Mr. Chief Justice Warren

delivered the opinion of the Court.

This case presents a constitutional question never addressed by this Court: whether a statutory scheme adopted by the State of Virginia to prevent marriages between persons solely on the basis of racial classifications violates the Equal Protection and Due Process Clauses of the Fourteenth Amendment.1 For reasons which seem to us to reflect the central meaning of those constitutional commands, we conclude that these statutes cannot stand consistently with the Fourteenth Amendment.

In June 1958, two residents of Virginia, Mildred Jeter, a Negro woman, and Richard Loving, a white man, were married in the District of Columbia pursuant to its laws. Shortly after their marriage, the Lovings returned to Virginia and established their marital abode in Caroline County. At the October Term, 1958, of the Circuit Court *3of Caroline County, a grand jury issued an indictment charging the Lovings with violating Virginia’s ban on interracial marriages. On January 6, 1959, the Lovings pleaded guilty to the charge and were sentenced to one year in jail; however, the trial judge suspended the sentence for a period of 25 years on the condition that the Lovings leave the State and not return to Virginia together for 25 years. He stated in an opinion that:

“Almighty God created the races white, black, yellow, malay and red, and he placed them on separate continents. And but for the interference with his arrangement there would be no cause for such marriages. The fact that he separated the races shows that he did not intend for the races to mix.”

After their convictions, the Lovings took up residence in the District of Columbia. On November 6, 1963, they filed a motion in the state trial court to vacate the judgment and set aside the sentence on the ground that the statutes which they had violated were repugnant to the Fourteenth Amendment. The motion not having been decided by October 28, 1964, the Lovings instituted a class action in the United States District Court for the Eastern District of Virginia requesting that a three-judge court be convened to declare the Virginia antimiscegenation statutes unconstitutional and to enjoin state- officials from enforcing their convictions. On January 22, 1965, the state trial judge denied the motion to vacate the sentences, and the Lovings perfected an appeal to the Supreme Court of Appeals of Virginia. On February 11, 1965, the three-judge District Court continued the case to allow the Lovings to present their constitutional claims to the highest state court.

The Supreme Court of Appeals upheld the constitutionality of the antimiscegenation statutes and, after *4modifying the sentence, affirmed the convictions.2 The Lovings appealed this decision, and we noted probable jurisdiction on December 12, 1966, 385 U. S. 986.

The two statutes under which appellants were convicted and sentenced are part of a comprehensive statutory scheme aimed at prohibiting and punishing interracial marriages. The Lovings were convicted of violating § 20-58 of the Virginia Code:

“Leaving State to evade law. — If any white person and colored person shall go out of this State, for the purpose of being married, and with the intention of returning, and be married out of it, and afterwards return to and reside in it, cohabiting as man and wife, they shall be punished as provided in § 20-59, and the marriage shall be governed by the same law as if it had been solemnized in this State. The fact of their cohabitation here as man and wife shall be evidence of their marriage.”

Section 20-59, which defines the penalty for miscegenation, provides:

“Punishment for marriage. — If any white person intermarry with a colored person, or any colored person intermarry with a white person, he shall be guilty of a felony and shall be punished by confinement in the penitentiary for not less than one nor more than five years.”

Other central provisions in the Virginia statutory scheme are § 20-57, which automatically voids all marriages between “a white person and a colored person” without any judicial proceeding,3 and §§ 20-54 and 1-14 which, *5respectively, define “white persons” and “colored persons and Indians” for purposes of the statutory prohibitions.4 The Lovings have never disputed in the course of this litigation that Mrs. Loving is a “colored person” or that Mr. Loving is a “white person” within the meanings given those terms by the Virginia statutes.

*6Virginia is now one of 16 States which prohibit and punish marriages on the basis of racial classifications.5 Penalties for miscegenation arose as an incident to slavery and have been common in Virginia since the colonial period.6 The present statutory scheme dates from the adoption of the Racial Integrity Act of 1924, passed during the period of extreme nativism which followed the end of the First World War. The central features of this Act, and current Virginia law, are the absolute prohibition of a “white person” marrying other than another “white person,” 7 a prohibition against issuing marriage licenses until the issuing official is satisfied that *7the applicants’ statements as to their race are correct,8 certificates of “racial composition” to be kept by both local and state registrars,9 and the carrying forward of earlier prohibitions against racial’ intermarriage.10

I.

In upholding the constitutionality of these provisions in the decision below, the Supreme Court of Appeals of Virginia referred to its 1955 decision in Naim v. Naim, 197 Va. 80, 87 S. E. 2d 749, as stating the reasons supporting the validity of these laws. In Naim, the state court concluded that the State’s legitimate purposes were “to preserve the racial integrity of its citizens,” and to prevent “the corruption of blood,” “a mongrel breed of citizens,” and “the obliteration of racial pride,” obviously an endorsement of the doctrine of White Supremacy. Id., at 90, 87 S. E. 2d, at 756. The court also reasoned that marriage has traditionally been subject to state regulation without federal intervention, and, consequently, the regulation of marriage should be left to exclusive state control by the Tenth Amendment.

While the state court is no doubt correct in asserting that marriage is a social relation subject to the State’s police power, Maynard v. Hill, 125 U. S. 190 (1888), the State does not contend in its argument before this Court that its powers to regulate marriage are un--limited notwithstanding the commands of the Fourteenth Amendment. Nor could it do so in light of Meyer v. Nebraska, 262 U. S. 390 (1923), and Skinner v. Oklahoma, 316 U. S. 535 (1942). Instead, the State argues that the meaning of the Equal Protection Clause, as illuminated by the statements of the Framers, is only that state penal laws containing an interracial element *8as part of the definition of the offense must apply equally to whites and Negroes in the sense that members of each race are punished to the same degree. Thus, the State contends that, because its miscegenation statutes punish equally both the white and the Negro participants in an interracial marriage, these statutes, despite their reliance on racial classifications, do not constitute an invidious discrimination based upon race. The second argument advanced by the State assumes the validity of its equal application theory. The argument is that, if the Equal Protection Clause does not outlaw miscegenation statutes because of their reliance on racial classifications, the question of constitutionality would thus become whether there was any rational basis for a State to treat interracial marriages differently from other marriages. On this question, the State argues, the scientific evidence is substantially in doubt and, consequently, this Court should defer to the wisdom of the state legislature in adopting its policy of discouraging interracial marriages.

Because we reject the notion that the mere “equal application” of a statute containing racial classifications is enough to remove the classifications from the Fourteenth Amendment’s proscription of all invidious racial discriminations, we do not accept the State’s contention that these statutes should be upheld if there is any possible basis for concluding that they serve a rational purpose. The mere fact of equal application does not mean that our analysis of these statutes should follow the approach we have taken in cases involving no racial discrimination where the Equal Protection Clause has been arrayed against a statute discriminating between the kinds of advertising which may be displayed on trucks in New York City, Railway Express Agency, Inc. v. New York, 336 U. S. 106 (1949), or an exemption in Ohio’s ad valorem tax for merchandise owned by a nonresident in a storage warehouse, Allied Stores of Ohio, *9Inc. v. Bowers, 358 U. S. 522 (1959). In these cases, involving distinctions not drawn according to race, the Court has merely asked whether there is any rational foundation for the discriminations, and has deferred to the wisdom of the state legislatures. In the case at bar, however, we deal with statutes containing racial classifications, and the fact of equal application does not immunize the statute from the very heavy burden of justification which the Fourteenth Amendment has traditionally required of state statutes drawn according to race.

The State argues that statements in the Thirty-ninth Congress about the time of the passage of the Fourteenth Amendment indicate that the Framers did not intend the Amendment to make unconstitutional state miscegenation laws. Many of the statements alluded to by the State concern the debates over the Freedmen’s Bureau Bill, which President Johnson vetoed, and the Civil Rights Act of 1866, 14 Stat. 27, enacted over his veto. While these statements have some relevance to the intention of Congress in submitting the Fourteenth Amendment, it must be understood that they pertained to the passage of specific statutes and not to the broader, organic purpose of a constitutional amendment. As for the various statements directly concerning the Fourteenth Amendment, we have said in connection with a related problem, that although these historical sources “cast some light” they are not sufficient to resolve the problem; “[a]t best, they are inconclusive. The most avid proponents of the post-War Amendments undoubtedly intended them to remove all legal distinctions among 'all persons bom or naturalized in the United States.’ Their opponents, just as certainly, were antagonistic to both the letter and the spirit of the Amendments and wished them to have the most limited effect.” Brown v. Board of Education, 347 U. S. 483, 489 (1954). See also Strauder *10v. West Virginia, 100 U. S. 303, 310 (1880). We have rejected the proposition that the debates in the Thirty-ninth Congress or in the state legislatures which ratified the Fourteenth Amendment supported the theory advanced by the State, that the requirement of equal protection of the laws is satisfied by penal laws defining offenses based on racial classifications so long as white and Negro participants in the offense were similarly punished. McLaughlin v. Florida, 379 U. S. 184 (1964).

The State finds support for its “equal application'’ theory in the decision of the Court in Pace v. Alabama, 106 U. S. 583 (1883). In that case, the Court upheld a conviction under an Alabama statute forbidding adultery or fornication between a white person and a Negro which imposed a greater penalty than that of a statute proscribing similar conduct by members of the same race. The Court reasoned that the statute could not be said to discriminate against Negroes because the punishment for each participant in the offense was the same. However, as recently as the 1964 Term, in rejecting the reasoning of that case, we stated “Pace represents a limited view of the Equal Protection Clause which has not withstood analysis in the subsequent decisions of this Court.” McLaughlin v. Florida, supra, at 188. As we there demonstrated, the Equal Protection Clause requires the consideration of whether the classifications drawn by any statute constitute an arbitrary and invidious discrimination. The clear and central purpose of the Fourteenth Amendment was to eliminate all official state sources of invidious racial discrimination in the States. Slaughter-House Cases, 16 Wall. 36, 71 (1873); Strauder v. West Virginia, 100 U. S. 303, 307-308 (1880); Ex parte Virginia, 100 U. S. 339, 344-345 (1880); Shelley v. Kraemer, 334 U. S. 1 (1948); Burton v. Wilmington Parking Authority, 365 U. S. 715 (1961).

*11There can be no question but that Virginia’s miscegenation statutes rest solely upon distinctions drawn according to race. The statutes proscribe generally accepted conduct if engaged in by members of different races. Over the years, this Court has consistently repudiated “[distinctions between citizens solely because of their ancestry” as being “odious to a free people whose institutions are founded upon the doctrine of equality.” Hirabayashi v. United States, 320 U. S. 81, 100 (1943). At the very least, the Equal Protection Clause demands that racial classifications, especially suspect in criminal statutes, be subjected to the “most rigid scrutiny,” Korematsu v. United States, 323 U. S. 214, 216 (1944), and, if they are ever to be upheld, they must be shown to be necessary to the accomplishment of some permissible state objective, independent of the racial discrimination which it was the object of the Fourteenth Amendment to eliminate. Indeed, two members of this Court have already stated that they “cannot conceive of a valid legislative purpose . . . which makes the color of a person’s skin the test of whether his conduct is a criminal offense.” McLaughlin v. Florida, supra, at 198 (Stewart, J., joined by Douglas, J., concurring).

There is patently no legitimate overriding purpose independent of invidious racial discrimination which justifies this classification. The fact that Virginia prohibits only interracial marriages involving white persons demonstrates that the racial classifications must stand on their own justification, as measures designed to maintain White Supremacy.11 We have consistently denied *12the constitutionality of measures which restrict the rights of citizens on account of race. There can be no doubt that restricting the freedom to marry solely because of racial classifications violates the central meaning of the Equal Protection Clause.

II.

These statutes also deprive the Lovings of liberty without due process of law in violation of the Due Process Clause of the Fourteenth Amendment. The freedom to marry has long been recognized as one of the vital personal rights essential to the orderly pursuit of happiness by free men.

Marriage is one of the “basic civil rights of man,” fundamental to our very existence and survival. Skinner v. Oklahoma, 316 U. S. 535, 541 (1942). See also Maynard v. Hill, 125 U. S. 190 (1888). To deny this fundamental freedom on so unsupportable a basis as the racial classifications embodied in these statutes, classifications so directly subversive of the principle of equality at the heart of the Fourteenth Amendment, is surely to deprive all the State’s citizens of liberty without due process of law. The Fourteenth Amendment requires that the freedom of choice to marry not be restricted by invidious racial discriminations. Under our Constitution, the freedom to marry, or not marry, a person of another race resides with the individual and cannot be infringed by the State.

These convictions must be reversed.

]It is so ordered.

*13Mr. Justice Stewart,

concurring.

I have previously expressed the belief that “it is simply not possible for a state law to be valid under our Constitution which makes the criminality of an act depend upon the race of the actor.” McLaughlin v. Florida, 379 U. S. 184, 198 (concurring opinion). Because I adhere to that belief, I concur in the judgment of the Court.

OPINION OF THE COURT

R.S. Smith, J.

We hold that the New York Constitution does not compel recognition of marriages between members of the same sex. Whether such marriages should be recognized is a question to be addressed by the Legislature.

Facts and Procedural History

Plaintiffs and petitioners (hereafter plaintiffs) are the members of 44 same-sex couples. Each couple tried unsuccessfully to obtain a marriage license. Plaintiffs then began these four lawsuits, seeking declaratory judgments that the restriction of marriage to opposite-sex couples is invalid under the State Constitution. Defendants and respondents (hereafter defendants) are the license-issuing authorities of New York City, Albany and Ithaca; the State Department of Health, which *357instructs local authorities about the issuance of marriage licenses; and the State itself. In Hernandez v Robles, Supreme Court granted summary judgment in plaintiffs’ favor; the Appellate Division reversed. In Samuels v New York State Department of Health, Matter of Kane v Marsolais and Seymour v Holcomb, Supreme Court granted summary judgment in defendants’ favor, and the Appellate Division affirmed. We now affirm the orders of the Appellate Division.

Discussion

I

All the parties to these cases now acknowledge, implicitly or explicitly, that the Domestic Relations Law limits marriage to opposite-sex couples. Some amici, however, suggest that the statute can be read to permit same-sex marriage, thus mooting the constitutional issues. We find this suggestion untenable.

Articles 2 and 3 of the Domestic Relations Law, which govern marriage, nowhere say in so many words that only people of different sexes may marry each other, but that was the universal understanding when articles 2 and 3 were adopted in 1909, an understanding reflected in several statutes. Domestic Relations Law § 12 provides that “the parties must solemnly declare . . . that they take each other as husband and wife.” Domestic Relations Law § 15 (1) (a) requires town and city clerks to obtain specified information from “the groom” and “the bride.” Domestic Relations Law § 5 prohibits certain marriages as incestuous, specifying opposite-sex combinations (brother and sister, uncle and niece, aunt and nephew), but not same-sex combinations. Domestic Relations Law § 50 says that the property of “a married woman . . . shall not be subject to her husband’s control.”

New York’s statutory law clearly limits marriage to opposite-sex couples. The more serious question is whether that limitation is consistent with the New York Constitution.

II

New York is one of many states in which supporters of same-sex marriage have asserted it as a state constitutional right. Several other state courts have decided such cases, under various state constitutional provisions and with divergent results (e.g., Goodridge v Department of Pub. Health, 440 Mass 309, 798 NE2d 941 [2003] [excluding same-sex couples from mar*358riage violates Massachusetts Constitution]; Standhardt v Superior Ct. ex rel. County of Maricopa, 206 Ariz 276, 77 P3d 451 [Ct App 2004] [constitutional right to marry under Arizona Constitution does not encompass marriage to same-sex partner]; Morrison v Sadler, 821 NE2d 15 [Ind 2005] [Indiana Constitution does not require judicial recognition of same-sex marriage]; Lewis v Harris, 378 NJ Super 168, 875 A2d 259 [2005] [limitation of marriage to members of opposite sex does not violate New Jersey Constitution]; Baehr v Lewin, 74 Haw 530, 852 P2d 44 [1993] [refusal of marriage licenses to couples of the same sex subject to strict scrutiny under Hawaii Constitution]; Baker v State, 170 Vt 194, 744 A2d 864 [1999] [denial to same-sex couples of benefits and protections afforded to married people violates Vermont Constitution]). Here, plaintiffs claim that, by limiting marriage to opposite-sex couples, the New York Domestic Relations Law violates two provisions of the State Constitution: the Due Process Clause (art I, § 6 [“No person shall be deprived of life, liberty or property without due process of law”]) and the Equal Protection Clause (art I, § 11 [“No person shall be denied the equal protection of the laws of this state or any subdivision thereof’]).

We approach plaintiffs’ claims by first considering, in section III below, whether the challenged limitation can be defended as a rational legislative decision. The answer to this question, as we show in section IV below, is critical at every stage of the due process and equal protection analysis.

Ill

It is undisputed that the benefits of marriage are many. The diligence of counsel has identified 316 such benefits in New York law, of which it is enough to summarize some of the most important: Married people receive significant tax advantages, rights in probate and intestacy proceedings, rights to support from their spouses both during the marriage and after it is dissolved, and rights to be treated as family members in obtaining insurance coverage and making health care decisions. Beyond this, they receive the symbolic benefit, or moral satisfaction, of seeing their relationships recognized by the State.

The critical question is whether a rational legislature could decide that these benefits should be given to members of opposite-sex couples, but not same-sex couples. The question is not, we emphasize, whether the Legislature must or should continue to limit marriage in this way; of course the Legislature *359may (subject to the effect of the federal Defense of Marriage Act [Pub L 104-199, 110 US Stat 2419]) extend marriage or some or all of its benefits to same-sex couples. We conclude, however, that there are at least two grounds that rationally support the limitation on marriage that the Legislature has enacted. Others have been advanced, but we will discuss only these two, both of which are derived from the undisputed assumption that marriage is important to the welfare of children.

First, the Legislature could rationally decide that, for the welfare of children, it is more important to promote stability, and to avoid instability, in opposite-sex than in same-sex relationships. Heterosexual intercourse has a natural tendency to lead to the birth of children; homosexual intercourse does not. Despite the advances of science, it remains true that the vast majority of children are born as a result of a sexual relationship between a man and a woman, and the Legislature could find that this will continue to be true. The Legislature could also find that such relationships are all too often casual or temporary. It could find that an important function of marriage is to create more stability and permanence in the relationships that cause children to be born. It thus could choose to offer an inducement — in the form of marriage and its attendant benefits — to opposite-sex couples who make a solemn, long-term commitment to each other.

The Legislature could find that this rationale for marriage does not apply with comparable force to same-sex couples. These couples can become parents by adoption, or by artificial insemination or other technological marvels, but they do not become parents as a result of accident or impulse. The Legislature could find that unstable relationships between people of the opposite sex present a greater danger that children will be born into or grow up in unstable homes than is the case with same-sex couples, and thus that promoting stability in opposite-sex relationships will help children more. This is one reason why the Legislature could rationally offer the benefits of marriage to opposite-sex couples only.

There is a second reason: The Legislature could rationally believe that it is better, other things being equal, for children to grow up with both a mother and a father. Intuition and experience suggest that a child benefits from having before his or her eyes, every day, living models of what both a man and a woman are like. It is obvious that there are exceptions to this general rule — some children who never know their fathers, or their *360mothers, do far better than some who grow up with parents of both sexes — but the Legislature could find that the general rule will usually hold.

Plaintiffs, and amici supporting them, argue that the proposition asserted is simply untrue: that a home with two parents of different sexes has no advantage, from the point of view of raising children, over a home with two parents of the same sex. Perhaps they are right, but the Legislature could rationally think otherwise.

To support their argument, plaintiffs and amici supporting them refer to social science literature reporting studies of same-sex parents and their children. Some opponents of same-sex marriage criticize these studies, but we need not consider the criticism, for the studies on their face do not establish beyond doubt that children fare equally well in same-sex and opposite-sex households. What they show, at most, is that rather limited observation has detected no marked differences. More definitive results could hardly be expected, for until recently few children have been raised in same-sex households, and there has not been enough time to study the long-term results of such child-rearing.

Plaintiffs seem to assume that they have demonstrated the irrationality of the view that opposite-sex marriages offer advantages to children by showing there is no scientific evidence to support it. Even assuming no such evidence exists, this reasoning is flawed. In the absence of conclusive scientific evidence, the Legislature could rationally proceed on the commonsense premise that children will do best with a mother and father in the home. (See Goodridge, 440 Mass at 358-359, 798 NE2d at 979-980 [Sosman, J., dissenting].) And a legislature proceeding on that premise could rationally decide to offer a special inducement, the legal recognition of marriage, to encourage the formation of opposite-sex households.

In sum, there are rational grounds on which the Legislature could choose to restrict marriage to couples of opposite sex. Plaintiffs have not persuaded us that this long-accepted restriction is a wholly irrational one, based solely on ignorance and prejudice against homosexuals. This is the question on which these cases turn. If we were convinced that the restriction plaintiffs attack were founded on nothing but prejudice — if we agreed with plaintiffs that it is comparable to the restriction in Loving v Virginia (388 US 1 [1967]), a prohibition on inter*361racial marriage that was plainly “designed to maintain White Supremacy” (id. at 11) — we would hold it invalid, no matter how long its history. As the dissent points out, a long and shameful history of racism lay behind the kind of statute invalidated in Loving.

But the historical background of Loving is different from the history underlying this case. Racism has been recognized for centuries — at first by a few people, and later by many more — as a revolting moral evil. This country fought a civil war to eliminate racism’s worst manifestation, slavery, and passed three constitutional amendments to eliminate that curse and its vestiges. Loving was part of the civil rights revolution of the 1950’s and 1960’s, the triumph of a cause for which many heroes and many ordinary people had struggled since our nation began.

It is true that there has been serious injustice in the treatment of homosexuals also, a wrong that has been widely recognized only in the relatively recent past, and one our Legislature tried to address when it enacted the Sexual Orientation Non-Discrimination Act four years ago (L 2002, ch 2). But the traditional definition of marriage is not merely a by-product of historical injustice. Its history is of a different kind.

The idea that same-sex marriage is even possible is a relatively new one. 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. A court should not lightly conclude that everyone who held this belief was irrational, ignorant or bigoted. We do not so conclude.

IV

Our conclusion that there is a rational basis for limiting marriage to opposite-sex couples leads us to hold that that limitation is valid under the New York Due Process and Equal Protection clauses, and that any expansion of the traditional definition of marriage should come from the Legislature.

This Court is the final authority as to the meaning of the New York Constitution. This does not mean, of course, that we ignore the United States Supreme Court’s interpretations of similarly worded clauses of the Federal Constitution. The governing principle is that our Constitution cannot afford less protection to our citizens than the Federal Constitution does, but it can give more (People v P.J. Video, 68 NY2d 296, 302 *362[1986]). We have at times found our Due Process Clause to be more protective of rights than its federal counterpart, usually in cases involving the rights of criminal defendants (e.g., People v LaValle, 3 NY3d 88 [2004]) or prisoners (e.g., Cooper v Morin, 49 NY2d 69 [1979]). In general, we have used the same analytical framework as the Supreme Court in considering due process cases, though our analysis may lead to different results. By contrast, we have held that our Equal Protection Clause “is no broader in coverage than the Federal provision” (Under 21, Catholic Home Bur. for Dependent Children v City of New York, 65 NY2d 344, 360 n 6 [1985]).

We find no inconsistency that is significant in this case between our due process and equal protection decisions and the Supreme Court’s. No precedent answers for us the question we face today; we reject defendants’ argument that the Supreme Court’s ruling without opinion in Baker v Nelson (409 US 810 [1972]) bars us from considering plaintiffs’ equal protection claims. But both New York and federal decisions guide us in applying the Due Process and Equal Protection clauses.

A. Due Process

In deciding the validity of legislation under the Due Process Clause, courts first inquire whether the legislation restricts the exercise of a fundamental right, one that is “deeply rooted in this Nation’s history and tradition” (Washington v Glucksberg, 521 US 702, 721 [1997], quoting Moore v East Cleveland, 431 US 494, 503 [1977] [plurality op]; Hope v Perales, 83 NY2d 563, 575 [1994]). In this case, whether the right in question is “fundamental” depends on how it is defined. The right to marry is unquestionably a fundamental right (Loving, 388 US at 12; Zablocki v Redhail, 434 US 374, 384 [1978]; Cooper, 49 NY2d at 79). The right to marry someone of the same sex, however, is not “deeply rooted”; it has not even been asserted until relatively recent times. The issue then becomes whether the right to marry must be defined to include a right to same-sex marriage.

Recent Supreme Court decisions show that the definition of a fundamental right for due process purposes may be either too narrow or too broad. In Lawrence v Texas (539 US 558, 566 [2003]), the Supreme Court criticized its own prior decision in Bowers v Hardwick (478 US 186, 190 [1986]) for defining the right at issue as the right of “homosexuals to engage in sodomy.” The Lawrence court plainly thought the right should *363have been defined more broadly, as a right to privacy in intimate relationships. On the other hand, in Washington v Glucksberg (521 US at 722, 723), the Court criticized a lower federal court for defining the right at issue too broadly as a “right to die”; the right at issue in Glucksberg, the Court said, was really the “right to commit suicide” and to have assistance in doing so.

The difference between Lawrence and Glucksberg is that in Glucksberg the relatively narrow definition of the right at issue was based on rational line-drawing. In Lawrence, by contrast, the court found the distinction between homosexual sodomy and intimate relations generally to be essentially arbitrary. Here, there are, as we have explained, rational grounds for limiting the definition of marriage to opposite-sex couples. This case is therefore, in the relevant way, like Glucksberg and not at all like Lawrence. Plaintiffs here do not, as the petitioners in Lawrence did, seek protection against state intrusion on intimate, private activity. They seek from the courts access to a state-conferred benefit that the Legislature has rationally limited to opposite-sex couples. We conclude that, by defining marriage as it has, the New York Legislature has not restricted the exercise of a fundamental right (see also concurring op of Judge Graffeo at 368-374).

Where no fundamental right is at issue, legislation is valid under the Due Process Clause if it is rationally related to legitimate government interests (Glucksberg, 521 US at 728; Hope, 83 NY2d at 577). Again, our earlier discussion answers this question. Protecting the welfare of children is a legitimate governmental interest, and we have shown above that there is a rational relationship between that interest and the limitation of marriage to opposite-sex couples. That limitation therefore does not deprive plaintiffs of due process of law.

B. Equal Protection

Plaintiffs claim that the distinction made by the Domestic Relations Law between opposite-sex and same-sex couples deprives them of the equal protection of the laws. This claim raises, first, the issue of what level of scrutiny should be applied to the legislative classification. The plaintiffs argue for strict scrutiny, on the ground that the legislation affects their fundamental right to marry (see Alevy v Downstate Med. Ctr. of State of N.Y., 39 NY2d 326, 332 [1976]) — a contention we rejected above. Alternatively, plaintiffs argue for so-called intermediate or heightened scrutiny on two grounds. They say that the legisla*364tion discriminates on the basis of sex, a kind of discrimination that has been held to trigger heightened scrutiny (e.g., United States v Virginia, 518 US 515, 532-533 [1996]). They also say that discrimination on the basis of sexual preference should trigger heightened scrutiny, a possibility we left open in Under 21, Catholic Home Bur. for Dependent Children v City of New York (65 NY2d at 364). We reject both of these arguments, and hold that the restriction of marriage to opposite-sex couples is subject only to rational basis scrutiny.

By limiting marriage to opposite-sex couples, New York is not engaging in sex discrimination. The limitation does not put men and women in different classes, and give one class a benefit not given to the other. Women and men are treated alike — they are permitted to marry people of the opposite sex, but not people of their own sex. This is not the kind of sham equality that the Supreme Court confronted in Loving; the statute there, prohibiting black and white people from marrying each other, was in substance anti-black legislation. Plaintiffs do not argue here that the legislation they challenge is designed to subordinate either men to women or women to men as a class.

However, the legislation does confer advantages on the basis of sexual preference. Those who prefer relationships with people of the opposite sex and those who prefer relationships with people of the same sex are not treated alike, since only opposite-sex relationships may gain the status and benefits associated with marriage. This case thus presents the question of what level of scrutiny is to be applied to legislation that classifies people on this basis. We held in Under 21 that “classifications based on sexual orientation” would not be subject to strict scrutiny, but left open the question of “whether some level of ‘heightened scrutiny’ would be applied” in such cases (id. at 364).

We resolve this question in this case on the basis of the Supreme Court’s observation that no more than rational basis scrutiny is generally appropriate “where individuals in the group affected by a law have distinguishing characteristics relevant to interests the State has the authority to implement” (Cleburne v Cleburne Living Center, Inc., 473 US 432, 441 [1985]). Perhaps that principle would lead us to apply heightened scrutiny to sexual preference discrimination in some cases, but not where we review legislation governing marriage and family relationships. A person’s preference for the sort of sexual activity that cannot lead to the birth of children is relevant to the *365State’s interest in fostering relationships that will serve children best. In this area, therefore, we conclude that rational basis scrutiny is appropriate.

Where rational basis scrutiny applies, “[t]he general rule is that 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” (id. at 440). Plaintiffs argue that a classification distinguishing between opposite-sex couples and same-sex couples cannot pass rational basis scrutiny, because if the relevant state interest is the protection of children, the category of those permitted to marry — opposite-sex couples — is both underinclusive and overinclusive. We disagree.

Plaintiffs argue that the category is underinclusive because, as we recognized above, same-sex couples, as well as opposite-sex couples, may have children. That is indeed a reason why the Legislature might rationally choose to extend marriage or its benefits to same-sex couples; but it could also, for the reasons we have explained, rationally make another choice, based on the different characteristics of opposite-sex and same-sex relationships. Our earlier discussion demonstrates that the definition of marriage to include only opposite-sex couples is not irrationally underinclusive.

In arguing that the definition is overinclusive, plaintiffs point out that many opposite-sex couples cannot have or do not want to have children. How can it be rational, they ask, to permit these couples, but not same-sex couples, to marry? The question is not a difficult one to answer. While same-sex couples and opposite-sex couples are easily distinguished, fimiting marriage to opposite-sex couples likely to have children would require grossly intrusive inquiries, and arbitrary and unreliable line-drawing. A legislature that regarded marriage primarily or solely as an institution for the benefit of children could rationally find that an attempt to exclude childless opposite-sex couples from the institution would be a very bad idea.

Rational basis scrutiny is highly indulgent towards the State’s classifications (see Heller v Doe, 509 US 312, 320-321 [1993]). Indeed, it is “a paradigm of judicial restraint” (Affronti v Crosson, 95 NY2d 713, 719 [2001], cert denied sub nom. Affronti v Lippman, 534 US 826 [2001]). We conclude that permitting marriage by all opposite-sex couples does not create an irrationally overnarrow or overbroad classification. The distinction between opposite-sex and same-sex couples enacted by the Legislature does not violate the Equal Protection Clause.

*366V

We hold, in sum, that the Domestic Relations Law’s limitation of marriage to opposite-sex couples is not unconstitutional. We emphasize once again that we are deciding only this constitutional question. It is not for us to say whether same-sex marriage is right or wrong. We have presented some (though not all) of the arguments against same-sex marriage because our duty to defer to the Legislature requires us to do so. We do not imply that there are no persuasive arguments on the other side — and we know, of course, that there are very powerful emotions on both sides of the question.

The dissenters assert confidently that “future generations” will agree with their view of this case (dissenting op at 396). We do not predict what people will think generations from now, but we believe the present generation should have a chance to decide the issue through its elected representatives. We therefore express our hope that the participants in the controversy over same-sex marriage will address their arguments to the Legislature; that the Legislature will listen and decide as wisely as it can; and that those unhappy with the result — as many undoubtedly will be — will respect it as people in a democratic state should respect choices democratically made.

Accordingly, the orders of the Appellate Division in each case should be affirmed without costs.

Graffeo, J.

(concurring). We are asked by the 44 same-sex couples who commenced these four cases to declare that the denial of marriage licenses to same-sex couples violates the Due Process and Equal Protection clauses of the New York Constitution. Plaintiffs and petitioners (collectively referred to as plaintiffs) are representative of many homosexual couples living in committed relationships in our state, some of whom are raising children. They seek the societal recognition and legal and financial benefits accorded by the State to legally married couples. Respondents are the State of New York, the State Department of Health and local officials from the cities of New York, Albany and Ithaca who are involved either in overseeing the New York marriage licensing process or issuing marriage licenses.

Plaintiffs assert that the restriction of marriage to opposite-sex couples impedes the fundamental right to marry and amounts to gender or sexual orientation discrimination that does not withstand any level of constitutional analysis, whether *367strict scrutiny, intermediate scrutiny or rational basis review. Because the determination of the proper level of constitutional review is crucial to the judicial resolution of the issues in this case, I write separately to elaborate on the standard of review that should be applied under the precedent of this Court and the United States Supreme Court. I conclude that rational basis analysis is appropriate and, applying this standard, I concur in the result reached by the plurality that an affirmance is warranted in each of these cases.

This Court has long recognized that “[fjrom time immemorial the State has exercised the fullest control over the marriage relation,” going so far as to observe that “[tihere are, in effect, three parties to every marriage, the man, the woman and the State” (Fearon v Treanor, 272 NY 268, 272 [1936], appeal dismissed 301 US 667 [1937]). The historical conception of marriage as a union between a man and a woman is reflected in the civil institution of marriage adopted by the New York Legislature. The cases before us present no occasion for this Court to debate whether the State Legislature should, as a matter of social welfare or sound public policy, extend marriage to same-sex couples. Our role is limited to assessing whether the current statutory scheme offends the Due Process or Equal Protection clauses of the New York Constitution. Because it does not, we must affirm. Absent a constitutional violation, we may not disturb duly enacted statutes to, in effect, substitute another policy preference for that of the Legislature.

The Statutory Scheme:

As a preliminary matter, although plaintiffs have abandoned the argument (raised in Supreme Court in both Kane and Seymour) that the Domestic Relations Law already authorizes same-sex marriage because it does not explicitly define marriage as a union between one man and one woman, several amici continue to suggest that this Court can avoid a constitutional analysis by simply interpreting the statutory scheme to allow same-sex marriage. Our role when construing a statute is to ascertain and implement the will of the Legislature unless we are prevented from doing so by constitutional infirmity. It would be inappropriate for us to interpret the Domestic Relations Law in a manner that virtually all concede would not comport with legislative intent.

There is no basis to conclude that, when the Legislature adopted the Domestic Relations Law more than a century ago, it contemplated the possibility of same-sex marriage, *368much less intended to authorize it. In fact, the Domestic Relations Law contains many references to married persons that demonstrate that the Legislature viewed marriage as a union between one woman and one man — as seen by references to the parties to a marriage as the “bride” and “groom” (Domestic Relations Law § 15 [1] [a]) and “wife” and “husband” (Domestic Relations Law §§ 6, 12, 221, 248; see also CPLR 4502 [b]). Notably, high courts of other states with statutory schemes comparable to New York’s have interpreted the pertinent statutes as not authorizing same-sex marriage (see Goodridge v Department of Pub. Health, 440 Mass 309, 798 NE2d 941 [2003]; Baker v Nelson, 291 Minn 310, 191 NW2d 185 [1971], appeal dismissed 409 US 810 [1972]). And several of our prior cases alluded to the fact that the Domestic Relations Law precludes same-sex couples from marrying (Levin v Yeshiva Univ., 96 NY2d 484, 494 [2001]; Braschi v Stahl Assoc. Co., 74 NY2d 201, 210 [1989]). Because the Domestic Relations Law does not authorize marriage between persons of the same sex, this Court must address plaintiffs’ constitutional challenges to the validity of the marriage scheme, which are at the heart of this litigation.

Due Process:

Plaintiffs argue that the Domestic Relations Law violates article I, § 6 of the New York Constitution, which provides that “[n]o person shall be deprived of life, liberty or property without due process of law.” Their substantive due process challenge is predicated on the assertion that the New York Constitution precludes the State from defining marriage as a union between one man and one woman because the right to privacy derived therein grants each individual the unqualified right to select and marry the person of his or her choice. If the Due Process Clause encompasses this right, and if it is one of the bundle of rights deemed “fundamental” as plaintiffs contend, the Domestic Relations Law would be subjected to the most demanding form of constitutional review, with the State having the burden to prove that it is narrowly tailored to serve compelling state interests.

But it is an inescapable fact that New York due process cases and the relevant federal case law cited therein do not support plaintiffs’ argument. While many US Supreme Court decisions recognize marriage as a fundamental right protected under the Due Process Clause, all of these cases understood the marriage *369right as involving a union of one woman and one man (see e.g. Turner v Safley, 482 US 78 [1987]; Zablocki v Redhail, 434 US 374 [1978]; Griswold v Connecticut, 381 US 479 [1965]; Skinner v Oklahoma ex rel. Williamson, 316 US 535 [1942]). Whether interpreting New York’s Due Process Clause or its federal counterpart (which is textually identical), when this Court has addressed the fundamental right to marry, it has relied on federal precedent and similarly used the word “marriage” in its traditional sense. For example, in Cooper v Morin, we grounded the right of pretrial detainees to have contact visits with family on the “fundamental right to marriage and family life . . . and to bear and rear children” (49 NY2d 69, 80 [1979], cert denied sub nom. Lombard v Cooper, 446 US 984 [1980]), citing US Supreme Court cases highlighting the link between marriage and procreation. As the Third Department aptly noted in Samuels, to ignore the meaning ascribed to the right to marry in these cases and substitute another meaning in its place is to redefine the right in question and to tear the resulting new right away from the very roots that caused the US Supreme Court and this Court to recognize marriage as a fundamental right in the first place.

Nor has this Court recognized a due process right to privacy distinct from that articulated by the US Supreme Court. Although our Court has interpreted the New York Due Process Clause more broadly than its federal counterpart on a few occasions, all of those cases involved the rights of criminal defendants, prisoners or pretrial detainees, or other confined individuals and implicated classic liberty concerns beyond the right to privacy. Most recently, in People v LaValle (3 NY3d 88 [2004]), the Court concluded that the anticipatory deadlock charge in the Death Penalty Act violated New York’s Due Process Clause, even though it may have been upheld under the United States Constitution. Likewise, in Cooper (49 NY2d 69 [1979]), we held that the New York Due Process Clause protected the right of pretrial detainees in a county jail to have nonconjugal contact visits with family members, even though no such right had been deemed protected under the federal Due Process Clause. Even then, our analysis did not turn on recognition of broader family privacy rights than those articulated by the Supreme Court. Rather, the analysis focused on rejection of the rational basis test that the Supreme Court then applied to *370assess jail regulations,1 with this Court instead adopting a test that “balanc[ed] . . . the harm to the individual resulting from the condition imposed against the benefit sought by the government through its enforcement” (id. at 79).

Most of our Due Process Clause decisions in the right to privacy realm have cited federal authority interchangeably with New York precedent, making no distinction between New York’s constitutional provision and the federal Due Process Clause (see e.g. Hope v Perales, 83 NY2d 563, 575 [1994]; Matter of Raquel Marie X., 76 NY2d 387 [1990], cert denied sub nom. Robert C. v Miguel T., 498 US 984 [1990]; Matter of Doe v Coughlin, 71 NY2d 48 [1987], cert denied 488 US 879 [1988]; Rivers v Katz, 67 NY2d 485 [1986]). Our Court has not recognized a fundamental right to marry that departs in any respect from the right defined by the US Supreme Court in cases like Skinner which acknowledged that marriage is “fundamental to the very existence and survival of the [human] race” because it is the primary institution supporting procreation and child-rearing (316 US at 541; see also Zablocki, 434 US 374; Griswold, 381 US 479). The binary nature of marriage — its inclusion of one woman and one man — reflects the biological fact that human procreation cannot be accomplished without the genetic contribution of both a male and a female. Marriage creates a supportive environment for procreation to occur and the resulting offspring to be nurtured. Although plaintiffs suggest that the connection between procreation and marriage has become anachronistic because of scientific advances in assisted reproduction technology, the fact remains that the vast majority of children are conceived naturally through sexual contact between a woman and a man.

Plaintiffs’ reliance on Loving v Virginia (388 US 1 [1967]) for the proposition that the US Supreme Court has established a fundamental “right to marry the spouse of one’s choice” outside the male/female construct is misplaced. In Loving, an interracial couple argued that Virginia’s antimiscegenation statute, which precluded “any white person in this State to marry any save a white person, or a person with no other admixture of *371blood than white and American Indian” (id. at 5 n 4), violated the federal Due Process and Equal Protection clauses. The statute made intermarriage in violation of its terms a felony carrying a potential jail sentence of one to five years. The Lovings — a white man and a black woman — had married in violation of the law and been convicted, prompting them to challenge the validity of the Virginia law.

The Supreme Court struck the statute on both equal protection and due process grounds, but the focus of the analysis was on the Equal Protection Clause. Noting that “[t]he clear and central purpose of the Fourteenth Amendment was to eliminate all official state sources of invidious racial discrimination in the States,” the Court applied strict scrutiny review to the racial classification, finding “no legitimate overriding purpose independent of invidious racial discrimination which justifies this classification” (id. at 10, 11). It made clear “that restricting the freedom to marry solely because of racial classifications violates the central meaning of the Equal Protection Clause” (id. at 12). There is no question that the Court viewed this antimiscegenation statute as an affront to the very purpose for the adoption of the Fourteenth Amendment — to combat invidious racial discrimination.

In its brief due process analysis, the Supreme Court reiterated that marriage is a right “fundamental to our very existence and survival” (id., citing Skinner, 316 US at 541) — a clear reference to the link between marriage and procreation. It reasoned: “To deny this fundamental freedom on so unsupportable a basis as the racial classifications embodied in these statutes ... is surely to deprive all the State’s citizens of liberty without due process of law” (id.). Although the Court characterized the right to marry as a “choice,” it did not articulate the broad “right to marry the spouse of one’s choice” suggested by plaintiffs here. Rather, the Court observed that “[t]he Fourteenth Amendment requires that the freedom of choice to marry not be restricted by invidious racial discriminations” (id. [emphasis added]).2 Needless to say, a statutory scheme that burdens a fundamental right by making conduct criminal based on the race of the individual who engages in it is inimical to the *372values embodied in the state and federal Due Process clauses. Far from recognizing a right to marry extending beyond the one woman and one man union,3 it is evident from the Loving decision that the Supreme Court viewed marriage as fundamental precisely because of its relationship to human procreation.4

Nor does the Supreme Court’s recent federal due process analysis in Lawrence v Texas (539 US 558 [2003]) support defining the fundamental marriage right in the manner urged by *373plaintiffs. In Lawrence, the Court overruled its prior decision in Bowers v Hardwick (478 US 186 [1986]) and struck as unconstitutional a Texas statute that criminalized consensual sodomy between adult individuals of the same sex. The holding in Lawrence is consistent with our Court’s decision in People v Onofre (51 NY2d 476 [1980], cert denied 451 US 987 [1981]), which invalidated under a federal due process analysis a New York Penal Law provision that criminalized consensual sodomy between nonmarried persons.

In Lawrence the Supreme Court did not create any new fundamental rights, nor did it employ a strict scrutiny analysis. It acknowledged that laws that criminalize sexual conduct between homosexuals

“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” (539 US at 567).

Criticizing the historical analysis in Bowers, it noted that, even though sodomy as well as other nonprocreative sexual activity had been proscribed, criminal statutes “directed at homosexual conduct as a distinct matter” {id. at 568) were of recent vintage, having developed in the last third of the 20th century, and therefore did not possess “ancient roots” (id. at 570).

Consistent with our analysis in Onofre, the Lawrence court held “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” (id. at 567) because “liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex” (id. at 572). It reasoned that “moral disapproval” — the only justification Texas proffered for its law — is never an adequate basis for a criminal statute, a conclusion similar to this Court’s observation in Onofre that “it is not the function of the Penal Law in our governmental policy to provide either a medium for the articulation or the apparatus for the intended enforcement of moral or theological values” (51 NY2d at 488 n 3). Thus, in striking the sodomy law, the Supreme Court found that “[t]he Texas statute furthers no legitimate state interest which can justify its intrusion into the personal and private life of the individual” (Lawrence, 539 US at 578).

*374The right affirmed by the Supreme Court in Lawrence is not comparable to the new right to marry plaintiffs assert here, nor is the Texas statute criminalizing homosexual sodomy analogous to the marriage statutes under review. The Domestic Relations Law is not a penal provision and New York has not attempted to regulate plaintiffs’ private sexual conduct or disturb the sanctity of their homes. And, in contrast to the Texas statute, New York’s marriage laws are part of a longstanding tradition with roots dating back long before the adoption of our State Constitution.

New York’s Due Process Clause simply does not encompass a fundamental right to marry the spouse of one’s choice outside the one woman/one man construct. Strict scrutiny review of the Domestic Relations Law is therefore not warranted and, insofar as due process analysis is concerned, the statutory scheme must be upheld unless plaintiffs prove that it is not rationally related to any legitimate state interest.

Equal Protection:

Plaintiffs contend that, even if strict scrutiny analysis is not appropriate under the Due Process Clause, a heightened standard of review is nonetheless mandated under the Equal Protection Clause because New York’s marriage laws create gender and sexual orientation classifications that require a more rigorous level of analysis than rational basis review.

The Equal Protection Clause, added to the New York Constitution in 1938, provides:

“No person shall be denied the equal protection of the laws of this state or any subdivision thereof. No person shall, because of race, color, creed or religion, be subjected to any discrimination in his or her civil rights by any other person or by any firm, corporation, or institution, or by the state or any agency or subdivision of the state” (NY Const, art I, § 11).

Soon after the adoption of this provision, this Court recognized that it was modeled after its federal counterpart and “embodies” the federal equal protection command (Dorsey v Stuyvesant Town Corp., 299 NY 512, 530 [1949], cert denied 339 US 981 [1950]; see also, Under 21, Catholic Home Bur. for Dependent Children v City of New York, 65 NY2d 344, 360 n 6 [1985] [“the State constitutional equal protection clause ... is no broader in coverage than the Federal provision”]). Accordingly, this Court has consistently cited federal cases and applied federal *375analysis to resolve equal protection claims brought under the federal and state constitutions (see e.g. Matter of Aliessa v Novello, 96 NY2d 418 [2001]; People v Liberta, 64 NY2d 152 [1984], cert denied 471 US 1020 [1985]).

The Equal Protection Clause “is essentially a direction that all persons similarly situated should be treated alike” (Cleburne v Cleburne Living Center, Inc., 473 US 432, 439 [1985]). Both the US Supreme Court and this Court have applied three levels of review to legislative classifications. “[W]hen a statute classifies by race, alienage, or national origin” (id. at 440), or when it burdens a fundamental right protected under the Due Process Clause, it is subjected to strict scrutiny meaning that it will be sustained only if it is narrowly tailored to serve a compelling state interest (see Golden v Clark, 76 NY2d 618, 623 [1990]). Classifications based on gender or illegitimacy are reviewed under an intermediate level of scrutiny — meaning they will be sustained if “substantially related to the achievement of an important governmental objective” (Liberta, 64 NY2d at 168; Clark v Jeter, 486 US 456 [1988]). Neither the Supreme Court nor this Court has recognized any other classifications as triggering heightened scrutiny and, therefore, all other statutory distinctions have been sustained if rationally related to a legitimate government interest (see e.g. Golden, 76 NY2d 618).

Plaintiffs argue that the Domestic Relations Law creates a classification based on gender that requires intermediate scrutiny because a woman cannot marry another woman due to her gender and a man cannot marry another man due to his gender. Respondents counter that the marriage laws are neutral insofar as gender is concerned because they treat all males and females equally — neither gender can marry a person of the same sex and both can marry persons of the opposite sex.

Respondents’ interpretation more closely comports with the analytical framework for gender discrimination applied by this Court and the Supreme Court. The precedent establishes that gender discrimination occurs when men and women are not treated equally and one gender is benefitted or burdened as opposed to the other. For example, in Liberta (64 NY2d 152), we held that the Penal Law’s restriction of the crime of forcible rape to male offenders constituted gender discrimination and the restriction was struck on the basis that it failed to meet the intermediate scrutiny standard. Men and women were not treated equally because only men could be convicted of forcible rape; women who engaged in precisely the same conduct could *376not be charged or convicted of the same offense. Similarly, in Mississippi Univ. for Women v Hogan (458 US 718 [1982]), the Supreme Court found that a publically-funded state university that refused to allow men admission to its nursing program had engaged in gender discrimination. The university improperly privileged female students by allowing them a benefit not available to similarly-situated male applicants. Likewise, in J. E. B. v Alabama ex rel. T. B. (511 US 127 [1994]), a prosecutor was determined to have engaged in gender discrimination when he exercised 9 of his 10 peremptory challenges to strike males from the venire panel resulting in an all-female jury. There, the prosecutor did not apply jury selection criteria equally among males and females — he used almost all of his challenges to exclude men from the jury.

Plaintiffs cite Loving for the proposition that a statute can discriminate even if it treats both classes identically. This misconstrues the Loving analysis because the antimiscegenation statute did not treat blacks and whites identically — it restricted who whites could marry (but did not restrict intermarriage between non-whites) for the purpose of promoting white supremacy. Virginia’s antimiscegenation statute was the quintessential example of invidious racial discrimination as it was intended to advantage one race and disadvantage all others, which is why the Supreme Court applied strict scrutiny and struck it down as violating the core interest of the Equal Protection Clause.

In contrast, neither men nor women are disproportionately disadvantaged or burdened by the fact that New York’s Domestic Relations Law allows only opposite-sex couples to marry— both genders are treated precisely the same way. As such, there is no gender classification triggering intermediate scrutiny.

Nor does the statutory scheme create a classification based on sexual orientation. In this respect, the Domestic Relations Law is facially neutral: individuals who seek marriage licenses are not queried concerning their sexual orientation and are not precluded from marrying if they are not heterosexual. Regardless of sexual orientation, any person can marry a person of the opposite sex. Certainly, the marriage laws create a classification that distinguishes between opposite-sex and same-sex couples and this has a disparate impact on gays and lesbians. However, a claim that a facially-neutral statute enacted without an invidious discriminatory intent has a disparate impact on a class (even a suspect class, such as one defined by race) is insufficient *377to establish an equal protection violation5 (see Campaign for Fiscal Equity v State of New York, 86 NY2d 307, 321 [1995]; People v New York City Tr. Auth., 59 NY2d 343, 350 [1983]; Washington v Davis, 426 US 229, 240 [1976]). Plaintiffs concede that the Domestic Relations Law was not enacted with an invidiously discriminatory intent — the Legislature did not craft the marriage laws for the purpose of disadvantaging gays and lesbians (cf. Romer v Evans, 517 US 620 [1996]). Hence, there is no basis to address plaintiffs’ argument that classifications based on sexual orientation should be subjected to intermediate scrutiny.

Rational Basis Review:

Thus, under both the Due Process and Equal Protection clauses, these cases turn on whether the Legislature’s decision to confine the institution of marriage to couples composed of one woman and one man is rationally related to any legitimate state interest. In Affronti v Crosson (95 NY2d 713, 719 [2001], cert denied sub nom. Affronti v Lippman, 534 US 826 [2001]), we explained that

“[t]he rational basis standard of review is a paradigm of judicial restraint. On rational basis review, a statute will be upheld unless the disparate treatment is so unrelated to the achievement of any combination of legitimate purposes that it is irrational. Since the challenged statute is presumed to be valid, the burden is on the one attacking the legislative arrangement to negative every conceivable basis which might support it . . . whether or not the basis has a foundation in the record. Thus, those challenging the legislative judgment must convince the court that the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decisionmaker.” (Internal quotation marks, citations, brackets and emphasis omitted.)

Especially in the realm of social or economic legislation, “the Equal Protection Clause allows the States wide latitude . . . and the Constitution presumes that even improvident decisions will eventually be rectified by the democratic processes” *378(Cleburne, 473 US at 440; see generally Lovelace v Gross, 80 NY2d 419, 427 [1992]).

In these cases, respondents articulate a number of interests that they claim are legitimate and are advanced by the current definition of marriage. Given the extremely deferential standard of review, plaintiffs cannot prevail unless they establish that no conceivable legitimate interest is served by the statutory scheme. This means that if this Court finds a rational connection between the classification and any single governmental concern, the marriage laws survive review under both the Due Process and Equal Protection clauses.

As set forth in the plurality opinion, plaintiffs have failed to negate respondents’ explanation that the current definition of marriage is rationally related to the State’s legitimate interest in channeling opposite-sex relationships into marriage because of the natural propensity of sexual contact between opposite-sex couples to result in pregnancy and childbirth. Of course, marriage can and does serve individual interests that extend well beyond creating an environment conducive to procreation and child-rearing, such as companionship and emotional fulfilment. But here we are concerned with the State’s interest in promoting the institution of marriage.

As Justice Robert Cordy pointed out in his dissent in Goodridge v Department of Pub. Health (440 Mass at 381-382, 798 NE2d at 995 [Cordy, J., dissenting]):

“Civil marriage is the institutional mechanism by which societies have sanctioned and recognized particular family structures, and the institution of marriage has existed as one of the fundamental organizing principles of human society. . . . Paramount among its many important functions, the institution of marriage has systematically provided for the regulation of heterosexual behavior, brought order to the resulting procreation, and ensured a stable family structure in which children will be reared, educated, and socialized. . . . [A]n orderly society requires some mechanism for coping with the fact that sexual intercourse [between a man and a woman] commonly results in pregnancy and childbirth. The institution of marriage is that mechanism.”

Since marriage was instituted to address the fact that sexual contact between a man and a woman naturally can result in *379pregnancy and childbirth, the Legislature’s decision to focus on opposite-sex couples is understandable. It is not irrational for the Legislature to provide an incentive for opposite-sex couples — for whom children may be conceived from casual, even momentary intimate relationships — to marry, create a family environment, and support their children. Although many same-sex couples share these family objectives and are competently raising children in a stable environment, they are simply not similarly situated to opposite-sex couples in this regard given the intrinsic differences in the assisted reproduction or adoption processes that most homosexual couples rely on to have children.

As respondents concede, the marriage classification is imperfect and could be viewed in some respects as overinclusive or underinclusive since not all opposite-sex couples procreate, opposite-sex couples who cannot procreate may marry, and opposite-sex partners can and do procreate outside of marriage. It is also true that children being raised in same-sex households would derive economic and social benefits if their parents could marry. But under rational basis review, the classification need not be perfectly precise or narrowly tailored — all that is required is a reasonable connection between the classification and the interest at issue. In light of the history and purpose of the institution of marriage, the marriage classification in the Domestic Relations Law meets that test.

The Legislature has granted the benefits (and responsibilities) of marriage to the class — opposite-sex couples — that it concluded most required the privileges and burdens the institution entails due to inherent procreative capabilities. This type of determination is a central legislative function and lawmakers are afforded leeway in fulfilling this function, especially with respect to economic and social legislation where issues are often addressed incrementally (see FCC v Beach Communications, Inc., 508 US 307, 315-316 [1993]). It may well be that the time has come for the Legislature to address the needs of same-sex couples and their families, and to consider granting these individuals additional benefits through marriage or whatever status the Legislature deems appropriate. Because the New York Constitution does not compel such a revision of the Domestic Relations Law, the decision whether or not to do so rests with our elected representatives.

*380Chief Judge Kaye

(dissenting). Plaintiffs (including petitioners) are 44 same-sex couples who wish to marry. They include a doctor, a police officer, a public school teacher, a nurse, an artist and a state legislator. Ranging in age from under 30 to 68, plaintiffs reflect a diversity of races, religions and ethnicities. They come from upstate and down, from rural, urban and suburban settings. Many have been together in committed relationships for decades, and many are raising children — from toddlers to teenagers. Many are active in their communities, serving on their local school board, for example, or their cooperative apartment building board. In short, plaintiffs represent a cross-section of New Yorkers who want only to live full lives, raise their children, better their communities and be good neighbors.

For most of us, leading a full life includes establishing a family. Indeed, most New Yorkers can look back on, or forward to, their wedding as among the most significant events of their lives. They, like plaintiffs, grew up hoping to find that one person with whom they would share their future, eager to express their mutual lifetime pledge through civil marriage. Solely because of their sexual orientation, however — that is, because of who they love — plaintiffs are denied the rights and responsibilities of civil marriage. This State has a proud tradition of affording equal rights to all New Yorkers. Sadly, the Court today retreats from that proud tradition.

I. Due Process

Under both the state and federal constitutions, the right to due process of law protects certain fundamental liberty interests, including the right to marry. Central to the right to marry is the right to marry the person of one’s choice (see e.g. Crosby v State of N.Y., Workers’ Compensation Bd., 57 NY2d 305, 312 [1982] [“clearly falling within (the right of privacy) are matters relating to the decision of whom one will marry”]; People v Shepard, 50 NY2d 640, 644 [1980] [“the government has been prevented from interfering with an individual’s decision about whom to marry”]). The deprivation of a fundamental right is subject to strict scrutiny and requires that the infringement be narrowly tailored to achieve a compelling state interest (see e.g. Carey v Population Services Int’l, 431 US 678, 686 [1977]).

Fundamental rights are those “which are, objectively, deeply rooted in this Nation’s history and tradition . . . and implicit in the concept of ordered liberty, such that neither liberty nor justice would exist if they were sacrificed” (Washington v *381Glucksberg, 521 US 702, 720-721 [1997] [internal quotation marks and citations omitted]). Again and again, the Supreme Court and this Court have made clear that the right to marry is fundamental (see e.g. Loving v Virginia, 388 US 1 [1967]; Zablocki v Redhail, 434 US 374 [1978]; Turner v Safley, 482 US 78 [1987]; Matter of Doe v Coughlin, 71 NY2d 48, 52 [1987]; Cooper v Morin, 49 NY2d 69, 80 [1979]; Levin v Yeshiva Univ., 96 NY2d 484, 500 [2001] [G.B. Smith, J., concurring] [“marriage is a fundamental constitutional right”]).

The Court concludes, however, that same-sex marriage is not deeply rooted in tradition, and thus cannot implicate any fundamental liberty. But fundamental rights, once recognized, cannot be denied to particular groups on the ground that these groups have historically been denied those rights. Indeed, in recasting plaintiffs’ invocation of their fundamental right to marry as a request for recognition of a “new” right to same-sex marriage, the Court misapprehends the nature of the liberty interest at stake. In Lawrence v Texas (539 US 558 [2003]), the Supreme Court warned against such error.

Lawrence overruled Bowers v Hardwick (478 US 186 [1986]), which had upheld a Georgia statute criminalizing sodomy. In so doing, the Lawrence court criticized Bowers for framing the issue presented too narrowly. Declaring that “Bowers was not correct when it was decided, and it is not correct today” (539 US at 578), Lawrence explained that Bowers purported to analyze — erroneously—whether the Constitution conferred a “fundamental right upon homosexuals to engage in sodomy” (539 US at 566 [citation omitted]). This was, however, the wrong question. The fundamental right at issue, properly framed, was the right to engage in private consensual sexual conduct — a right that applied to both homosexuals and heterosexuals alike. In narrowing the claimed liberty interest to embody the very exclusion being challenged, Bowers “disclose[d] the Court’s own failure to appreciate the extent of the liberty at stake” (Lawrence, 539 US at 567).

The same failure is evident here. An asserted liberty interest is not to be characterized so narrowly as to make inevitable the conclusion that the claimed right could not be fundamental because historically it has been denied to those who now seek to exercise it (see Planned Parenthood of Southeastern Pa. v Casey, 505 US 833, 847 [1992] [it is “tempting ... to suppose that the Due Process Clause protects only those practices, defined at the most specific level, that were protected against government *382interference by other rules of law when the Fourteenth Amendment was ratified. . . . But such a view would be inconsistent with our law”]).

Notably, the result in Lawrence was not affected by the fact, acknowledged by the Court, that there had been no long history of tolerance for homosexuality. Rather, in holding that “[pier-sons in a homosexual relationship may seek autonomy for the[ ] purpose[ of making intimate and personal choices], just as heterosexual persons do” (539 US at 574), Lawrence rejected the notion that fundamental rights it had already identified could be restricted based on traditional assumptions about who should be permitted their protection. As the Court noted, “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” (Lawrence, 539 US at 579; see also id. at 572 [“(history and tradition are the starting point but not in all cases the ending point of the substantive due process inquiry” (internal quotation marks and citation omitted)]; Cleburne v Cleburne Living Center, Inc., 473 US 432, 466 [1985] [Marshall, J., concurring in the judgment in part and dissenting in part] [“what once was a ‘natural’ and ‘self-evident’ ordering later comes to be seen as an artificial and invidious constraint on human potential and freedom”]).

Simply put, fundamental rights are fundamental rights. They are not defined in terms of who is entitled to exercise them.

Instead, the Supreme Court has repeatedly held that the fundamental right to marry must be afforded even to those who have previously been excluded from its scope — that is, to those whose exclusion from the right was “deeply rooted.”1 Well into the twentieth century, the sheer weight of precedent accepting the constitutionality of bans on interracial marriage was deemed sufficient justification in and of itself to perpetuate these discriminatory laws (see e.g. Jones v Lorenzen, 441 P2d 986, 989 *383[Okla 1965] [upholding antimiscegenation law since the “great weight of authority holds such statutes constitutional”]) — much as defendants now contend that same-sex couples should be prohibited from marrying because historically they always have been.

Just 10 years before Loving declared unconstitutional state laws banning marriage between persons of different races, 96% of Americans were opposed to interracial marriage (see brief of NAACP Legal Defense and Educational Fund, Inc., as amicus curiae in support of plaintiffs, at 5). Sadly, many of the arguments then raised in support of the antimiscegenation laws were identical to those made today in opposition to same-sex marriage (see e.g. Kinney v Commonwealth, 71 Va [30 Gratt] 858, 869 [1878] [marriage between the races is “unnatural” and a violation of God’s will]; Pace v State, 69 Ala 231, 232 [1881] [“amalgamation” of the races would produce a “degraded civilization”]; see also Lonas v State, 50 Tenn [3 Heisk] 287, 310 [1871] [“(t)he laws of civilization demand that the races be kept apart”]).

To those who appealed to history as a basis for prohibiting interracial marriage, it was simply inconceivable that the right of interracial couples to marry could be deemed “fundamental.” Incredible as it may seem today, during the lifetime of every Judge on this Court, interracial marriage was forbidden in at least a third of American jurisdictions. In 1948, New York was one of only 18 states in the nation that did not have such a ban. By 1967, when Loving was decided, 16 states still outlawed marriages between persons of different races. Nevertheless, even though it was the ban on interracial marriage — not interracial marriage itself — that had a long and shameful national tradition, the Supreme Court determined that interracial couples could not be deprived of their fundamental right to marry.

Unconstitutional infringements on the right to marry are not limited to impermissible racial restrictions. Inasmuch as the fundamental right to marry is shared by “all the State’s citizens” (Loving, 388 US at 12), the State may not, for example, require individuals with child support obligations to obtain court approval before getting married (see Zablocki, 434 US 374 [1978]). Calling Loving the “leading decision of this Court on the right to marry,” Justice Marshall made clear in Zablocki that Loving

“could have rested solely on the ground that the *384statutes discriminated on the basis of race in violation of the Equal Protection Clause. But the Court went on to hold that laws arbitrarily deprived the couple of a fundamental liberty protected by the Due Process Clause, the freedom to marry. . . .

“Although Loving arose in the context of racial discrimination, prior and subsequent decisions of this Court confirm that the right to marry is of fundamental importance for all individuals” (434 US at 383-384 [citation omitted]).

Similarly, in Turner (482 US 78 [1987]), the Supreme Court determined that the right to marry was so fundamental that it could not be denied to prison inmates (see also Boddie v Connecticut, 401 US 371 [1971] [state requirement that indigent individuals pay court fees to obtain divorce unconstitutionally burdened fundamental right to marry]).

Under our Constitution, discriminatory views about proper marriage partners can no more prevent same-sex couples from marrying than they could different-race couples. Nor can “deeply rooted” prejudices uphold the infringement of a fundamental right (see People v Onofre, 51 NY2d 476, 490 [1980] [“disapproval by a majority of the populace . . . may not substitute for the required demonstration of a valid basis for intrusion by the State in an area of important personal decision”]). For these reasons, the NAACP Legal Defense and Educational Fund, as amicus, contends that

“[although the historical experiences in this country of African Americans, on the one hand, and gay men and lesbians, on the other, are in many important ways quite different, the legal questions raised here and in Loving are analogous. The state law at issue here, like the law struck down in Loving, restricts an individual’s right to marry the person of his or her choice. We respectfully submit that the decisions below must be reversed if this Court follows the reasoning of the United States Supreme Court’s decision in Loving” (brief of NAACP Legal Defense and Educational Fund, Inc., as amicus curiae in support of plaintiffs, at 3-4; see also brief of New York County Lawyers’ Association and National Black Justice Coalition, as amici curiae in support of plaintiffs [detailing history of antimiscegenation laws and public attitudes toward interracial marriage]).

*385It is no answer that same-sex couples can be excluded from marriage because “marriage,” by definition, does not include them. In the end, “an argument that marriage is heterosexual because it ‘just is’ amounts to circular reasoning” (Halpern v Attorney Gen. of Can., 65 OR3d 161, 172 OAC 276, 11 71 [2003]). “To define the institution of marriage by the characteristics of those to whom it always has been accessible, in order to justify the exclusion of those to whom it never has been accessible, is conclusory and bypasses the core question we are asked to decide” (Goodridge v Department of Pub. Health, 440 Mass 309, 348, 798 NE2d 941, 972-973 [2003] [Greaney, J., concurring]).

The claim that marriage has always had a single and unalterable meaning is a plain distortion of history. In truth, the common understanding of “marriage” has changed dramatically over the centuries (see brief of Professors of History and Family Law, as amici curiae in support of plaintiffs). Until well into the nineteenth century, for example, marriage was defined by the doctrine of coverture, according to which the wife’s legal identity was merged into that of her husband, whose property she became. A married woman, by definition, could not own property and could not enter into contracts.2 Such was the very “meaning” of marriage. Only since the mid-twentieth century has the institution of marriage come to be understood as a relationship between two equal partners, founded upon shared intimacy and mutual financial and emotional support. Indeed, as amici professors note, “The historical record shows that, through adjudication and legislation, all of New York’s sex-specific rules for marriage have been invalidated save for the one at issue here.”

That restrictions on same-sex marriage are prevalent cannot in itself justify their retention. After all, widespread public opposition to interracial marriage in the years before Loving could not sustain the antimiscegenation laws. “[T]he 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” (Lawrence, 539 US at 577-578 [internal quotation marks and citation omitted]; see also id. at 571 [fundamental right to engage in private consensual sexual conduct extends to homosexuals, notwithstanding that “for centuries there have been powerful voices to condemn homo*386sexual conduct as immoral”]). The long duration of a constitutional wrong cannot justify its perpetuation, no matter how strongly tradition or public sentiment might support it.

II. Equal Protection

By virtue of their being denied entry into civil marriage, plaintiff couples are deprived of a number of statutory benefits and protections extended to married couples under New York law. Unlike married spouses, same-sex partners may be denied hospital visitation of their critically ill life partners. They must spend more of their joint income to obtain equivalent levels of health care coverage. They may, upon the death of their partners, find themselves at risk of losing the family home. The record is replete with examples of the hundreds of ways in which committed same-sex couples and their children are deprived of equal benefits under New York law. Same-sex families are, among other things, denied equal treatment with respect to intestacy, inheritance, tenancy by the entirety, taxes, insurance, health benefits, medical decisionmaking, workers’ compensation, the right to sue for wrongful death and spousal privilege. Each of these statutory inequities, as well as the discriminatory exclusion of same-sex couples from the benefits and protections of civil marriage as a whole, violates their constitutional right to equal protection of the laws.

Correctly framed, the question before us is not whether the marriage statutes properly benefit those they are intended to benefit — any discriminatory classification does that — but whether there exists any legitimate basis for excluding those who are not covered by the law. That the language of the licensing statute does not expressly reference the implicit exclusion of same-sex couples is of no moment {see Domestic Relations Law § 13 [“persons intended to be married” must obtain a marriage license]). The Court has, properly, construed the statutory scheme as prohibiting same-sex marriage. That being so, the statute, in practical effect, becomes identical to — and, for purposes of equal protection analysis, must be analyzed as if it were — one explicitly providing that “civil marriage is hereby established for couples consisting of a man and a woman,” or, synonymously, “marriage between persons of the same sex is prohibited.”

On three independent grounds, this discriminatory classification is subject to heightened scrutiny, a test that defendants concede it cannot pass.

*387A. Heightened Scrutiny

1. Sexual Orientation Discrimination

Homosexuals meet the constitutional definition of a suspect class, that is, a group whose defining characteristic is “so seldom relevant to the achievement of any legitimate state interest that laws grounded in such considerations are deemed to reflect prejudice and antipathy — a view that those in the burdened class are not as worthy or deserving as others” (Cleburne, 473 US at 440). Accordingly, any classification discriminating on the basis of sexual orientation must be narrowly tailored to meet a compelling state interest (see e.g. Alevy v Downstate Med. Ctr. of State of N.Y., 39 NY2d 326, 332 [1976]; Matter of Aliessa v Novello, 96 NY2d 418, 431 [2001]).

“No single talisman can define those groups likely to be the target of classifications offensive to the Fourteenth Amendment and therefore warranting heightened or strict scrutiny” (Cleburne, 473 US at 472 n 24 [Marshall, J., concurring in the judgment in part and dissenting in part]). Rather, such scrutiny is to be applied when analyzing legislative classifications involving groups who “may well be the target of the sort of prejudiced, thoughtless, or stereotyped action that offends principles of equality found in” the Constitution (id. at 472).

Although no single factor is dispositive, the Supreme Court has generally looked to three criteria in determining whether a group subject to legislative classification must be considered “suspect.” First, the Court has considered whether the group has historically been subjected to purposeful discrimination. Homosexuals plainly have been, as the Legislature expressly found when it recently enacted the Sexual Orientation NonDiscrimination Act (SONDA), barring discrimination against homosexuals in employment, housing, public accommodations, education, credit and the exercise of civil rights. Specifically, the Legislature found

“that many residents of this state have encountered prejudice on account of their sexual orientation, and that this prejudice has severely limited or actually prevented access to employment, housing and other basic necessities of life, leading to deprivation and suffering. The legislature further recognizes that this prejudice has fostered a general climate of hostility and distrust, leading in some instances to *388physical violence against those perceived to be homosexual or bisexual” (L 2002, ch 2, § 1; see also brief of Parents, Families & Friends of Lesbians and Gays, Inc., et al., as amici curiae in support of plaintiffs, at 22-49 [detailing history of state-sanctioned discrimination against gays and lesbians]).

Second, the Court has considered whether the trait used to define the class is unrelated to the ability to perform and participate in society. When the State differentiates among its citizens “on the basis of stereotyped characteristics not truly indicative of their abilities” (Massachusetts Bd. of Retirement v Murgia, 427 US 307, 313 [1976]), the legislative classification must be closely scrutinized. Obviously, sexual orientation is irrelevant to one’s ability to perform or contribute.

Third, the Court has taken into account the group’s relative political powerlessness. Defendants contend that classifications based on sexual orientation should not be afforded heightened scrutiny because, they claim, homosexuals are sufficiently able to achieve protection from discrimination through the political process, as evidenced by the Legislature’s passage of SONDA in 2002. SONDA, however, was first introduced in 1971. It failed repeatedly for 31 years, until it was finally enacted just four years ago. Further, during the Senate debate on the Hate Crimes Act of 2000, one Senator noted that “[i]t’s no secret that for years we could have passed a hate-crimes bill if we were willing to take out gay people, if we were willing to take out sexual orientation” (New York State Senate Debate on Senate Bill S 4691-A, June 7, 2000, at 4609 [statement of Senator Schneider-man]; accord id. at 4548-4549 [statement of Senator Connor]). The simple fact is that New York has not enacted anything approaching comprehensive statewide domestic partnership protections for same-sex couples, much less marriage or even civil unions.

In any event, the Supreme Court has never suggested that racial or sexual classifications are not (or are no longer) subject to heightened scrutiny because of the passage of even comprehensive civil rights laws (see Cleburne, 473 US at 467 [Marshall, J., concurring in the judgment in part and dissenting in part]). Indeed, sex discrimination was first held to deserve heightened scrutiny in 1973 — after passage of title VII of the Civil Rights Act of 1964 and the Equal Pay Act of 1963, federal laws prohibiting sex discrimination. Such measures acknowledge — rather *389than mark the end of — a history of purposeful discrimination (see Frontiero v Richardson, 411 US 677, 687-688 [1973] [citing antidiscrimination legislation to support conclusion that classifications based on sex merit heightened scrutiny]).

Nor is plaintiffs’ claim legitimately answered by the argument that the licensing statute does not discriminate on the basis of sexual orientation since it permits homosexuals to marry persons of the opposite sex and forbids heterosexuals to marry persons of the same sex. The purported “right” of gays and lesbians to enter into marriages with different-sex partners to whom they have no innate attraction cannot possibly cure the constitutional violation actually at issue here. “The right to marry is the right of individuals, not of . . . groups” (Perez v Sharp, 32 Cal 2d 711, 716, 198 P2d 17, 20 [1948]). “Human beings are bereft of worth and dignity by a doctrine that would make them as interchangeable as trains” (32 Cal 2d at 725, 198 P2d at 25). Limiting marriage to opposite-sex couples undeniably restricts gays and lesbians from marrying their chosen same-sex partners whom “to [them] may be irreplaceable” (id.) — and thus constitutes discrimination based on sexual orientation.3

2. Sex Discrimination

The exclusion of same-sex couples from civil marriage also discriminates on the basis of sex, which provides a further basis for requiring heightened scrutiny. Classifications based on sex must be substantially related to the achievement of important governmental objectives (see e.g. Craig v Boren, 429 US 190, 197 [1976]), and must have an “exceedingly persuasive justification” (Mississippi Univ. for Women v Hogan, 458 US 718, 724 [1982] [citations omitted]).

Under the Domestic Relations Law, a woman who seeks to marry another woman is prevented from doing so on account of her sex — that is, because she is not a man. If she were, she would be given a marriage license to marry that woman. That *390the statutory scheme applies equally to both sexes does not alter the conclusion that the classification here is based on sex. The “equal application” approach to equal protection analysis was expressly rejected by the Supreme Court in Loving: “[W]e reject the notion that the mere ‘equal application’ of a statute containing [discriminatory] classifications is enough to remove the classifications from the [constitutional] proscription of all invidious . . . discriminations” (388 US at 8). Instead, the Loving court held that “[t]here can be no question but that Virginia’s miscegenation statutes rest solely upon distinctions drawn according to race [where the] statutes proscribe generally accepted conduct if engaged in by members of different races” (id. at 11; see also Johnson v California, 543 US 499, 506 [2005]; McLaughlin v Florida, 379 US 184, 191 [1964]; Anderson v Martin, 375 US 399, 403-404 [1964]; Shelley v Kraemer, 334 US 1, 21-22 [1948]; J. E. B. v Alabama ex rel. T. B., 511 US 127, 141-142 [1994] [government exercise of peremptory challenges on the basis of gender constitutes impermissible sex discrimination even though based on gender stereotyping of both men and women]).

3. Fundamental Right

“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” (Lawrence, 539 US at 575). Because, as already discussed, the legislative classification here infringes on the exercise of the fundamental right to marry, the classification cannot be upheld unless it is necessary to the achievement of a compelling state interest (see Onofre, 51 NY2d at 492 n 6; Alevy, 39 NY2d at 332; Eisenstadt v Baird, 405 US 438, 447 n 7 [1972]). “[C]ritical examination of the state interests advanced in support of the classification is required” (Zablocki, 434 US at 383 [internal quotation marks and citations omitted]). And if “the means selected by the State for achieving” even “legitimate and substantial interests” unnecessarily impinge on the right to marry, the statutory distinction “cannot be sustained” (id. at 388).

B. Rational-Basis Analysis

Although the classification challenged here should be analyzed using heightened scrutiny, it does not satisfy even rational-basis review, which requires that the classification “rationally further *391a legitimate state interest” (Affronti v Crosson, 95 NY2d 713, 718 [2001], cert denied sub nom. Affronti v Lippman, 534 US 826 [2001]). Rational-basis review requires both the existence of a legitimate interest and that the classification rationally advance that interest. Although a number of interests have been proffered in support of the challenged classification at issue, none is rationally furthered by the exclusion of same-sex couples from marriage. Some fail even to meet the threshold test of legitimacy.

Properly analyzed, equal protection requires that it be the legislated distinction that furthers a legitimate state interest, not the discriminatory law itself (see e.g. Cooper, 49 NY2d at 78; Romer v Evans, 517 US 620, 633 [1996]). Were it otherwise, an irrational or invidious exclusion of a particular group would be permitted so long as there was an identifiable group that benefitted from the challenged legislation. In other words, it is not enough that the State have a legitimate interest in recognizing or supporting opposite-sex marriages. The relevant question here is whether there exists a rational basis for excluding same-sex couples from marriage, and, in fact, whether the State’s interests in recognizing or supporting opposite-sex marriages are rationally furthered by the exclusion.

1. Children

Defendants primarily assert an interest in encouraging procreation within marriage. But while encouraging opposite-sex couples to marry before they have children is certainly a legitimate interest of the State, the exclusion of gay men and lesbians from marriage in no way furthers this interest. There are enough marriage licenses to go around for everyone.

Nor does this exclusion rationally further the State’s legitimate interest in encouraging heterosexual married couples to procreate. Plainly, the ability or desire to procreate is not a prerequisite for marriage. The elderly are permitted to marry, and many same-sex couples do indeed have children. Thus, the statutory classification here — which prohibits only same-sex couples, and no one else, from marrying — is so grossly underinclusive and overinclusive as to make the asserted rationale in promoting procreation “impossible to credit” (Romer, 517 US at 635).4 Indeed, even the Lawrence dissenters observed that “encouragement of procreation” could not “possibly” be a justification *392for denying marriage to gay and lesbian couples, “since the sterile and the elderly are allowed to marry” (539 US at 605 [Scalia, J., dissenting]; see also Lapides v Lapides, 254 NY 73, 80 [1930] [“inability to bear children” does not justify án annulment under the Domestic Relations Law]).

Of course, there are many ways in which the government could rationally promote procreation — for example, by giving tax breaks to couples who have children, subsidizing child care for those couples, or mandating generous family leave for parents. Any of these benefits — and many more — might convince people who would not otherwise have children to do so. But no one rationally decides to have children because gays and lesbians are excluded from marriage.

In holding that prison inmates have a fundamental right to marry — even though they cannot procreate — the Supreme Court has made it clear that procreation is not the sine qua non of marriage. “Many important attributes of marriage remain . . . after taking into account the limitations imposed by prison life. . . . [I]nmate marriages, like others, are expressions of emotional support and public commitment. These elements are an important and significant aspect of the marital relationship” (Turner, 482 US at 95-96). Nor is there any conceivable rational basis for allowing prison inmates to marry, but not homosexuals. It is, of course, no answer that inmates could potentially procreate once they are released — that is, once they are no longer prisoners — since, as nonprisoners, they would then undeniably have a right to marry even in the absence of Turner.

Marriage is about much more than producing children, yet same-sex couples are excluded from the entire spectrum of protections that come with civil marriage — purportedly to encourage other people to procreate. Indeed, the protections that the State gives to couples who do marry — such as the right to own property as a unit or to make medical decisions for each other — are focused largely on the adult relationship, rather than on the couple’s possible role as parents. Nor does the *393plurality even attempt to explain how offering only heterosexuals the right to visit a sick loved one in the hospital, for example, conceivably furthers the State’s interest in encouraging opposite-sex couples to have children, or indeed how excluding same-sex couples from each of the specific legal benefits of civil marriage — even apart from the totality of marriage itself — does not independently violate plaintiffs’ rights to equal protection of the laws. The breadth of protections that the marriage laws make unavailable to gays and lesbians is “so far removed” from the State’s asserted goal of promoting procreation that the justification is, again, “impossible to credit” (Romer, 517 US at 635).

The State plainly has a legitimate interest in the welfare of children, but excluding same-sex couples from marriage in no way furthers this interest. In fact, it undermines it. Civil marriage provides tangible legal protections and economic benefits to married couples and their children, and tens of thousands of children are currently being raised by same-sex couples in New York. Depriving these children of the benefits and protections available to the children of opposite-sex couples is antithetical to their welfare, as defendants do not dispute (see e.g. Baker v State, 170 Vt 194, 219, 744 A2d 864, 882 [1999] [“(i)f anything, the exclusion of same-sex couples from the legal protections incident to marriage exposes their children to the precise risks that the State argues the marriage laws are designed to secure against”]; cf. Matter of Jacob, 86 NY2d 651, 656 [1995] [“(t)o rule otherwise would mean that the thousands of New York children actually being raised in homes headed by two unmarried persons could have only one legal parent, not the two who want them”]). The State’s interest in a stable society is rationally advanced when families are established and remain intact irrespective of the gender of the spouses.

Nor may the State legitimately seek either to promote heterosexual parents over homosexual parents, as the plurality posits, or to discourage same-sex parenting. First, granting such a preference to heterosexuals would be an acknowledgment of purposeful discrimination against homosexuals, thus constituting a flagrant equal protection violation. Second, such a preference would be contrary to the stated public policy of New York, and therefore irrational (see 18 NYCRR 421.16 [h] [2] [applicants to be adoptive parents “shall not be rejected solely on the basis of homosexuality”]; see also Jacob, 86 NY2d at 668 [same-sex partner of a legal parent may adopt that parent’s *394child; “(a)ny proffered justification for rejecting (adoptions) based on a governmental policy disapproving of homosexuality or encouraging marriage would not apply”]; brief of American Psychological Association et al., as amici curiae in support of plaintiffs, at 34-43 [collecting the results of social scientific research studies which conclude that children raised by same-sex parents fare no differently from, and do as well as, those raised by opposite-sex parents in terms of the quality of the parent-child relationship and the mental health, development and social adjustment of the child]; brief of Association to Benefit Children et al., as amici curiae in support of plaintiffs, at 31-35 [same conclusion]).5

2. Moral Disapproval

The government cannot legitimately justify discrimination against one group of persons as a mere desire to preference another group (see Metropolitan Life Ins. Co. v Ward, 470 US 869, 882 and n 10 [1985]). Further, the Supreme Court has held that classifications “drawn for the purpose of disadvantaging the group burdened by the law” can never be legitimate (Romer, 517 US at 633), and that “a bare . . . desire to harm a politically unpopular group cannot constitute a legitimate governmental interest” (Department of Agriculture v Moreno, 413 US 528, 534 [1973]; see also Onofre, 51 NY2d at 490 [“disapproval by a majority of the populace . . . may not substitute for the required demonstration of a valid basis for intrusion by the State in an area of important personal decision”]; Palmore v Sidoti, 466 US 429, 433 [1984] [“(p)rivate biases may be outside the reach of the law, but the law cannot, directly or indirectly, give them effect”]; Lawrence, 539 US at 571 [no legitimate basis to penalize gay and lesbian relationships notwithstanding that “for centuries there have been powerful voices to condemn homosexual conduct as immoral”]; id. at 583 [O’Connor, J., concurring in the judgment] [“(m)oral disapproval” of homosexuals cannot be a legitimate state interest]).

*3953. Tradition

That civil marriage has traditionally excluded same-sex couples — i.e., that the “historic and cultural understanding of marriage” has been between a man and a woman — cannot in itself provide a rational basis for the challenged exclusion. To say that discrimination is “traditional” is to say only that the discrimination has existed for a long time. A classification, however, cannot be maintained merely “for its own sake” (Romer, 517 US at 635). Instead, the classification (here, the exclusion of gay men and lesbians from civil marriage) must advance a state interest that is separate from the classification itself (see Romer, 517 US at 633, 635). Because the “tradition” of excluding gay men and lesbians from civil marriage is no different from the classification itself, the exclusion cannot be justified on the basis of “history.” Indeed, the justification of “tradition” does not explain the classification; it merely repeats it. Simply put, a history or tradition of discrimination — no matter how entrenched — does not make the discrimination constitutional (see also Goodridge, 440 Mass at 332 n 23, 798 NE2d at 961 n 23 [“it is circular reasoning, not analysis, to maintain that marriage must remain a heterosexual institution because that is what it historically has been”]).6

4. Uniformity

The State asserts an interest in maintaining uniformity with the marriage laws of other states. But our marriage laws currently are not uniform with those of other states. For example, New York — unlike most other states in the nation— permits first cousins to marry (see Domestic Relations Law § 5). This disparity has caused no trouble, however, because well-settled principles of comity resolve any conflicts. The same well-settled principles of comity would resolve any conflicts arising from any disparity involving the recognition of same-sex marriages.

It is, additionally, already impossible to maintain uniformity among all the states, inasmuch as Massachusetts has now legalized same-sex marriage. Indeed, of the seven jurisdictions that border New York State, only Pennsylvania cur*396rently affords no legal status to same-sex relationships. Massachusetts, Ontario and Quebec all authorize same-sex marriage; Vermont and Connecticut provide for civil unions (see Vt Stat Ann, tit 15, § 1204 [a]; Conn Gen Stat § 46b-38nn); and New Jersey has a statewide domestic partnership law (see NJ Stat Ann § 26:8A-1 et seq.). Moreover, insofar as a number of localities within New York offer domestic partnership registration, even the law within the state is not uniform. Finally, and most fundamentally, to justify the exclusion of gay men and lesbians from civil marriage because “others do it too” is no more a justification for the discriminatory classification than the contention that the discrimination is rational because it has existed for a long time. As history has well taught us, separate is inherently unequal.

III. The Legislature

The Court ultimately concludes that the issue of same-sex marriage should be addressed by the Legislature. If the Legislature were to amend the statutory scheme by making it gender neutral, obviously the instant controversy would disappear. But this Court cannot avoid its obligation to remedy constitutional violations in the hope that the Legislature might some day render the question presented academic. After all, by the time the Court decided Loving in 1967, many states had already repealed their antimiscegenation laws. Despite this trend, however, the Supreme Court did not refrain from fulfilling its constitutional obligation.

The fact remains that although a number of bills to authorize same-sex marriage have been introduced in the Legislature over the past several years, none has ever made it out of committee (see 2005 NY Senate-Assembly Bill S 5156, A 7463; 2005 NY Assembly Bill A 1823; 2003 NY Senate Bill S 3816; 2003 NY Assembly Bill A 7392; 2001 NY Senate Bill S 1205; see also 2005 NY Senate-Assembly Bill S 1887-A, A 3693-A [proposing establishment of domestic partnerships]; 2004 NY Senate-Assembly Bill S 3393-A, A 7304-A [same]).

It is uniquely the function of the Judicial Branch to safeguard individual liberties guaranteed by the New York State Constitution, and to order redress for their violation. The Court’s duty to protect constitutional rights is an imperative of the separation of powers, not its enemy.

I am confident that future generations will look back on today’s decision as an unfortunate misstep.

*397Judges G.B. Smith and Read concur with Judge R.S. Smith; Judge Gbaffeo concurs in result in a separate opinion in which Judge G.B. Smith concurs; Chief Judge Kaye dissents in another opinion in which Judge Ciparick concurs; Judge Rosenblatt taking no part.

In each case: Order affirmed, without costs.

Justice KENNEDY delivered the opinion of the Court.

*749Two women then resident in New York were married in a lawful ceremony in Ontario, Canada, in 2007. Edith Windsor *750and 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, *751however, 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 *752unconstitutional 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 *2683marriages 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).

*753Edith 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 (C.A.2 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 *754defend 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 *2684scrutiny 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 *755affirmed 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. ----, 133 S.Ct. 786, 184 L.Ed.2d 527 (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, 127 S.Ct. 2553, 168 L.Ed.2d 424 (2007) (plurality *2685opinion) (emphasis deleted). Windsor suffered a redressable injury *756when 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 *757Warth v. Seldin, 422 U.S. 490, 498, 95 S.Ct. 2197, 45 L.Ed.2d 343 (1975). The latter are "essentially matters of judicial self-governance." Id., at 500, 95 S.Ct. 2197. The Court has kept these two strands separate: "Article III standing, which enforces the Constitution's case-or-controversy requirement, see Lujan v. Defenders of Wildlife, 504 U.S. 555, 559-562, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992) ; and prudential standing, which embodies 'judicially self-imposed limits on the exercise of federal jurisdiction,' Allen [v. Wright, ] 468 U.S. [737,] 751, 104 S.Ct. 3315 [82 L.Ed.2d 556 (1984) ]." Elk Grove Unified School Dist. v. Newdow, 542 U.S. 1, 11-12, 124 S.Ct. 2301, 159 L.Ed.2d 98 (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.' "

*2686Lujan, supra, at 560-561, 112 S.Ct. 2130 (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, 100 S.Ct. 1166, 63 L.Ed.2d 427 (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, 95 S.Ct. 2197.

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 *758order directing the Treasury to pay money is "a real and immediate economic injury," Hein, 551 U.S., at 599, 127 S.Ct. 2553, 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, 103 S.Ct. 2764, 77 L.Ed.2d 317 (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, 103 S.Ct. 2764. 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 *759deported; if we uphold [the statute], the INS will execute its order and deport him.' " Id., at 939-940, 103 S.Ct. 2764 (quoting Chadha v. INS, 634 F.2d 408, 419 (C.A.9 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, 103 S.Ct. 2764. 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 *2687intended to enforce the challenged law against that party." Id., at 940, n. 12, 103 S.Ct. 2764.

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, 100 S.Ct. 1166, see also Camreta v. Greene, 563 U.S. ----, ----, 131 S.Ct. 2020, 2030, 179 L.Ed.2d 1118 (2011) ( "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, 100 S.Ct. 1166.

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 *760courts an inquiry as to the constitutionality of the legislative act.' " Ashwander v. TVA, 297 U.S. 288, 346, 56 S.Ct. 466, 80 L.Ed. 688 (1936) (Brandeis, J., concurring) (quoting Chicago & Grand Trunk R. Co. v. Wellman, 143 U.S. 339, 345, 12 S.Ct. 400, 36 L.Ed. 176 (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, 82 S.Ct. 691, 7 L.Ed.2d 663 (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, 95 S.Ct. 2197. 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, 103 S.Ct. 2764. 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, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000).

*761In the case now before the Court the attorneys for BLAG present a substantial argument for the constitutionality of § 3 of *2688DOMA. 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 (C.A.1 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 *762District 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. ----, ----, 132 S.Ct. 1421, 1427-1428, 182 L.Ed.2d 423 (2012) (quoting Marbury v. Madison, 1 Cranch 137, 177, 2 L.Ed. 60 (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.

*2689The 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 *763suggestion 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 *764same-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 same-sex 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 same-sex 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 *2690been 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. ----, 133 S.Ct. 1943, 186 L.Ed.2d 43 (2013) ; see also Ridgway v. Ridgway, 454 U.S. 46, 102 S.Ct. 49, 70 L.Ed.2d 39 (1981) ; Wissner v. Wissner, 338 U.S. 655, 70 S.Ct. 398, 94 L.Ed. 424 (1950). This is one example of the general principle that when the Federal Government acts in the *765exercise 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, 4 L.Ed. 579 (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);

*766Wash. 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.

*2691In 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, 87 S.Ct. 1817, 18 L.Ed.2d 1010 (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, 95 S.Ct. 553, 42 L.Ed.2d 532 (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, 63 S.Ct. 207, 87 L.Ed. 279 (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 *767States on the subject of marriage and divorce." Haddock v. Haddock, 201 U.S. 562, 575, 26 S.Ct. 525, 50 L.Ed. 867 (1906) ; see also In re Burrus, 136 U.S. 586, 593-594, 10 S.Ct. 850, 34 L.Ed. 500 (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 state-law policy decisions with respect to domestic relations. In De Sylva v. Ballentine, 351 U.S. 570, 76 S.Ct. 974, 100 L.Ed. 1415 (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, 76 S.Ct. 974. 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, 112 S.Ct. 2206, 119 L.Ed.2d 468 (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, 112 S.Ct. 2206 (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, 50 S.Ct. 154, 74 L.Ed. 489 (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, *768but a handful-such as Iowa and Washington, see Iowa Code § 595.19 (2009); Wash. Rev.Code § 26.04.020 *2692(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, 116 S.Ct. 1620, 134 L.Ed.2d 855 (1996) (quoting Louisville Gas & Elec. Co. v. Coleman, 277 U.S. 32, 37-38, 48 S.Ct. 423, 72 L.Ed. 770 (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.

*769In 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. ----, ----, 131 S.Ct. 2355, 2359, 180 L.Ed.2d 269 (2011). 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, 123 S.Ct. 2472, 156 L.Ed.2d 508 (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 *2693roots 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.

*770See U.S. Const., Amdt. 5; Bolling v. Sharpe, 347 U.S. 497, 74 S.Ct. 693, 98 L.Ed. 884 (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, 93 S.Ct. 2821, 37 L.Ed.2d 782 (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 2692 (quoting Romer, supra, at 633, 116 S.Ct. 1620). 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."

*771H.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 *2694for 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.

*772DOMA's principal effect is to identify a subset of state-sanctioned 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 state-sanctioned 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, 123 S.Ct. 2472, 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), *773507(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 *2695official, § 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 same-sex 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 *774eligibility. 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, 74 S.Ct. 693; Adarand Constructors, Inc. v. Penã, 515 U.S. 200, 217-218, 115 S.Ct. 2097, 132 L.Ed.2d 158 (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.

*775The 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 *2696disability 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 2707 - 2708 (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 state-by-state *776variations 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 2689. 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 2694, 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 2692, 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 2696 -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 2681. Justice SCALIA believes this is a " 'bald, unreasoned disclaime[r].' "

*2697Post, at 2709. In my view, though, the disclaimer is a logical and necessary consequence of the argument the majority has chosen to *777adopt. 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 2690, 2692. 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 2692. 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 2692.

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 2693; 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 *778offspring, property interests, and the enforcement of marital responsibilities.' " Ante, at 2694, 2690. 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 2693, 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, --- U.S., at ----, 133 S.Ct. 1521. 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 *2698power 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 *779but 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 2688. 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 *780branches 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 2688. 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 *2699majority 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 2688. *781In 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, free-standing 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, 94 S.Ct. 2925, 41 L.Ed.2d 706 (1974) ; United States v. Richardson, 418 U.S. 166, 179, 94 S.Ct. 2940, 41 L.Ed.2d 678 (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, 56 S.Ct. 466, 80 L.Ed. 688 (1936) (concurring opinion) (quoting Chicago & Grand Trunk R. Co. v. Wellman, 143 U.S. 339, 345, 12 S.Ct. 400, 36 L.Ed. 176 (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, 112 S.Ct. 2130, 119 L.Ed.2d 351 (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 *782court 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 *2700What 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.

*783The closest we have ever come to what the Court blesses today was our opinion in INS v. Chadha, 462 U.S. 919, 103 S.Ct. 2764, 77 L.Ed.2d 317 (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, 103 S.Ct. 2764. 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, 103 S.Ct. 2764. 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.

*784Id., at 939, 103 S.Ct. 2764. That remark is true, as a description of the judicial review conducted in the Court of Appeals, where the Houses of Congress *2701had 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, 103 S.Ct. 2764. 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, 103 S.Ct. 2764), 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 2685. 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 2686 -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.

*785The 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, 88 S.Ct. 1942, 20 L.Ed.2d 947 (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, 127 S.Ct. 2553, 168 L.Ed.2d 424 (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, 100 S.Ct. 1166, 63 L.Ed.2d 427 (1980), the District Court had entered judgment in the individual plaintiff's favor based on the *2702defendant 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. ----, 131 S.Ct. 2020, 179 L.Ed.2d 1118 (2011). *786There 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 2687, 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 *787Presidential 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, 2 L.Ed. 60 (1803) that "[i]t is emphatically the province and duty of the judicial department to say what the law is." Ante, at 2688 (internal quotation *2703marks 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, 12 L.Ed. 1067 (1850) :

*788"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, 2 L.Ed. 60.

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 *789powers to act, but to correct a perceived inadequacy in the execution of its laws.3 This would lay to rest Tocqueville's *2704praise 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 *790by 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, 117 S.Ct. 2312, 138 L.Ed.2d 849 (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, decades-long 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 (C.A.D.C.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 *791what "[a]mbition ... counteract[ing] ambition," The Federalist, No. 51, at 322 (J. Madison), is all about. If majorities in both *2705Houses 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 *792Constitution," 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 2681. 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 2681. 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 *2706Amendment [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 2695. The only possible interpretation of *793this 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, 74 S.Ct. 693, 98 L.Ed. 884 (1954), Department of Agriculture v. Moreno, 413 U.S. 528, 93 S.Ct. 2821, 37 L.Ed.2d 782 (1973), and Romer v. Evans, 517 U.S. 620, 116 S.Ct. 1620, 134 L.Ed.2d 855 (1996) -all of which are equal-protection 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, 123 S.Ct. 2472, 156 L.Ed.2d 508 (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 2694.

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 (C.A.2 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, 116 S.Ct. 2264, 135 L.Ed.2d 735 (1996) (SCALIA, J., dissenting). As nearly as I can tell, the Court agrees with that; its opinion *794does 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, 113 S.Ct. 2637, 125 L.Ed.2d 257 (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 2695; that it violates "basic due process" principles, ante, at 2693; 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 2692. 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 *2707does 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, 117 S.Ct. 2258, 138 L.Ed.2d 772 (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, 117 S.Ct. 2258 (quoting Palko v. Connecticut, 302 U.S. 319, 325, 58 S.Ct. 149, 82 L.Ed. 288 (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-due-process 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 2693. It is this proposition with which I will therefore engage.

*795B

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, 123 S.Ct. 2472, 156 L.Ed.2d 508 (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, 88 S.Ct. 1673, 20 L.Ed.2d 672 (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 2693. 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, 107 S.Ct. 2573, 96 L.Ed.2d 510 (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 *796evidence, 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 2693. I imagine that this is because it is harder to maintain the illusion *2708of 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, 892 N.Y.S.2d 272, 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 *797exemption 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 2695) "to disparage and to injure" same-sex couples. It says that the motivation for DOMA was to "demean," ibid .; to "impose inequality," ante, at 2694; to "impose ... a stigma," ante, at 2692; to deny people "equal dignity," ibid. ; to brand gay people as "unworthy," ante, at 2694; 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 *798reasoned disagreement. To question its high-handed *2709invalidation 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 2696, 2695. I have heard such "bald, unreasoned disclaimer[s]" before. Lawrence, 539 U.S., at 604, 123 S.Ct. 2472. 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, 123 S.Ct. 2472. Now we are told that DOMA is invalid because it "demeans the couple, whose moral and sexual choices the Constitution protects," ante, at 2694 -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.

*799I do not mean to suggest disagreement with THE CHIEF JUSTICE's view, ante, pp. 2696 - 2697 (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 2691. 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 2694:

"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 *2710sexual relationships, but not other couples, of both rights and responsibilities."

Or try this passage, from ante, at 2694:

"[DOMA] This state law tells those couples, and all the world, that their otherwise valid marriages relationships *800are unworthy of federalstate 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 2694 -which does not even require alteration, except as to the invented number:

"And it humiliates tens ofthousands 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 2696. 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 2695, 2696. 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 *801marriage-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 *2711and 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 *802In 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 same-sex 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 *803of 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 *2712so 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, --- U.S., at ----, 133 S.Ct. 1521 -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, 100 S.Ct. 1166, 63 L.Ed.2d 427 (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.1

*804A 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. ----, ----, 131 S.Ct. 2020, 2028, 179 L.Ed.2d 1118 (2011) (quoting Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-561, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992) ). In the present case, the House of Representatives, which has authorized BLAG to represent its interests in this matter,2 suffered just such an injury.

In INS v. Chadha, 462 U.S. 919, 103 S.Ct. 2764, 77 L.Ed.2d 317 (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, 103 S.Ct. 2764 (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. *, 103 S.Ct. 2764. 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 *805that the statute *2713is inapplicable or unconstitutional." Id ., at 940, 103 S.Ct. 2764.

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, 59 S.Ct. 972, 83 L.Ed. 1385 (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, 59 S.Ct. 972 ("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, 59 S.Ct. 972 ("[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 *806constitutional 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, 117 S.Ct. 2312, 138 L.Ed.2d 849 (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, 117 S.Ct. 2312; 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)).

*2714Second, 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 *807ground that their votes have been completely nullified." 521 U.S., at 823, 117 S.Ct. 2312. 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, 103 S.Ct. 2764. See also 2 U.S.C. § 288h(7) (Senate Legal Counsel shall defend the constitutionality of Acts of Congress when placed in issue).3 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, 74 S.Ct. 693, 98 L.Ed. 884 (1954). The Court rests its holding on related arguments. See ante, at 2694 - 2695.

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.

*808The 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 2695, 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, 117 S.Ct. 2258, 138 L.Ed.2d 772 (1997) ; Snyder v. Massachusetts, 291 U.S. 97, 105, 54 S.Ct. 330, 78 L.Ed. 674 (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, 117 S.Ct. 2258 (quoting Palko v. Connecticut, *2715302 U.S. 319, 325-326, 58 S.Ct. 149, 82 L.Ed. 288 (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 2000.4

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 *809judges. 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 same-sex 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.5 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).6 Others *810think *2716that 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.

*811III

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 2695 -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, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971) (quoting F.S. Royster Guano Co. v. Virginia, 253 U.S. 412, 415, 40 S.Ct. 560, 64 L.Ed. 989 (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, 92 S.Ct. 251.

*2717*812So, 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, 127 S.Ct. 2738, 168 L.Ed.2d 508 (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, 105 S.Ct. 3249, 87 L.Ed.2d 313 (1985) ; cf. id ., at 452-453, 105 S.Ct. 3249 (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, 116 S.Ct. 2264, 135 L.Ed.2d 735 (1996) ; id., at 567, 116 S.Ct. 2264 (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, 105 S.Ct. 3249. 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, 101 S.Ct. 1200, 67 L.Ed.2d 437 (1981) (plurality opinion). 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, 92 S.Ct. 251, *813and based on "outmoded notions of the relative capabilities of men and women," Cleburne,supra, at 441, 105 S.Ct. 3249, 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, 92 S.Ct. 251.

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, 105 S.Ct. 3249. 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, 116 S.Ct. 1620, 134 L.Ed.2d 855 (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. Penã, 515 U.S. 200, 218, 115 S.Ct. 2097, 132 L.Ed.2d 158 (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, 105 S.Ct. 3249.

In asking the Court to determine that § 3 of DOMA is subject to and violates heightened scrutiny, Windsor and the *2718United 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.

*814The 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, 821 N.Y.S.2d 770, 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 "consent-based" 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 *815this 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.7 Because our constitutional *2719order *816assigns 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, 111 S.Ct. 1759, 114 L.Ed.2d 233 (1991) ("[T]he government 'may make a value judgment favoring childbirth over abortion' " (quoting Maher v. Roe, 432 U.S. 464, 474, 97 S.Ct. 2376, 53 L.Ed.2d 484 (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 2693 ("As the title and dynamics of the bill indicate, its purpose is to *817discourage enactment of state same-sex marriage laws and to restrict the freedom and choice of couples married under those *2720laws 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 (C.A.1 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 2695 - 2696 (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 *818Constitution 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.