management, funds, or property cannot be squared with the [Fourteenth] Amendment's command that no State shall deny to any person within its jurisdiction the equal protection of the laws"); Grove City College v. Bell, 465 U. S. 555, 565 (1984) (case arising under Title IX of the Education Amendments of 1972 and stating that "[t]he economic effect of direct and indirect assistance often is indistinguishable"). When the Government was pressed at oral argument concerning the implications of these cases for private single-sex education if government-provided single-sex education is unconstitutional, [600] it stated that the implications will not be so disastrous, since States can provide funding to racially segregated private schools, "depend[ing] on the circumstances," Tr. of Oral Arg. 56. I cannot imagine what those "circumstances" might be, and it would be as foolish for privateschool administrators to think that that assurance from the Justice Department will outlive the day it was made, as it was for VMI to think that the Justice Department's "unequivoca[l]" support for an intermediate-scrutiny standard in this litigation would survive the Government's loss in the courts below.

The only hope for state-assisted single-sex private schools is that the Court will not apply in the future the principles of law it has applied today. That is a substantial hope, I am happy and ashamed to say. After all, did not the Court today abandon the principles of law it has applied in our earlier sex-classification cases? And does not the Court positively invite private colleges to rely upon our ad-hocery by assuring them this litigation is "unique"? I would not advise the foundation of any new single-sex college (especially an all-male one) with the expectation of being allowed to receive any government support; but it is too soon to abandon in despair those single-sex colleges already in existence. It will certainly be possible for this Court to write a future opinion that ignores the broad principles of law set forth today, and that characterizes as utterly dispositive the opinion's perceptions that VMI was a uniquely prestigious allmale institution, conceived in chauvinism, etc., etc. I will not join that opinion.

* * *

Justice Brandeis said it is "one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country." New State Ice Co. v. Liebmann, 285 U. S. 262, 311 [601] (1932) (dissenting opinion). But it is one of the unhappy incidents of the federal system that a self-righteous Supreme Court, acting on its Members' personal view of what would make a "`more perfect Union,' " ante, at 558 (a criterion only slightly more restrictive than a "more perfect world"), can impose its own favored social and economic dispositions nationwide. As today's disposition, and others this single Term, show, this places it beyond the power of a "single courageous State," not only to introduce novel dispositions that the Court frowns upon, but to reintroduce, or indeed even adhere to, disfavored dispositions that are centuries old. See, e. g., BMW of North America, Inc. v. Gore, 517 U. S. 559 (1996); Romer v. Evans, 517 U. S. 620 (1996). The sphere of self-government reserved to the people of the Republic is progressively narrowed.

In the course of this dissent, I have referred approvingly to the opinion of my former colleague, Justice Powell, in Mississippi Univ. for Women v. Hogan, 458 U. S. 718 (1982). Many of the points made in his dissent apply with equal force here—in particular, the criticism of judicial opinions that purport to be "narro[w]" but whose "logic" is "sweepin[g]." Id., at 745-746, n. 18. But there is one statement with which I cannot agree. Justice Powell observed that the Court's decision in Hogan, which struck down a single-sex program offered by the Mississippi University for Women, had thereby "[l]eft without honor . . . an element of diversity that has characterized much of American education and enriched much of American life." Id., at 735. Today's decision does not leave VMI without honor; no court opinion can do that.

In an odd sort of way, it is precisely VMI's attachment to such old-fashioned concepts as manly "honor" that has made it, and the system it represents, the target of those who today succeed in abolishing public single-sex education. The record contains a booklet that all first-year VMI students [602] (the so-called "rats") were required to keep in their possession at all times. Near the end there appears the following period piece, entitled "The Code of a Gentleman":

"Without a strict observance of the fundamental Code of Honor, no man, no matter how `polished,' can be considered a gentleman. The honor of a gentleman demands the inviolability of his word, and the incorruptibility of his principles. He is the descendant of the knight, the crusader; he is the defender of the defenseless and the champion of justice . . . or he is not a Gentleman.

"A Gentleman . . .

"Does not discuss his family affairs in public or with acquaintances.

"Does not speak more than casually about his girl friend.

"Does not go to a lady's house if he is affected by alcohol. He is temperate in the use of alcohol.

"Does not lose his temper; nor exhibit anger, fear, hate, embarrassment, ardor or hilarity in public.

"Does not hail a lady from a club window.

"A gentleman never discusses the merits or demerits of a lady.

"Does not mention names exactly as he avoids the mention of what things cost.

"Does not borrow money from a friend, except in dire need. Money borrowed is a debt of honor, and must be repaid as promptly as possible. Debts incurred by a deceased parent, brother, sister or grown child are assumed by honorable men as a debt of honor.

"Does not display his wealth, money or possessions.

"Does not put his manners on and off, whether in the club or in a ballroom. He treats people with courtesy, no matter what their social position may be.

[603] "Does not slap strangers on the back nor so much as lay a finger on a lady.

"Does not `lick the boots of those above' nor `kick the face of those below him on the social ladder.'

"Does not take advantage of another's helplessness or ignorance and assumes that no gentleman will take advantage of him.

"A Gentleman respects the reserves of others, but demands that others respect those which are his.

"A Gentleman can become what he wills to be. . . ."

I do not know whether the men of VMI lived by this code; perhaps not. But it is powerfully impressive that a public institution of higher education still in existence sought to have them do so. I do not think any of us, women included, will be better off for its destruction.

[1] Together with No. 94-2107, Virginia et al. v. United States, also on certiorari to the same court.

[2] Briefs of amici curiae urging reversal in No. 94-1941 were filed for the State of Maryland et al. by J. Joseph Curran, Jr., Attorney General of Maryland, and Andrew H. Baida, Assistant Attorney General, and by the Attorneys General for their respective jurisdictions as follows: Margery S. Bronster of Hawaii, Scott Harshbarger of Massachusetts, Frankie Sue Del Papa of Nevada, C. Sebastian Aloot of the Northern Mariana Islands, and Theodore R. Kulongoski of Oregon; for the Employment Law Center et al. by Patricia A. Shiu and Judith Kurtz; and for the National Women's Law Center et al. by Robert N. Weiner, Marcia D. Greenberger, Sara L. Mandelbaum, Janet Gallagher, Mary Wyckoff, Steven R. Shapiro, and Susan Deller Ross.

Briefs of amici curiae urging affirmance in No. 94-1941 were filed for the State of South Carolina et al. by Charles Molony Condon, Attorney General, Treva Ashworth, Deputy Attorney General, Kenneth P. Woodington, Senior Assistant Attorney General, Reginald I. Lloyd, Assistant Attorney General, and M. Dawes Cooke, Jr.; and for Kenneth E. Clark et al. by James C. Roberts and George A. Somerville.

Briefs of amici curiae were filed in both cases for the State of Wyoming et al. by William U. Hill, Attorney General of Wyoming, Thomas W. Corbett, Jr., Attorney General of Pennsylvania, and Bradley B. Cavedo; for Bennett College et al. by Wendy S. White; for the Center for Military Readiness et al. by Mellissa Wells-Petry and Jordan W. Lorence; for the Employment Law Center et al. by Patricia A. Shiu and Judith Kurtz; for the Independent Women's Forum et al. by Anita K. Blair and C. Douglas Welty; for Mary Baldwin College by Craig T. Merritt and Richard K. Willard; for the South Carolina Institute of Leadership for Women by Julianne Farnsworth; for Wells College et al. by David M. Lascell; for Women's Schools Together, Inc., et al. by John C. Danforth and Thomas C. Walsh; and for Nancy Mellette by Valorie K. Vojdik, Henry Weisburg, Suzanne E. Coe, and Robert R. Black.

Briefs of amici curiae were filed in No. 94-1941 for the American Association of University Professors et al. by Joan E. Bertin and Ann H. Franke; and for Rhonda Cornum et al. by Allan L. Gropper.

Daniel F. Kolb, Herbert J. Hansell, Paul C. Saunders, Norman Redlich, Barbara R. Arnwine, Thomas J. Henderson, and Richard T. Seymour filed a brief for the Lawyers' Committee for Civil Rights Under Law as amicus curiae in No. 94-2107.

[3] During the Civil War, school teaching became a field dominated by women. See A. Scott, The Southern Lady: From Pedestal to Politics, 1830-1930, p. 82 (1970).

[4] Historically, most of Virginia's public colleges and universities were single sex; by the mid-1970's, however, all except VMI had become coeducational. 766 F. Supp. 1407, 1418-1419 (WD Va. 1991). For example, Virginia's legislature incorporated Farmville Female Seminary Association in 1839, the year VMI opened. 1839 Va. Acts, ch. 167. Originally providing instruction in "English, Latin, Greek, French, and piano" in a "home atmosphere," R. Sprague, Longwood College: A History 7-8, 15 (1989) (Longwood College), Farmville Female Seminary became a public institution in 1884 with a mission to train "white female teachers for public schools," 1884 Va. Acts, ch. 311. The school became Longwood College in 1949, Longwood College 136, and introduced coeducation in 1976, id. , at 133.

[5] The District Court allowed the VMI Foundation and the VMI Alumni Association to intervene as defendants. 766 F. Supp., at 1408.

[6] Six judges voted to rehear the case en banc, four voted against rehearing, and three were recused. The Fourth Circuit's local Rule permits rehearing en banc only on the vote of a majority of the Circuit's judges in regular active service (currently 13) without regard to recusals. See 52 F. 3d, at 91, and n. 1.

[7] As Thomas Jefferson stated the view prevailing when the Constitution was new:

"Were our State a pure democracy . . . there would yet be excluded from their deliberations . . . [w]omen, who, to prevent depravation of morals and ambiguity of issue, could not mix promiscuously in the public meetings of men." Letter from Thomas Jefferson to Samuel Kercheval (Sept. 5, 1816), in 10 Writings of Thomas Jefferson 45-46, n. 1 (P. Ford ed. 1899).

[8] The Court has thus far reserved most stringent judicial scrutiny for classifications based on race or national origin, but last Term observed that strict scrutiny of such classifications is not inevitably "fatal in fact." Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 237 (1995) (internal quotation marks omitted).

[9] Several amici have urged that diversity in educational opportunities is an altogether appropriate governmental pursuit and that single-sex schools can contribute importantly to such diversity. Indeed, it is the mission of some single-sex schools "to dissipate, rather than perpetuate, traditional gender classifications." See Brief for Twenty-six Private Women's Colleges as Amici Curiae 5. We do not question the Commonwealth's prerogative evenhandedly to support diverse educational opportunities. We address specifically and only an educational opportunity recognized by the District Court and the Court of Appeals as "unique," see 766 F. Supp., at 1413, 1432; 976 F. 2d, at 892, an opportunity available only at Virginia's premier military institute, the Commonwealth's sole single-sex public university or college. Cf. Mississippi Univ. for Women v. Hogan, 458 U. S. 718, 720, n. 1 (1982) ("Mississippi maintains no other single-sex public university or college. Thus, we are not faced with the question of whether States can provide `separate but equal' undergraduate institutions for males and females.").

[10] On this point, the dissent sees fire where there is no flame. See post, at 596-598, 598-600. "Both men and women can benefit from a single-sex education," the District Court recognized, although "the beneficial effects" of such education, the court added, apparently "are stronger among women than among men." 766 F. Supp., at 1414. The United States does not challenge that recognition. Cf. C. Jencks & D. Riesman, The Academic Revolution 297-298 (1968):

"The pluralistic argument for preserving all-male colleges is uncomfortably similar to the pluralistic argument for preserving all-white colleges . . . . The all-male college would be relatively easy to defend if it emerged from a world in which women were established as fully equal to men. But it does not. It is therefore likely to be a witting or unwitting device for preserving tacit assumptions of male superiority—assumptions for which women must eventually pay."

[11] Dr. Edward H. Clarke of Harvard Medical School, whose influential book, Sex in Education, went through 17 editions, was perhaps the most well-known speaker from the medical community opposing higher education for women. He maintained that the physiological effects of hard study and academic competition with boys would interfere with the development of girls' reproductive organs. See E. Clarke, Sex in Education 38-39, 62-63 (1873); id. , at 127 ("identical education of the two sexes is a crime before God and humanity, that physiology protests against, and that experience weeps over"); see also H. Maudsley, Sex in Mind and in Education 17 (1874) ("It is not that girls have not ambition, nor that they fail generally to run the intellectual race [in coeducational settings], but it is asserted that they do it at a cost to their strength and health which entails life-long suffering, and even incapacitates them for the adequate performance of the natural functions of their sex."); C. Meigs, Females and Their Diseases 350 (1848) (after five or six weeks of "mental and educational discipline," a healthy woman would "lose . . . the habit of menstruation" and suffer numerous ills as a result of depriving her body for the sake of her mind).

[12] Virginia's Superintendent of Public Instruction dismissed the coeducational idea as "`repugnant to the prejudices of the people' " and proposed a female college similar in quality to Girton, Smith, or Vassar.2 History of Women's Education 254 (quoting Dept. of Interior, 1 Report of Commissioner of Education, H. R. Doc. No. 5, 58th Cong., 2d Sess., 438 (1904)).

[13] See post, at 566, 598-599, 603. Forecasts of the same kind were made regarding admission of women to the federal military academies. See, e. g., Hearings on H. R. 9832 et al. before Subcommittee No. 2 of the House Committee on Armed Services, 93d Cong., 2d Sess., 137 (1975) (statement of Lt. Gen. A. P. Clark, Superintendent of U. S. Air Force Academy) ("It is my considered judgment that the introduction of female cadets will inevitably erode this vital atmosphere."); id., at 165 (statement of Hon. H. H. Callaway, Secretary of the Army) ("Admitting women to West Point would irrevocably change the Academy. . . . The Spartan atmosphere— which is so important to producing the final product—would surely be diluted, and would in all probability disappear.").

[14] See 766 F. Supp., at 1413 (describing testimony of expert witness David Riesman: "[I]f VMI were to admit women, it would eventually find it necessary to drop the adversative system altogether, and adopt a system that provides more nurturing and support for the students."). Such judgments have attended, and impeded, women's progress toward full citizenship stature throughout our Nation's history. Speaking in 1879 in support of higher education for females, for example, Virginia State Senator C. T. Smith of Nelson recounted that legislation proposed to protect the property rights of women had encountered resistance. 10 Educ. J. Va. 213 (1879). A Senator opposing the measures objected that "there [was] no formal call for the [legislation]," and "depicted in burning eloquence the terrible consequences such laws would produce." Ibid. The legislation passed, and a year or so later, its sponsor, C. T. Smith, reported that "not one of [the forecast "terrible consequences"] has or ever will happen, even unto the sounding of Gabriel's trumpet." Ibid. See also supra, at 537-538.

[15] Women cadets have graduated at the top of their class at every federal military academy. See Brief for Lieutenant Colonel Rhonda Cornum et al. as Amici Curiae 11, n. 25; cf. Defense Advisory Committee on Women in the Services, Report on the Integration and Performance of Women at West Point 64 (1992).

[16] Brief for Lieutenant Colonel Rhonda Cornum, supra, at 5-9 (reporting the vital contributions and courageous performance of women in the military); see Mintz, President Nominates 1st Woman to Rank of Three-Star General, Washington Post, Mar. 27, 1996, p. A19, col. 1 (announcing President's nomination of Marine Corps Major General Carol Mutter to rank of Lieutenant General; Mutter will head corps manpower and planning); Tousignant, A New Era for the Old Guard, Washington Post, Mar. 23, 1996, p. C1, col. 2 (reporting admission of Sergeant Heather Johnsen to eliteInfantry unit that keeps round-the-clock vigil at Tomb of the Unknowns in Arlington National Cemetery).

[17] Inclusion of women in settings where, traditionally,they were not wanted inevitably entails a period of adjustment. As one West Point cadet squad leader recounted: "[T]he classes of `78 and `79 see the women as women, but the classes of `80and `81 see them as classmates." U. S. Military Academy, A. Vitters, Report of Admission of Women (Project Athena II) 84 (1978) (internal quotation marks omitted).

[18] VMI has successfully managed another notable change. The school admitted its first African-American cadets in 1968. See The VMI Story 347-349 (students no longer sing "Dixie," salute the Confederate flag or the tomb of General Robert E. Lee at ceremonies and sports events). As the District Court noted, VMI established a program on "retention of black cadets" designed to offer academic and social-cultural support to "minority members of a dominantly white and tradition-oriented student body." 766 F. Supp., at 1436-1437. The school maintains a "special recruitment program for blacks" which, the District Court found, "has had little,if any, effect on VMI's method of accomplishing its mission." Id., at 1437.

[19] As earlier observed, see supra, at 529, Judge Phillips, in dissent, measured Virginia's plan against a paradigm arrangement, one that "could survive equal protection scrutiny": single-sex schools with "substantially comparable curricular and extra-curricular programs, funding, physical plant, administration and support services, . . . faculty[,] and library resources." 44 F. 3d 1229, 1250 (CA4 1995). Cf. Bray v. Lee, 337 F. Supp. 934 (Mass. 1972) (holding inconsistent with the Equal Protection Clause admission of males to Boston's Boys Latin School with a test score of 120 or higher (up to a top score of 200) while requiring a score, on the same test, of at least 133 for admission of females to Girls Latin School, but not ordering coeducation). Measuring VMI/VWIL against the paradigm, Judge Phillips said, "reveals how far short the [Virginia] plan falls from providing substantially equal tangible and intangible educational benefits to men and women." 44 F. 3d, at 1250.

[20] Both programs include an honor system. Students at VMI are expelled forthwith for honor code violations, see 766 F. Supp., at 1423; the system for VWIL students, see 852 F. Supp., at 496-497, is less severe, see Tr. 414-415 (testimony of Mary Baldwin College President Cynthia Tyson).

[21] Admitting women to VMI would undoubtedly require alterations necessary to afford members of each sex privacy from the other sex in living arrangements, and to adjust aspects of the physical training programs. See Brief for Petitioner 27-29; cf.note following 10 U. S. C. § 4342 (academic and other standards for women admitted to the Military, Naval, and Air Force Academies "shall be the same as those required for male individuals, except for those minimum essential adjustments in such standards required because of physiological differences between male and female individuals"). Experience shows such adjustments are manageable. See U. S. Military Academy, A. Vitters, N. Kinzer, & J. Adams, Report of Admission of Women (Project Athena I-IV) (1977-1980) (4-year longitudinal study of the admission of women to West Point); Defense Advisory Committee on Women in the Services, Report on the Integration and Performance of Women at West Point 17-18 (1992).

[22] Virginia's prime concern, it appears, is that "plac[ing] men and women into the adversative relationship inherent in the VMI program . . . would destroy, at least for that period of the adversative training, any sense of decency that still permeates the relationship between the sexes." 44 F. 3d, at 1239; see supra, at 540-546. It is an ancient and familiar fear. Compare In re Lavinia Goodell, 39 Wis. 232, 246 (1875) (denying female applicant's motion for admission to the bar of its court, Wisconsin Supreme Court explained: "Discussions are habitually necessary in courts of justice, which are unfit for female ears. The habitual presence of women at these would tend to relax the public sense of decency and propriety."), with Levine, Closing Comments, 6 Law & Inequality 41 (1988) (presentation at Eighth Circuit Judicial Conference, Colorado Springs, Colo., July 17, 1987) (footnotes omitted):

"Plato questioned whether women should be afforded equal opportunity to become guardians, those elite Rulers of Platonic society. Ironically, in that most undemocratic system of government, the Republic, women's native ability to serve as guardians was not seriously questioned. The concern was over the wrestling and exercise class in which all candidates for guardianship had to participate, for rigorous physical and mental training were prerequisites to attain the exalted status of guardian. And in accord with Greek custom, those exercise classes were conducted in the nude. Plato concluded that their virtue would clothe the women's nakedness and that Platonic society would not thereby be deprived of the talent of qualified citizens for reasons of mere gender."

For Plato's full text on the equality of women, see 2 The Dialogues of Plato 302-312 (B. Jowett transl., 4th ed. 1953). Virginia, not bound to ancient Greek custom in its "rigorous physical and mental training" programs, could more readily make the accommodations necessary to draw on "the talent of [all] qualified citizens." Cf. supra, at 550-551, n. 19.

[23] R. Morris, The Forging of the Union, 1781-1789, p. 193 (1987); see id. , at 191, setting out letter to a friend from Massachusetts patriot (later second President) John Adams, on the subject of qualifications for voting in his home State:

"[I]t is dangerous to open so fruitful a source of controversy and altercation as would be opened by attempting to alter the qualifications of voters; there will be no end of it. New claims will arise; women will demand a vote; lads from twelve to twenty-one will think their rights not enough attended to; and every man who has not a farthing, will demand an equal voice with any other, in all acts of state. It tends to confound and destroy all distinctions, and prostrate all ranks to one common level." Letter from John Adams to James Sullivan (May 26, 1776), in 9 Works of John Adams 378 (C. Adams ed. 1854).

[24] The dissent equates our conclusion that VMI's "asserted interest in promoting diversity" is not "`genuine,' " with a "charge" that the diversity rationale is "a pretext for discriminating against women." Post, at 579-580. Of course, those are not the same thing. I do not read the Court as saying that the diversity rationale is a pretext for discrimination, and I would not endorse such a proposition. We may find that diversity was not the Commonwealth's real reason without suggesting, or having to show, that the real reason was "antifeminism," post, at 580. Our cases simply require that the proffered purpose for the challenged gender classification be the actual purpose, although not necessarily recorded. See ante, at 533, 535-536. The dissent also says that the interest in diversity is so transparent that having to articulate it is "absurd on its face."Post, at 592. Apparently, that rationale was not obvious to the Mission Study Committee which failed to list it among its reasons for maintaining VMI's all-men admissions policy.

[25] Accord, ante, at 541 ("In sum . . . , neither the goal of producing citizensoldiers, VMI's raison d'être, nor VMI's implementing methodology is inherently unsuitable to women" (internal quotation marks omitted; emphasis added)); ante, at 542 ("[T]he question is whether the Commonwealth can constitutionally deny to women who have the will and capacity, the training and attendant opportunities that VMI uniquely affords"); ante, at 547-548 (the "violation" is that "equal protection [has been] denied to women ready, willing, and able to benefit from educational opportunities of the kind VMI offers");ante, at 550 ("As earlier stated, see supra, at 541-542, generalizations about `the way women are,' estimates of what is appropriate for most women, no longer justify denying opportunity to women whose talent and capacity place them outside the average description").

[26] This statement is supported by other evidence in the record demonstrating, by reference to both public and private institutions, that Virginia actively seeks tofoster its "`richheritage of pluralism and diversity in higher education,' " 1969 Report of the Virginia Commission on Constitutional Revision, quoted in relevant part at Lodged Materials 53; that Virginia views "`[o]ne special characteristic of the Virginia system [as being] its diversity,' " 1989 Virginia Plan for Higher Education, quoted in relevant part at Lodged Materials 64; and that in the Commonwealth's view "[h]igher education resources should be viewed as a whole—public and private"—because "`Virginia needs the diversity inherent in a dual system of higher education,' " 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia, quoted in 766 F. Supp. 1407, 1420 (WD Va. 1991). See also Budget Initiatives for 1990-1992 of State Council of Higher Education for Virginia 10 (June 21, 1989) (Budget Initiatives),quoted at n. 3,infra. It should be noted (for this point will be crucial to my later discussion) that these official reports quoted here, in text and footnote, regard the Commonwealth's educational system—public and private —as a unitary one.

[27] The Commonwealth provides tuition assistance, scholarship grants, guaranteed loans, and work-study funds for residents of Virginia who attend private colleges in the Commonwealth. See, e. g., Va. Code Ann. §§ 23-38.11 to 23-38.19 (1993 and Supp. 1995) (Tuition Assistance Grant Act); §§ 23-38.30 to 23-38.44:3 (Virginia Student Assistance Authorities); Va. Code Ann. §§ 23-38.45 to 23-38.53 (1993) (College Scholarship Assistance Act); §§ 23-38.53:1 to 23-38.53:3 (Virginia Scholars Program); §§ 23-38.70, 23-38.71 (Virginia Work-Study Program). These programs involve substantial expenditures: for example, Virginia appropriated $4,413,750 (not counting federal funds it also earmarked) for the College Scholarship Assistance Program for both 1996 and 1997, and for the Tuition Assistance Grant Program appropriated $21,568,000 for 1996 and $25,842,000 for 1997. See 1996 Va. Appropriations Act, ch. 912, pt. 1, § 160.

In addition, as the parties stipulated in the District Court, the Commonwealth provides other financial support and assistance to private institutions—including single-sex colleges—through low-cost building loans, state-funded services contracts, and other programs. See, e. g., Va. Code Ann. §§ 23-30.39 to 23.30.58 (1993) (Educational Facilities Authority Act). The State Council of Higher Education for Virginia, in a 1989 document not created for purposes of this litigation but introduced into evidence, has described these various programs as a "means by which the Commonwealth can provide funding to its independent institutions, thereby helping to maintain a diverse system of higher education." Budget Initiatives 10.

[28] The Court, unfamiliar with the Commonwealth's policy of diverse and independent institutions, and in any event careless of state and local traditions, must be forgiven by Virginians for quoting a reference to "`the Charlottesville campus' " of the University of Virginia. See ante, at 538. The University of Virginia, an institution even older than VMI, though not as old as another of the Commonwealth's universities, the College of William and Mary, occupies the portion of Charlottesville known, not as the "campus," but as "the grounds." More importantly, even if it were a "campus," there would be no need to specify "the Charlottesville campus," as one might refer to the Bloomington or Indianapolis campus of Indiana University. Unlike university systems with which the Court is perhaps more familiar, such as those in New York (e. g., the State University of New York at Binghamton or Buffalo), Illinois (University of Illinois at Urbana-Champaign or at Chicago), and California (University of California, Los Angeles, or University of California, Berkeley), there is only one University of Virginia. It happens (because Thomas Jefferson lived near there) to be located at Charlottesville. To many Virginians it is known, simply, as "the University," which suffices to distinguish it from the Commonwealth's other institutions offering 4-year college instruction, which include Christopher Newport College, Clinch Valley College, the College of William and Mary, George Mason University, James Madison University, Longwood College, Mary Washington University, Norfolk State University,Old Dominion University, Radford University, Virginia Commonwealth University, Virginia Polytechnic Institute and State University, Virginia State University—and, of course, VMI.

[29] The Court's do-it-yourself approach to factfinding, which throughout is contrary to our well-settled rule that we will not "undertake to review concurrent findings of fact by two courts below in the absence of a very obvious and exceptional showing of error," Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U. S. 271, 275 (1949) (and cases cited), is exemplified by its invocation of the experience of the federal military academies to prove that not much change would occur. See ante, at 542, n. 11; 544-545, and n. 15; 550-551, n. 19. In fact, the District Court noted that "the West Point experience" supported the theory that a coeducational VMI would have to "adopt a [different] system," for West Point found it necessary upon becoming coeducational to "move away" from its adversative system. 766 F. Supp., at 1413, 1440. "Without a doubt . . . VMI's present methods of training and education would have to be changed as West Point's were." Id., at 1413, n. 8; accord, 976 F. 2d 890, 896-897 (CA4 1992) (upholding District Court's findings that "the unique characteristics of VMI's program," including its "unique methodology," "would be destroyed by coeducation").

[30] The Court is incorrect in suggesting that the Court of Appeals applied a "deferential" "brand of review inconsistent with the more exacting standard our precedent requires." Ante, at 555. That court "inquir[ed] (1) whether the state's objective is legitimate and important,' and (2) whether `the requisite direct, substantial relationship between objective and means is present,' " 44 F. 3d, at 1235 (quoting Mississippi Univ. for Women v. Hogan, 458 U. S. 718, 725 (1982)). To be sure, such review is "deferential" to a degree that the Court's new standard is not, for it is intermediate scrutiny. (The Court cannot evade this point or prove the Court of Appeals too deferential by stating that that court "devised another test, a `substantive comparability' inquiry,' " ante, at 555 (quoting 44 F. 3d, at 1237), for as that court explained, its "substantive comparability" inquiry was an "additional step" that it engrafted on "th[e] traditional test" of intermediate scrutiny, ibid. (emphasis added).)

[31] The concurrence states that it "read[s] the Court" not "as saying that the diversity rationale is a pretext" for discriminating against women, but as saying merely that the diversity rationale is not genuine. Ante, at 562, n. The Court itself makes no such disclaimer, which would be difficult to credit inasmuch as the foundation for its conclusion that the diversity rationale is not "genuin[e]," ante, at 539, is its antecedent discussion of Virginia's "deliberate" actions over the past century and a half, based on "[f]amiliar arguments," that sought to enforce once "widely held views about women's proper place," ante, at 537, 538.

[32] In this regard, I note that the Court—which I concede is under no obligation to do so—provides no example of a program that would pass muster under its reasoning today: not even, for example, a football or wrestling program. On the Court's theory, any woman ready, willing, and physically able to participate in such a program would, as a constitutional matter, be entitled to do so.

CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT.

Blanch Freedman argued the cause for petitioner. With her on the briefs was Robert Brown.

Nathan Lewin argued the cause for respondent. On the brief were Solicitor General Marshall, Assistant Attorney General Vinson and Philip R. Monahan.

Briefs of amici curiae, urging reversal, were filed by David Carliner, Nanette Dembitz and Alan H. Levine for the American Civil Liberties Union et al., and by the Homosexual Law Reform Society of America.

MR. JUSTICE CLARK delivered the opinion of the Court.

The petitioner, an alien, has been ordered deported to Canada as one who upon entry into this country was a homosexual and therefore “afflicted with psychopathic personality” and excludable under § 212 (a) (4) of the Immigration and Nationality Act of 1952, 66 Stat. 182. 8 U. S. C. § 1182 (a) (4).[*] Petitioner’s appeal from the 119*119 finding of the Special Inquiry Officer was dismissed by the Board of Immigration Appeals, without opinion, and his petition for review in the Court of Appeals was dismissed, with one judge dissenting. 363 F. 2d 488. It held that the term “psychopathic personality,” as used by the Congress in § 212 (a) (4), was a term of art intended to exclude homosexuals from entry into the United States. It further found that the term was not void for vagueness and was, therefore, not repugnant to the Fifth Amendment’s Due Process Clause. We granted certiorari, 385 U. S. 927, and now affirm.

I.

Petitioner, a Canadian national, was first admitted to this country on June 22, 1955, at the age of 21. His last entry was in 1959, at which time he was returning from a short trip to Canada. His mother and stepfather and three of his brothers and sisters live in the United States. In 1963 he applied for citizenship and submitted to the Naturalization Examiner an affidavit in which he admitted that he was arrested in New York in October 1959, on a charge of sodomy, which was later reduced to simple assault and thereafter dismissed on default of the complainant. In 1964, petitioner, at the request of the Government, submitted another affidavit which revealed the full history of his sexual deviate behavior. It stated that his first homosexual experience occurred when he was 14 years of age, some seven years before his entry into the United States. Petitioner was evidently a passive participant in this encounter. His next episode was at age 16 and occurred in a public park in Halifax, Nova Scotia. Petitioner was the active participant in this affair. During the next five years immediately preceding 120*120 his first entry into the United States petitioner had homosexual relations on an average of three or four times a year. He also stated that prior to his entry he had engaged in heterosexual relations on three or four occasions. During the eight and one-half years immediately subsequent to his entry, and up to the time of his second statement, petitioner continued to have homosexual relations on an average of three or four times a year. Since 1959 petitioner had shared an apartment with a man with whom he had had homosexual relations.

The 1964 affidavit was submitted to the Public Health Service for its opinion as to whether petitioner was excludable for any reason at the time of his entry. The Public Health Service issued a certificate in 1964 stating that in the opinion of the subscribing physicians petitioner “was afflicted with a class A condition, namely, psychopathic personality, sexual deviate” at the time of his admission. Deportation proceedings were then instituted. “No serious question,” the Special Inquiry Officer found, “has been raised either by the respondent [petitioner here], his counsel or the psychiatrists [employed by petitioner] who have submitted reports on the respondent as to his sexual deviation.” Indeed, the officer found that both of petitioner’s psychiatrists “concede that the respondent has been a homosexual for a number of years but conclude that by reason of such sexual deviation, the respondent is not a psychopathic personality.” Finding against petitioner on the facts, the issue before the officer was reduced to the purely legal question of whether the term “psychopathic personality” included homosexuals and if it suffered illegality because of vagueness.

II.

The legislative history of the Act indicates beyond a shadow of a doubt that the Congress intended the phrase “psychopathic personality” to include homosexuals such as petitioner.

121*121 Prior to the 1952 Act the immigration law excluded “persons of constitutional psychopathic inferiority.” 39 Stat. 875, as amended, 8 U. S. C. § 136 (a) (1946 ed.). Beginning in 1950, a subcommittee of the Senate Committee on the Judiciary conducted a comprehensive study of the immigration laws and in its report found “that the purpose of the provision against `persons with constitutional psychopathic inferiority’ will be more adequately served by changing that term to `persons afflicted with psychopathic personality,’ and that the classes of mentally defectives should be enlarged to include homosexuals and other sex perverts.” S. Rep. No. 1515, 81st Cong., 2d Sess., p. 345. The resulting legislation was first introduced as S. 3455 and used the new phrase “psychopathic personality.” The bill, however, contained an additional clause providing for the exclusion of aliens “who are homosexuals or sex perverts.” As the legislation progressed (now S. 2550 in the 82d Congress), however, it omitted the latter clause “who are homosexuals or sex perverts” and used only the phrase “psychopathic personality.” The omission is explained by the Judiciary Committee Report on the bill:

“The provisio[n] of S. 716 [one of the earlier bills not enacted] which specifically excluded homosexuals and sex perverts as a separate excludable class does not appear in the instant bill. The Public Health Service has advised that the provision for the exclusion of aliens afflicted with psychopathic personality or a mental defect which appears in the instant bill is sufficiently broad to provide for the exclusion of homosexuals and sex perverts. This change of nomenclature is not to be construed in any way as modifying the intent to exclude all aliens who are sexual deviates.” (Emphasis supplied.) S. Rep. No. 1137, 82d Cong., 2d Sess., p. 9.

122*122 Likewise, a House bill, H. R. 5678, adopted the position of the Public Health Service that the phrase “psychopathic personality” excluded from entry homosexuals and sex perverts. The report that accompanied the bill shows clearly that the House Judiciary Committee adopted the recommendation of the Public Health Service that “psychopathic personality” should be used in the Act as a phrase that would exclude from admission homosexuals and sex perverts. H. R. Rep. No. 1365, 82d Cong., 2d Sess. It quoted at length, and specifically adopted, the Public Health Service report which recommended that the term “psychopathic personality” be used to “specify such types of pathologic behavior as homosexuality or sexual perversion.” We, therefore, conclude that the Congress used the phrase “psychopathic personality” not in the clinical sense, but to effectuate its purpose to exclude from entry all homosexuals and other sex perverts.

Petitioner stresses that only persons afflicted with psychopathic personality are excludable. This, he says, is “a condition, physical or psychiatric, which may be manifested in different ways, including sexual behavior.” Petitioner’s contention must fall by his own admissions. For over six years prior to his entry petitioner admittedly followed a continued course of homosexual conduct. The Public Health Service doctors found and certified that at the time of his entry petitioner “was afflicted with a class A condition, namely, psychopathic personality, sexual deviate . . . .” It was stipulated that if these doctors were to appear in the case they would testify to this effect and that “no useful purpose would be served by submitting this additional psychiatric material [furnished by petitioner’s doctors] to the United States Public Health Service . . . .” The Government clearly established that petitioner was a homosexual at entry. Having substantial support in the record, we do not now disturb that finding, especially since petitioner admitted 123*123 being a homosexual at the time of his entry. The existence of this condition over a continuous and uninterrupted period prior to and at the time of petitioner’s entry clearly supports the ultimate finding upon which the order of deportation was based.

III.

Petitioner says, even so, the section as construed is constitutionally defective because it did not adequately warn him that his sexual affliction at the time of entry could lead to his deportation. It is true that this Court has held the “void for vagueness” doctrine applicable to civil as well as criminal actions. See Small Co. v. Am. Sugar Ref. Co., 267 U. S. 233, 239 (1925). However, this is where “the exaction of obedience to a rule or standard. . . was so vague and indefinite as really to be no rule or standard at all. . . .” In short, the exaction must strip a participant of his rights to come within the principle of the cases. But the “exaction” of § 212 (a) (4) never applied to petitioner’s conduct after entry. The section imposes neither regulation of nor sanction for conduct. In this situation, therefore, no necessity exists for guidance so that one may avoid the applicability of the law. The petitioner is not being deported for conduct engaged in after his entry into the United States, but rather for characteristics he possessed at the time of his entry. Here, when petitioner first presented himself at our border for entrance, he was already afflicted with homosexuality. The pattern was cut, and under it he was not admissible.

The constitutional requirement of fair warning has no applicability to standards such as are laid down in § 212 (a) (4) for admission of aliens to the United States. It has long been held that the Congress has plenary power to make rules for the admission of aliens and to exclude those who possess those characteristics which Congress has forbidden. See The Chinese Exclusion Case, 124*124 130 U. S. 581 (1889). Here Congress commanded that homosexuals not be allowed to enter. The petitioner was found to have that characteristic and was ordered deported. The basis of the deportation order was his affliction for a long period of time prior to entry, i. e., six and one-half years before his entry. It may be, as some claim, that “psychopathic personality” is a medically ambiguous term, including several separate and distinct afflictions. Noyes, Modern Clinical Psychiatry 410 (3d ed. 1948). But the test here is what the Congress intended, not what differing psychiatrists may think. It was not laying down a clinical test, but an exclusionary standard which it declared to be inclusive of those having homosexual and perverted characteristics. It can hardly be disputed that the legislative history of § 212 (a) (4) clearly shows that Congress so intended.

But petitioner says that he had no warning and that no interpretation of the section had come down at the time of his 1955 entry. Therefore, he argues, he was unaware of the fact that homosexual conduct engaged in after entry could lead to his deportation. We do not believe that petitioner’s post-entry conduct is the basis for his deportation order. At the time of his first entry he had continuously been afflicted with homosexuality for over six years. To us the statute is clear. It fixes “the time of entry” as the crucial date and the record shows that the findings of the Public Health Service doctors and the Special Inquiry Officer all were based on that date. We find no indication that the post-entry evidence was of any consequence in the ultimate decision of the doctors, the hearing officer or the court. Indeed, the proof was uncontradicted as to petitioner’s characteristic at the time of entry and this brought him within the excludable class. A standard applicable solely to time of entry could hardly be vague as to post-entry conduct.

125*125 The petitioner raises other points, including the claim that an “arriving alien” under the Act is entitled to medical examination. Since he is not an “arriving alien” subject to exclusion, but a deportable alien within an excludable class—who through error was permitted entry— it is doubtful if the requirement would apply. But we need not go into the question since petitioner was twice offered examination and refused to submit himself. He can hardly be heard to complain now. The remaining contentions are likewise without merit.

Affirmed.

MR. JUSTICE BRENNAN dissents for the reasons stated by Judge Moore of the Court of Appeals, 363 F. 2d 488, 496-499.

MR. JUSTICE DOUGLAS, with whom MR. JUSTICE FORTAS concurs, dissenting.

The term “psychopathic personality” is a treacherous one like “communist” or in an earlier day “Bolshevik.” A label of this kind when freely used may mean only an unpopular person. It is much too vague by constitutional standards for the imposition of penalties or punishment.

Cleckley defines “psychopathic personality” as one who has the following characteristics:

(1) Superficial charm and good “intelligence.” (2) Absence of delusions and other signs of irrational “thinking.” (3) Absence of “nervousness” or psycho-neurotic manifestations. (4) Unreliability. (5) Untruthfulness and insincerity. (6) Lack of remorse or shame. (7) Inadequately motivated antisocial behavior. (8) Poor judgment and failure to learn by experience. (9) Pathologic egocentricity and incapacity for love. (10) General poverty in major affective reactions. (11) Specific loss of insight. 126*126 (12) Unresponsiveness in general interpersonal relations. (13) Fantastic and uninviting behavior with drink and sometimes without. (14) Suicide rarely carried out. (15) Sex life impersonal, trivial and poorly integrated. (16) Failure to follow any life plan. Cleckley, The Mask of Sanity 238-255 (1941).

The word “psychopath” according to some means “a sick mind.” Guttmacher & Weihofen, Psychiatry and the Law 86 (1952):

“In the light of present knowledge, most of the individuals called psychopathic personalities should probably be considered as suffering from neurotic character disorders. They are, for the most part, unhappy persons, harassed by tension and anxiety, who are struggling against unconscious conflicts which were created during the very early years of childhood. The nature and even the existence of these conflicts which drive them restlessly on are unknown to them. When the anxiety rises to a certain pitch, they seek relief through some antisocial act. The frequency with which this pattern recurs in the individual is dependent in part upon the intensity of the unconscious conflict, upon the tolerance for anxiety, and upon chance environmental situations which may heighten or decrease it. One of the chief diagnostic criteria of this type of neurotically determined delinquency is the repetitiveness of the pattern. The usual explanation, as for example, that the recidivistic check-writer has just `got in the habit of writing bad checks’ is meaningless.” Id., at 88-89.

Many experts think that it is a meaningless designation. “Not yet is there any common agreement . . . as to classification or . . . etiology.” Noyes, Modern Clinical Psychiatry 410 (3d ed. 1948). “The only conclusion that seems warrantable is that, at some time or 127*127 other and by some reputable authority, the term psychopathic personality has been used to designate every conceivable type of abnormal character.” Curran & Mallinson, Psychopathic Personality, 90 J. Mental Sci. 266, 278. See also Guttmacher, Diagnosis and Etiology of Psychopathic Personalities as Perceived in Our Time, in Current Problems in Psychiatric Diagnosis 139, 154 (Hoch & Zubin ed. 1953); Tappan, Sexual Offences and the Treatment of Sexual Offenders in the United States, in Sexual Offences 500, 507 (Radzinowicz ed. 1957). It is much too treacherously vague a term to allow the high penalty of deportation to turn on it.

When it comes to sex, the problem is complex. Those “who fail to reach sexual maturity (hetero-sexuality), and who remain at a narcissistic or homosexual stage” are the products “of heredity, of glandular dysfunction, [or] of environmental circumstances.” Henderson, Psychopathic Constitution and Criminal Behaviour, in Mental Abnormality and Crime 105, 114 (Radzinowicz & Turner ed. 1949).

The homosexual is one, who by some freak, is the product of an arrested development:

“All people have originally bisexual tendencies which are more or less developed and which in the course of time normally deviate either in the direction of male or female. This may indicate that a trace of homosexuality, no matter how weak it may be, exists in every human being. It is present in the adolescent stage, where there is a considerable amount of undifferentiated sexuality.” Abrahamsen, Crime and the Human Mind 117 (1944).

Many homosexuals become involved in violations of laws; many do not. Kinsey reported:

“It is not possible to insist that any departure from the sexual mores, or any participation in socially taboo activities, always, or even usually, involves 128*128 a neurosis or psychosis, for the case histories abundantly demonstrate that most individuals who engage in taboo activities make satisfactory social adjustments. There are, in actuality, few adult males who are particularly disturbed over their sexual histories. Psychiatrists, clinical psychologists, and others who deal with cases of maladjustment, sometimes come to feel that most people find difficulty in adjusting their sexual lives; but a clinic is no place to secure incidence figures. The incidence of tuberculosis in a tuberculosis sanitarium is no measure of the incidence of tuberculosis in the population as a whole; and the incidence of disturbance over sexual activities, among the persons who come to a clinic, is no measure of the frequency of similar disturbances outside of clinics. The impression that such `sexual irregularities’ as `excessive’ masturbation, pre-marital intercourse, responsibility for a pre-marital pregnancy, extra-marital intercourse, mouth-genital contacts, homosexual activity, or animal intercourse, always produce psychoses and abnormal personalities is based upon the fact that the persons who do go to professional sources for advice are upset by these things. “It is unwarranted to believe that particular types of sexual behavior are always expressions of psychoses or neuroses. In actuality, they are more often expressions of what is biologically basic in mammalian and anthropoid behavior, and of a deliberate disregard for social convention. Many of the socially and intellectually most significant persons in our histories, successful scientists, educators, physicians, clergymen, business men, and persons of high position in governmental affairs, have socially taboo items in their sexual histories, and among them they have accepted nearly the whole range 129*129 of so-called sexual abnormalities. Among the socially most successful and personally best adjusted persons who have contributed to the present study, there are some whose rates of outlet are as high as those in any case labelled nymphomania or satyriasis in the literature, or recognized as such in the clinic.” Kinsey, Sexual Behavior in the Human Male 201-202 (1948).

It is common knowledge that in this century homosexuals have risen high in our own public service—both in Congress and in the Executive Branch—and have served with distinction. It is therefore not credible that Congress wanted to deport everyone and anyone who was a sexual deviate, no matter how blameless his social conduct had been nor how creative his work nor how valuable his contribution to society. I agree with Judge Moore, dissenting below, that the legislative history should not be read as imputing to Congress a purpose to classify under the heading “psychopathic personality” every person who had ever had a homosexual experience:

“Professor Kinsey estimated that `at least 37 per cent’ of the American male population has at least one homosexual experience, defined in terms of physical contact to the point of orgasm, between the beginning of adolescence and old age.[1] Kinsey, Pomeroy 130*130 & Martin, Sexual Behavior in the Human Male 623 (1948). Earlier estimates had ranged from one per cent to 100 per cent. Id. at 616-622. The sponsors of Britain’s current reform bill on homosexuality have indicated that one male in 25 is a homosexual in Britain.[2] To label a group so large `excludable aliens’ would be tantamount to saying that Sappho, Leonardo da Vinci, Michelangelo, Andre Gide, and perhaps even Shakespeare, were they to come to life again, would be deemed unfit to visit our shores.[3] Indeed, so broad a definition might well comprise more than a few members of legislative bodies.” 363 F. 2d 488, 497-498.

The Public Health Service, from whom Congress borrowed the term “psychopathic personality” (H. R. Rep. No. 1365, 82d Cong., 2d Sess., 46-47) admits that the term is “vague and indefinite.” Id., at 46.

131*131 If we are to hold, as the Court apparently does, that any acts of homosexuality suffice to deport the alien, whether or not they are part of a fabric of antisocial behavior, then we face a serious question of due process. By that construction a person is judged by a standard that is almost incapable of definition. I have already quoted from clinical experts to show what a wide range the term “psychopathic personality” has. Another expert⁴ classifies such a person under three headings:

Acting: (1) inability to withstand tedium, (2) lack of a sense of responsibility, (3) a tendency to “blow up” under pressure, (4) maladjustment to law and order, and (5) recidivism.

Feeling: they tend to (1) be emotionally deficient, narcissistic, callous, inconsiderate, and unremorseful, generally projecting blame on others, (2) have hair-trigger emotions, exaggerated display of emotion, and be irritable and impulsive, (3) be amoral (socially and sexually) and (4) worry, but do nothing about it.

Thinking: they display (1) defective judgment, living for the present rather than for the future, and (2) inability to profit from experience, i. e., they are able to realize the consequences intelligently, but not to evaluate them.

We held in Jordan v. De George, 341 U. S. 223, that the crime of a conspiracy to defraud the United States of taxes involved “moral turpitude” and made the person subject to deportation. That, however, was a term that has “deep roots in the law.” Id., at 227. But the grab-bag —”psychopathic personality”—has no “deep roots” whatsoever.[5] Caprice of judgment is almost certain under this broad definition. Anyone can be caught who is unpopular, who is off-beat, who is nonconformist.

132*132 Deportation is the equivalent to banishment or exile. Fong Haw Tan v. Phelan, 333 U. S. 6, 10. Though technically not criminal, it practically may be. The penalty is so severe that we have extended to the resident alien the protection of due process. Wong Yang Sung v. McGrath, 339 U. S. 33. Even apart from deportation cases, we look with suspicion at those delegations of power so broad as to allow the administrative staff the power to formulate the fundamental policy. See Watkins v. United States, 354 U. S. 178, 203-205; Kent v. Dulles, 357 U. S. 116. In the Watkins case we were protecting important First Amendment rights. In the Kent case we were protecting the right to travel, an important ingredient of a person’s “liberty” within the meaning of the Fifth Amendment. We deal here also with an aspect of “liberty” and the requirements of due process. They demand that the standard be sufficiently clear as to fore-warn those who may otherwise be entrapped and to provide full opportunity to conform. “Psychopathic personality” is so broad and vague as to be hardly more than an epithet. The Court seeks to avoid this question by saying that the standard being applied relates only to what petitioner had done prior to his entry, not to his postentry conduct. But at least half of the questioning of this petitioner related to his postentry conduct.

Moreover, the issue of deportability under § 212 (a) of the Immigration and Nationality Act of 1952 turns on whether petitioner is “afflicted with psychopathic personality.” On this I think he is entitled to a hearing to satisfy both the statute and the requirement of due process.

One psychiatrist reported:

“On psychiatric examination of Mr. Boutilier, there was no indication of delusional trend or hallucinatory phenomena. He is not psychotic. From his own account, he has a psychosexual problem 133*133 but is beginning treatment for this disorder. Diagnostically, I would consider him as having a Character Neurosis, believe that the prognosis in therapy is reasonably good and do not think he represents any risk of decompensation into a dependent psychotic reaction nor any potential for frank criminal activity.”

Another submitted a long report ending as follows:

“The patient’s present difficulties obviously weigh very heavily upon him. He feels as if he has made his life in this country and is deeply disturbed at the prospect of being cut off from the life he has created for himself. He talks frankly about himself. What emerged out of the interview was not a picture of a psychopath but that of a dependent, immature young man with a conscience, an awareness of the feelings of others and a sense of personal honesty. His sexual structure still appears fluid and immature so that he moves from homosexual to heterosexual interests as well as abstinence with almost equal facility. His homosexual orientation seems secondary to a very constricted, dependent personality pattern rather than occurring in the context of a psychopathic personality. My own feeling is that his own need to fit in and be accepted is so great that it far surpasses his need for sex in any form.

“I do not believe that Mr. Boutilier is a psychopath.”

In light of these statements, I cannot say that it has been determined that petitioner was “afflicted” in the statutory sense either at the time of entry or at present. “Afflicted” means possessed or dominated by. Occasional acts would not seem sufficient. “Afflicted” means a way of life, an accustomed pattern of conduct. Whatever disagreement there is as to the meaning of “psychopathic 134*134 personality,” it has generally been understood to refer to a consistent, lifelong pattern of behavior conflicting with social norms without accompanying guilt. Cleckley, supra, at 29.[6] Nothing of that character was 135*135 shown to exist at the time of entry. The fact that he presently has a problem, as one psychiatrist said, does not mean that he is or was necessarily “afflicted” with homosexuality. His conduct is, of course, evidence material to the issue. But the informed judgment of experts is needed to make the required finding. We cruelly mutilate the Act when we hold otherwise. For we make the word of the bureaucrat supreme, when it was the expertise of the doctors and psychiatrists on which Congress wanted the administrative action to be dependent.

[*] “SEC. 212. (a) Except as otherwise provided in this Act, the following classes of aliens shall be ineligible to receive visas and shall be excluded from admission into the United States:

.....

”(4) Aliens afflicted with psychopathic personality, epilepsy, or a mental defect . . . .”

Section 241 (a) (1) of the Immigration and Nationality Act, 66 Stat. 204, 8 U. S. C. § 1251 (a) (1), provides that: “Any alien in the United States . . . shall, upon the order of the Attorney General, be deported who—(1) at the time of entry was within one or more of the classes of aliens excludable by the law existing at the time of such entry . . . .”

[1] “Homosexual activity in the human male is much more frequent than is ordinarily realized . . . . In the youngest unmarried group, more than a quarter (27.3%) of the males have some homosexual activity to the point of orgasm . . . . The incidence among these single males rises in successive age groups until it reaches a maximum of 38.7 per cent between 36 and 40 years of age.

“High frequencies do not occur as often in the homosexual as they do in some other kinds of sexual activity . . . . Populations are more homogeneous in regard to this outlet. This may reflect the difficulties involved in having frequent and regular relations in a socially taboo activity. Nevertheless, there are a few of the younger adolescent males who have homosexual frequencies of 7 or more per week, and between 26 and 30 the maximum frequencies run to 15 per week. By 50 years of age the most active individual is averaging only 5.0 per week.

“For single, active populations, the mean frequencies of homosexual contacts . . . rise more or less steadily from near once per week . . . for the younger adolescent boys to nearly twice as often . . . for males between the ages of 31 and 35. They stand above once a week through age 50.” Kinsey, Sexual Behavior in the Human Male 259-261 (1948).

[2] Report, Committee on Homosexual Offenses and Prostitution (1957).

[3] Sigmund Freud wrote in 1935:

“Homosexuality is assuredly no advantage, but it is nothing to be ashamed of, no vice, no degradation, it cannot be classified as an illness; we consider it to be a variation of the sexual function produced by a certain arrest of sexual development. Many highly respectable individuals of ancient and modern times have been homosexuals, several of the greatest men among them (Plato, Michelangelo, Leonardo da Vinci, etc.). It is a great injustice to persecute homosexuality as a crime, and cruelty too. If you do not believe me, read the books of Havelock Ellis.” Ruitenbeek, The Problem of Homosexuality in Modern Society 1 (1963).

[4] Caldwell, Constitutional Psychopathic State (Psychopathic Personality) Studies of Soldiers in the U. S. Army, 3 J. Crim. Psychopathology 171-172 (1941).

[5] See Lindman & McIntyre, The Mentally Disabled and the Law 299 (1961).

[6] There is good indication that Congress intended the term “afflicted with psychopathic personality” to refer only to those individuals demonstrating “developmental defects or pathological trends in the personality structure manifest[ed] by lifelong patterns of action or behavior . . . .” U. S. Public Health Service, Report on Medical Aspects of H. R. 2379, U. S. Code Cong. & Admin. News 1700 (1952). The provision for exclusion of persons afflicted with psychopathic personality replaced the section of the 1917 Act, 39 Stat. 875, providing for the exclusion of “persons of constitutional psychopathic inferiority.” The purpose of that clause was “to keep out `tainted blood,’ that is, `persons who have medical traits which would harm the people of the United States if those traits were introduced in this country, or if those possessing those traits were added to those in this country who unfortunately are so afflicted.’ ” The Immigration and Naturalization System of the United States, S. Rep. No. 1515, 81st Cong., 2d Sess., 343 (1950). The Senate subcommittee which had been charged with making an investigation of the immigration laws concluded that “the exclusion of persons with `constitutional psychopathic inferiority’ was aimed at keeping out of the country aliens with a propensity to mental aberration, those with an inherent likelihood of becoming mental cases, as indicated by their case history.” Ibid. It concluded that “the purpose of the provision against `persons with constitutional psychopathic inferiority’ will be more adequately served by changing that term to `persons afflicted with psychopathic personality,’ and that the classes of mentally defectives should be enlarged to include homosexuals and other sex perverts.” Id., at 345. Senate Report 1515 accompanied Senate bill 3455, which included among excludable aliens ”[a]liens afflicted with psychopathic personality,” and ”[a]liens who are homosexuals or sex perverts.” The bill was redrafted and became S. 716, with its counterpart in the House being H. R. 2379; the material provisions remained the same as in S. 3455. In response to the House’s request for its opinion on the new provisions, the Public Health Service noted that:

“The conditions classified within the group of psychopathic personalities are, in effect, disorders of the personality. They are characterized by developmental defects or pathological trends in the personality structure manifest by lifelong patterns of action or behavior, rather than by mental or emotional symptoms. Individuals with such a disorder may manifest a disturbance of intrinsic personality patterns, exaggerated personality trends, or are persons ill primarily in terms of society and the prevailing culture. The latter or sociopathic reactions are frequently symptomatic of a severe underlying neurosis or psychosis and frequently include those groups of individuals suffering from addiction or sexual deviation.” U. S. Code Cong. & Admin. News 1700 (1952).

The letter setting forth the views of the Public Health Service went on to say, with respect to the exclusion of “homosexuals or sex perverts”:

“Ordinarily, persons suffering from disturbances in sexuality are included within the classification of `psychopathic personality with pathologic sexuality.’ This classification will specify such types of pathologic behavior as homosexuality or sexual perversion which includes sexual sadism, fetishism, transvestism, pedophilia, etc.” Id., at 1701. The bill which was finally enacted, H. R. 5678, provided for exclusion of ”[a]liens afflicted with psychopathic personality,” but did not provide for exclusion of aliens who are homosexuals or sex perverts, as had its predecessors. The House Report, H. R. Rep. No. 1365, which accompanied the bill incorporated the full report of the Public Health Service (H. R. Rep. No. 1365, 82d Cong., 2d Sess., at 46-48) and indicated that the “recommendations contained in the . . . report have been followed.” Id., at 48.

This legislative history indicates that the term “afflicted with psychopathic personality” was used in a medical sense and was meant to refer to lifelong patterns of action that are pathologic and symptomatic of grave underlying neurosis or psychosis. Homosexuality and sex perversion, as a subclass, are limited to the same afflictions.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF MICHIGAN.

Anne R. Davidow argued the cause and filed a brief for appellants. Larry S. Davidow was also of counsel.

Edmund E. Shepherd, Solicitor General of Michigan, argued the cause for appellees. With him on the brief were Eugene F. Black, Attorney General, Daniel J. O’Hara and Charles M.A. Martin, Assistant Attorneys General.

465*465 MR. JUSTICE FRANKFURTER delivered the opinion of the Court.

As part of the Michigan system for controlling the sale of liquor, bartenders are required to be licensed in all cities having a population of 50,000 or more, but no female may be so licensed unless she be “the wife or daughter of the male owner” of a licensed liquor establishment. Section 19a of Act 133 of the Public Acts of Michigan, 1945, Mich. Stat. Ann. § 18.990 (1) (Cum. Supp. 1947). The case is here on direct appeal from an order of the District Court of three judges, convened under § 266 of the old Judicial Code, now 28 U.S.C. § 2284, denying an injunction to restrain the enforcement of the Michigan law. The claim, denied below, one judge dissenting, 74 F. Supp. 735, and renewed here, is that Michigan cannot forbid females generally from being barmaids and at the same time make an exception in favor of the wives and daughters of the owners of liquor establishments. Beguiling as the subject is, it need not detain us long. To ask whether or not the Equal Protection of the Laws Clause of the Fourteenth Amendment barred Michigan from making the classification the State has made between wives and daughters of owners of liquor places and wives and daughters of non-owners, is one of those rare instances where to state the question is in effect to answer it.

We are, to be sure, dealing with a historic calling. We meet the alewife, sprightly and ribald, in Shakespeare, but centuries before him she played a role in the social life of England. See, e.g., Jusserand, English Wayfaring Life in the Middle Ages, 133, 134, 136-37 (1889). The Fourteenth Amendment did not tear history up by the roots, and the regulation of the liquor traffic is one of the oldest and most untrammeled of legislative powers. Michigan could, beyond question, forbid all women from working behind a bar. This is so despite the vast changes 466*466 in the social and legal position of women. The fact that women may now have achieved the virtues that men have long claimed as their prerogatives and now indulge in vices that men have long practiced, does not preclude the States from drawing a sharp line between the sexes, certainly in such matters as the regulation of the liquor traffic. See the Twenty-First Amendment and Carter v. Virginia, 321 U.S. 131. The Constitution does not require legislatures to reflect sociological insight, or shifting social standards, any more than it requires them to keep abreast of the latest scientific standards.

While Michigan may deny to all women opportunities for bartending, Michigan cannot play favorites among women without rhyme or reason. The Constitution in enjoining the equal protection of the laws upon States precludes irrational discrimination as between persons or groups of persons in the incidence of a law. But the Constitution does not require situations “which are different in fact or opinion to be treated in law as though they were the same.” Tigner v. Texas, 310 U.S. 141, 147. Since bartending by women may, in the allowable legislative judgment, give rise to moral and social problems against which it may devise preventive measures, the legislature need not go to the full length of prohibition if it believes that as to a defined group of females other factors are operating which either eliminate or reduce the moral and social problems otherwise calling for prohibition. Michigan evidently believes that the oversight assured through ownership of a bar by a barmaid’s husband or father minimizes hazards that may confront a barmaid without such protecting oversight. This Court is certainly not in a position to gainsay such belief by the Michigan legislature. If it is entertainable, as we think it is, Michigan has not violated its duty to afford equal protection of its laws. We cannot cross-examine either actually or argumentatively the mind of Michigan legislators 467*467 nor question their motives. Since the line they have drawn is not without a basis in reason, we cannot give ear to the suggestion that the real impulse behind this legislation was an unchivalrous desire of male bartenders to try to monopolize the calling.

It would be an idle parade of familiar learning to review the multitudinous cases in which the constitutional assurance of the equal protection of the laws has been applied. The generalities on this subject are not in dispute; their application turns peculiarly on the particular circumstances of a case. Thus, it would be a sterile inquiry to consider whether this case is nearer to the nepotic pilotage law of Louisiana, sustained in Kotch v. Pilot Commissioners, 330 U.S. 552, than it is to the Oklahoma sterilization law, which fell in Skinner v. Oklahoma, 316 U.S. 535. Suffice it to say that “A statute is not invalid under the Constitution because it might have gone farther than it did, or because it may not succeed in bringing about the result that it tends to produce.” Roschen v. Ward, 279 U.S. 337, 339.

Nor is it unconstitutional for Michigan to withdraw from women the occupation of bartending because it allows women to serve as waitresses where liquor is dispensed. The District Court has sufficiently indicated the reasons that may have influenced the legislature in allowing women to be waitresses in a liquor establishment over which a man’s ownership provides control. Nothing need be added to what was said below as to the other grounds on which the Michigan law was assailed.

Judgment affirmed.

MR. JUSTICE RUTLEDGE, with whom MR. JUSTICE DOUGLAS and MR. JUSTICE MURPHY join, dissenting.

While the equal protection clause does not require a legislature to achieve “abstract symmetry”[1] or to classify 468*468 with “mathematical nicety,”[2] that clause does require lawmakers to refrain from invidious distinctions of the sort drawn by the statute challenged in this case.[3]

The statute arbitrarily discriminates between male and female owners of liquor establishments. A male owner, although he himself is always absent from his bar, may employ his wife and daughter as barmaids. A female owner may neither work as a barmaid herself nor employ her daughter in that position, even if a man is always present in the establishment to keep order. This inevitable result of the classification belies the assumption that the statute was motivated by a legislative solicitude for the moral and physical well-being of women who, but for the law, would be employed as barmaids. Since there could be no other conceivable justification for such discrimination against women owners of liquor establishments, the statute should be held invalid as a denial of equal protection.

[1] Patsone v. Pennsylvania, 232 U.S. 138, 144.

[2] Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78-82; see also Tigner v. Texas, 310 U.S. 141, 147; Bain Peanut Co. v. Pinson, 282 U.S. 499, 501; Bryant v. Zimmerman, 278 U.S. 63, 73-77; Miller v. Wilson, 236 U.S. 373, 384.

[3] Cf. Skinner v. Oklahoma, 316 U.S. 535; Missouri ex rel. Gaines v. Canada, 305 U.S. 337; McCabe v. Atchison, T. & S.F.R. Co., 235 U.S. 151; Yick Wo v. Hopkins, 118 U.S. 356. And see Kotch v. Pilot Commissioners, 330 U.S. 552, dissenting opinion 564.