Plessy v. Ferguson.

May 18, 1896

*538This was a petition for writs of prohibition and certiorari originally filed in the supreme court of the state by Plessy, the plaintiff in error, against the Hon. John H. Ferguson, judge of the criminal district court for the parish of Orleans, and setting forth, in substance, the following facts:

That petitioner was a citizen of the United States and a resident of the state of Louisiana, of mixed descent, in the proportion of seven-eighths Caucasian and one-eighth African blood; that the mixture of colored blood was not discernible in him, and that he was entitled to every recognition, right, privilege, and immunity secured to the citizens of the United States of the white race by its constitution and laws; that on June 7, 1892, he engaged and paid for a first-class passage on the East Louisiana Railway, from New Orleans to Covington, in the same state, and thereupon entered a passenger train, and took possession of a vacant seat in a coach where passengers of the white race were accommodated; that such railroad company was incorporated by the laws of Louisiana as a common carrier, and was not authorized to distinguish between citizens according to their race, but, notwithstanding this, petitioner was required by the conductor, under penalty of ejection from said train and imprisonment, to vacate said coach, and occupy another seat, in a coach assigned by said company for persons not of the white race, and for no other reason than that petitioner was of the colored race; that, upon petitioner's refusal to comply with such order, he was, with the aid of a police officer, forcibly ejected from said coach, and hurried off to, and imprisoned in, the parish jail of*539 New Orleans, and there held to answer a charge made by such officer to the effect that he was guilty of having criminally violated an act of the general assembly of the state, approved July 10, 1890, in such case made and provided.

The petitioner was subsequently brought before the recorder of the city for preliminary examination, and committed for trial to the criminal district court for the parish of Orleans, where an information was filed against him in the matter above set forth, for a violation of the above act, which act the petitioner affirmed to be null and void, because in conflict with the constitution of the United States; that petitioner interposed a plea to such information, based upon the unconstitutionality of the act of the general assembly, to which the district attorney, on behalf of the state, filed a demurrer; that, upon issue being joined upon such demurrer and plea, the court sustained the demurrer, overruled the plea, and ordered petitioner to plead over to the facts set forth in the information, and that, unless the judge of the said court be enjoined by a writ of prohibition from further proceeding in such case, the court will proceed to fine and sentence petitioner to imprisonment, and thus deprive him of his constitutional rights set forth in his said plea, notwithstanding the unconstitutionality of the act under which he was being prosecuted; that no appeal lay from such sentence, and petitioner was without relief or remedy except by writs of prohibition and certiorari. Copies of the information and other proceedings in the criminal district court were annexed to the petition as an exhibit.

Upon the filing of this petition, an order was issued upon the respondent to show cause why a writ of prohibition should not issue, and be made perpetual, and a further order that the record of the proceedings had in the criminal cause be certified and transmitted to the supreme court.

To this order the respondent made answer, transmitting a certified copy of the proceedings, asserting the constitutionality of the law, and averring that, instead of pleading or admitting that he belonged to the colored race, the said Plessy declined and refused, either by pleading or otherwise, to ad*540mit that he was in any sense or in any proportion a colored man.

The case coming on for hearing before the supreme court, that court was of opinion that the law under which the prosecution was had was constitutional and denied the relief prayed for by the petitioner (Ex parte Plessy, 45 La. Ann. 80, 11 South. 948); whereupon petitioner prayed for a writ of error from this court, which was allowed by the chief justice of the supreme court of Louisiana.

Mr. Justice Harlan dissenting.

A. W. Tourgee and S. F. Phillips, for plaintiff in error.

Alex. Porter Morse, for defendant in error.

Mr. Justice BROWN, after stating the facts in the foregoing language, delivered the opinion of the court.

This case turns upon the constitutionality of an act of the general assembly of the state of Louisiana, passed in 1890, providing for separate railway carriages for the white and colored races. Acts 1890, No. 111, p. 152.

The first section of the statute enacts 'that all railway companies carrying passengers in their coaches in this state, shall provide equal but separate accommodations for the white, and colored races, by providing two or more passenger coaches for each passenger train, or by dividing the passenger coaches by a partition so as to secure separate accommodations: provided, that this section shall not be construed to apply to street railroads. No person or persons shall be permitted to occupy seats in coaches, other than the ones assigned to them, on account of the race they belong to.'

By the second section it was enacted 'that the officers of such passenger trains shall have power and are hereby required*541 to assign each passenger to the coach or compartment used for the race to which such passenger belongs; any passenger insisting on going into a coach or compartment to which by race he does not belong, shall be liable to a fine of twenty-five dollars, or in lieu thereof to imprisonment for a period of not more than twenty days in the parish prison, and any officer of any railroad insisting on assigning a passenger to a coach or compartment other than the one set aside for the race to which said passenger belongs, shall be liable to a fine of twenty-five dollars, or in lieu thereof to imprisonment for a period of not more than twenty days in the parish prison; and should any passenger refuse to occupy the coach or compartment to which he or she is assigned by the officer of such railway, said officer shall have power to refuse to carry such passenger on his train, and for such refusal neither he nor the railway company which he represents shall be liable for damages in any of the courts of this state.'

The third section provides penalties for the refusal or neglect of the officers, directors, conductors, and employees of railway companies to comply with the act, with a proviso that 'nothing in this act shall be construed as applying to nurses attending children of the other race.' The fourth section is immaterial.

The information filed in the criminal district court charged, in substance, that Plessy, being a passenger between two stations within the state of Louisiana, was assigned by officers of the company to the coach used for the race to which he belonged, but he insisted upon going into a coach used by the race to which he did not belong. Neither in the information nor plea was his particular race or color averred.

The petition for the writ of prohibition averred that petitioner was seven-eights Caucasian and one-eighth African blood; that the mixture of colored blood was not discernible in him; and that he was entitled to every right, privilege, and immunity secured to citizens of the United States of the white race; and that, upon such theory, he took possession of a vacant seat in a coach where passengers of the white race were accommodated, and was ordered by the conductor to vacate*542 said coach, and take a seat in another, assigned to persons of the colored race, and, having refused to comply with such demand, he was forcibly ejected, with the aid of a police officer, and imprisoned in the parish jail to answer a charge of having violated the above act.

The constitutionality of this act is attacked upon the ground that it conflicts both with the thirteenth amendment of the constitution, abolishing slavery, and the fourteenth amendment, which prohibits certain restrictive legislation on the part of the states.

1. That it does not conflict with the thirteenth amendment, which abolished slavery and involuntary servitude, except § a punishment for crime, is too clear for argument. Slavery implies involuntary servitude,—a state of bondage; the ownership of mankind as a chattel, or, at least, the control of the labor and services of one man for the benefit of another, and the absence of a legal right to the disposal of his own person, property, and services. This amendment was said in the Slaughter-House Cases, 16 Wall. 36, to have been intended primarily to abolish slavery, as it had been previously known in this country, and that it equally forbade Mexican peonage or the Chinese coolie trade, when they amounted to slavery or involuntary servitude, and that the use of the word 'servitude' was intended to prohibit the use of all forms of involuntary slavery, of whatever class or name. It was intimated, however, in that case, that this amendment was regarded by the statesmen of that day as insufficient to protect the colored race from certain laws which had been enacted in the Southern states, imposing upon the colored race onerous disabilities and burdens, and curtailing their rights in the pursuit of life, liberty, and property to such an extent that their freedom was of little value; and that the fourteenth amendment was devised to meet this exigency.

So, too, in the Civil Rights Cases, 109 U. S. 3, 3 Sup. Ct. 18, it was said that the act of a mere individual, the owner of an inn, a public conveyance or place of amusement, refusing accommodations to colored people, cannot be justly regarded as imposing any badge of slavery or servitude upon the applicant, but*543 only as involving an ordinary civil injury, properly cognizable by the laws of the state, and presumably subject to redress by those laws until the contrary appears. 'It would be running the slavery question into the ground,' said Mr. Justice Bradley, 'to make it apply to every act of discrimination which a person may see fit to make as to the guests he will entertain, or as to the people he will take into his coach or cab or car, or admit to his concert or theater, or deal with in other matters of intercourse or business.'

A statute which implies merely a legal distinction between the white and colored races—a distinction which is founded in the color of the two races, and which must always exist so long as white men are distinguished from the other race by color—has no tendency to destroy the legal equality of the two races, or re-establish a state of involuntary servitude. Indeed, we do not understand that the thirteenth amendment is strenuously relied upon by the plaintiff in error in this connection.

2. By the fourteenth amendment, all persons born or naturalized in the United States, and subject to the jurisdiction thereof, are made citizens of the United States and of the state wherein they reside; and the states are forbidden from making or enforcing any law which shall abridge the privileges or immunities of citizens of the United States, or shall deprive any person of life, liberty, or property without due process of law, or deny to any person within their jurisdiction the equal protection of the laws.

The proper construction of this amendment was first called to the attention of this court in the Slaughter-House Cases, 16 Wall. 36, which involved, however, not a question of race, but one of exclusive privileges. The case did not call for any expression of opinion as to the exact rights it was intended to secure to the colored race, but it was said generally that its main purpose was to establish the citizenship of the negro, to give definitions of citizenship of the United States and of the states, and to protect from the hostile legislation of the states the privileges and immunities of citizens of the United States, as distinguished from those of citizens of the states.

*544The object of the amendment was undoubtedly to enforce the absolute equality of the two races before the law, but, in the nature of things, it could not have been intended to abolish distinctions based upon color, or to enforce social, as distinguish d from political, equality, or a commingling of the two races upon terms unsatisfactory to either. Laws permitting, and even requiring, their separation, in places where they are liable to be brought into contact, do not necessarily imply the inferiority of either race to the other, and have been generally, if not universally, recognized as within the competency of the state legislatures in the exercise of their police power. The most common instance of this is connected with the establishment of separate schools for white and colored children, which have been held to be a valid exercise of the legislative power even by courts of states where the political rights of the colored race have been longest and most earnestly enforced.

One of the earliest of these cases is that of Roberts v. City of Boston, 5 Cush. 198, in which the supreme judicial court of Massachusetts held that the general school committee of Boston had power to make provision for the instruction of colored children in separate schools established exclusively for them, and to prohibit their attendance upon the other schools.

'The great principle,' said Chief Justice Shaw, 'advanced by the learned and eloquent advocate for the plaintiff [Mr. Charles Sumner], is that, by the constitution and laws of Massachusetts, all persons, without distinction of age or sex, birth or color, origin or condition, are equal before the law. * * * But, when this great principle comes to be applied to the actual and various conditions of persons in society, it will not warrant the assertion that men and women are legally clothed with the same civil and political powers, and that children and adults are legally to have the same functions and be subject to the same treatment; but only that the rights of all, as they are settled and regulated by law, are equally entitled to the paternal consideration and protection of the law for their maintenance and security.'

It was held that the powers of the committee extended to the establish*545ment of separate schools for children of different ages, sexes and colors, and that they might also establish special schools for poor and neglected children, who have become too old to attend the primary school, and yet have not acquired the rudiments of learning, to enable them to enter the ordinary schools. Similar laws have been enacted by congress under its general power of legislation over the District of Columbia (sections 281-283, 310, 319, Rev. St. D. C.), as well as by the legislatures of many of the states, and have been generally, if not uniformly, sustained by the courts. State v. McCann, 21 Ohio St. 210; Lehew v. Brummell (Mo. Sup.) 15 S. W. 765; Ward v. Flood, 48 Cal. 36; Bertonneau v. Directors of City Schools, 3 Woods, 177, Fed. Cas. No. 1,361; People v. Gallagher, 93 N. Y. 438; Cory v. Carter, 48 Ind. 337; Dawson v. Lee, 83 Ky. 49.

Laws forbidding the intermarriage of the two races may be said in a technical sense to interfere with the freedom of contract, and yet have been universally recognized as within the police power of the state. State v. Gibson, 36 Ind. 389.

The distinction between laws interfering with the political equality of the negro and those requiring the separation of the two races in schools, theaters, and railway carriages has been frequently drawn by this court. Thus, in Strauder v. West Virginia, 100 U. S. 303, it was held that a law of West Virginia limiting to white male persons 21 years of age, and citizens of the state, the right to sit upon juries, was a discrimination which implied a legal inferiority in civil society, which lessened the security of the right of the colored race, and was a step towards reducing them to a condition of servility. Indeed, the right of a colored man that, in the selection of jurors to pass upon his life, liberty, and property, there shall be no exclusion of his race, and no discrimination against them because of color, has been asserted in a number of cases. Virginia v. Rivers, 100 U. S. 313; Neal v. Delaware, 103 U. S. 370; Bush v. Com., 107 U. S. 110, 1 Sup. Ct. 625; Gibson v. Mississippi, 162 U. S. 565, 16 Sup. Ct. 904. So, where the laws of a particular locality or the charter of a particular railway corporation has provided that no person shall be excluded from the cars on account of*546 color, we have held that this meant that persons of color should travel in the same car as white ones, and that the enactment was not satisfied by the company providing cars assigned exclusively to people of color, though they were as good as those which they assigned exclusively to white persons. Railroad Co. v. Brown, 17 Wall. 445.

Upon the other hand, where a statute of Louisiana required those engaged in the transportation of passengers among the states to give to all persons traveling within that state, upon vessels employed in that business, equal rights and privileges in all parts of the vessel, without distinction on account of race or color, and subjected to an action for damages the owner of such a vessel who excluded colored passengers on account of their color from the cabin set aside by him for the use of whites, it was held to be, so far as it applied to interstate commerce, unconstitutional and void. Hall v. De Cuir, 95 U. S. 485. The court in this case, however, expressly disclaimed that it had anything whatever to do with the statute as a regulation of internal commerce, or affecting anything else than commerce among the states.

In the Civil Rights Cases, 109 U. S. 3, 3 Sup. Ct. 18, it was held that an act of congress entitling all persons within the jurisdiction of the United States to the full and equal enjoyment of the accommodations, advantages, facilities, and privileges of inns, public conveyances, on land or water, theaters, and other places of public amusement, and made applicable to citizens of every race and color, regardless of any previous condition of servitude, was unconstitutional and void, upon the ground that the fourteenth amendment was prohibitory upon the states only, and the legislation authorized to be adopted by congress for enforcing it was not direct legislation on matters respecting which the states were prohibited from making or enforcing certain laws, or doing certain acts, but was corrective legislation, such as might be necessary or proper for counter-acting and redressing the effect of such laws or acts. In delivering the opinion of the court, Mr. Justice Bradley observed that the fourteenth amendment 'does not invest congress with power to legislate upon subjects that are within the*547 domain of state legislation, but to provide modes of relief against state legislation or state action of the kind referred to. It does not authorize congress to create a code of municipal law for the regulation of private rights, but to provide modes of redress against the operation of state laws, and the action of state officers, executive or judicial, when these are subversive of the fundamental rights specified in the amendment. Positive rights and privileges are undoubtedly secured by the fourteenth amendment; but they are secured by way of prohibition against state laws and state proceedings affecting those rights and privileges, and by power given to congress to legislate for the purpose of carrying such prohibition into effect; and such legislation must necessarily be predicated upon such supposed state laws or state proceedings, and be directed to the correction of their operation and effect.'

Much nearer, and, indeed, almost directly in point, is the case of the Louisville, N. O. & T. Ry. Co. v. State, 133 U. S. 587, 10 Sup. Ct. 348, wherein the railway company was indicted for a violation of a statute of Mississippi, enacting that all railroads carrying passengers should provide equal, but separate, accommodations for the white and colored races, by providing two or more passenger cars for each passenger train, or by dividing the passenger cars by a partition, so as to secure separate accommodations. The case was presented in a different aspect from the one under consideration, inasmuch as it was an indictment against the railway company for failing to provide the separate accommodations, but the question considered was the constitutionality of the law. In that case, the supreme court of Mississippi (66 Miss. 662, 6 South. 203) had held that the statute applied solely to commerce within the state, and, that being the construction of the state statute by its highest court, was accepted as conclusive. 'If it be a matter,' said the court (page 591, 133 U. S., and page 348, 10 Sup. Ct.), 'respecting commerce wholly within a state, and not interfering with commerce between the states, then, obviously, there is no violation of the commerce clause of the federal constitution. * * * No question arises under this section as to the power of the state to separate in different compartments interstate pas*548sengers, or affect, in any manner, the privileges and rights of such passengers. All that we can consider is whether the state has the power to require that railroad trains within her limits shall have separate accommodations for the two races. That affecting only commerce within the state is no invasion of the power given to congress by the commerce clause.'

A like course of reasoning applies to the case under consideration, since the supreme court of Louisiana, in the case of State v. Judge, 44 La. Ann. 770, 11 South. 74, held that the statute in question did not apply to interstate passengers, but was confined in its application to passengers traveling exclusively within the borders of the state. The case was decided largely upon the authority of Louisville, N. O. & T. Ry. Co. v. State, 66 Miss. 662, 6 South, 203, and affirmed by this court in 133 U. S. 587, 10 Sup. Ct. 348. In the present case no question of interference with interstate commerce can possibly arise, since the East Louisiana Railway appears to have been purely a local line, with both its termini within the state of Louisiana. Similar statutes for the separation of the two races upon public conveyances were held to be constitutional in Railroad v. Miles, 55 Pa. St. 209; Day v. Owen 5 Mich. 520; Railway Co. v. Williams, 55 Ill. 185; Railroad Co. v. Wells, 85 Tenn. 613; 4 S. W. 5; Railroad Co. v. Benson, 85 Tenn. 627, 4 S. W. 5; The Sue, 22 Fed. 843; Logwood v. Railroad Co., 23 Fed. 318; McGuinn v. Forbes, 37 Fed. 639; People v. King (N. Y. App.) 18 N. E. 245; Houck v. Railway Co., 38 Fed. 226; Heard v. Railroad Co., 3 Inter St. Commerce Com. R. 111, 1 Inter St. Commerce Com. R. 428.

While we think the enforced separation of the races, as applied to the internal commerce of the state, neither abridges the privileges or immunities of the colored man, deprives him of his property without due process of law, nor denies him the equal protection of the laws, within the meaning of the fourteenth amendment, we are not prepared to say that the conductor, in assigning passengers to the coaches according to their race, does not act at his peril, or that the provision of the second section of the act that denies to the passenger compensa*549tion in damages for a refusal to receive him into the coach in which he properly belongs is a valid exercise of the legislative power. Indeed, we understand it to be conceded by the state's attorney that such part of the act as exempts from liability the railway company and its officers is unconstitutional. The power to assign to a particular coach obviously implies the power to determine to which race the passenger belongs, as well as the power to determine who, under the laws of the particular state, is to be deemed a white, and who a colored, person. This question, though indicated in the brief of the plaintiff in error, does not properly arise upon the record in this case, since the only issue made is as to the unconstitutionality of the act, so far as it requires the railway to provide separate accommodations, and the conductor to assign passengers according to their race.

It is claimed by the plaintiff in error that, in an mixed community, the reputation of belonging to the dominant race, in this instance the white race, is 'property,' in the same sense that a right of action or of inheritance is property. Conceding this to be so, for the purposes of this case, we are unable to see how this statute deprives him of, or in any way affects his right to, such property. If he be a white man, and assigned to a colored coach, he may have his action for damages against the company for being deprived of his so-called 'property.' Upon the other hand, if he be a colored man, and be so assigned, he has been deprived of no property, since he is not lawfully entitled to the reputation of being a white man.

In this connection, it is also suggested by the learned counsel for the plaintiff in error that the same argument that will justify the state legislature in requiring railways to provide separate accommodations for the two races will also authorize them to require separate cars to be provided for people whose hair is of a certain color, or who are aliens, or who belong to certain nationalities, or to enact laws requiring colored people to walk upon one side of the street, and white people upon the other, or requiring white men's houses to be painted white, and colored men's black, or their vehicles or business signs to be of different colors, upon the theory that one side*550 of the street is as good as the other, or that a house or vehicle of one color is as good as one of another color. The reply to all this is that every exercise of the police power must be reasonable, and extend only to such laws as are enacted in good faith for the promotion of the public good, and not for the annoyance or oppression of a particular class. Thus, in Yick Wo v. Hopkins, 118 U. S. 356, 6 Sup. Ct. 1064, it was held by this court that a municipal ordinance of the city of San Francisco, to regulate the carrying on of public laundries within the limits of the municipality, violated the provisions of the constitution of the United States, if it conferred upon the municipal authorities arbitrary power, at their own will, and without regard to discretion, in the legal sense of the term, to give or withhold consent as to persons or places, without regard to the competency of the persons applying or the propriety of the places selected for the carrying on of the business. It was held to be a covert attempt on the part of the municipality to make an arbitrary and unjust discrimination against the Chinese race. While this was the case of a municipal ordinance, a like principle has been held to apply to acts of a state legislature passed in the exercise of the police power. Railroad Co. v. Husen, 95 U. S. 465; Louisville & N. R. Co. v. Kentucky, 161 U. S. 677, 16 Sup. Ct. 714, and cases cited on page 700, 161 U. S., and page 714, 16 Sup. Ct.; Daggett v. Hudson, 43 Ohio St. 548, 3 N. E. 538; Capen v. Foster, 12 Pick. 485; State v. Baker, 38 Wis. 71; Monroe v. Collins, 17 Ohio St. 665; Hulseman v. Rems, 41 Pa. St. 396; Osman v. Riley, 15 Cal. 48.

So far, then, as a conflict with the fourteenth amendment is concerned, the case reduces itself to the question whether the statute of Louisiana is a reasonable regulation, and with respect to this there must necessarily be a large discretion on the part of the legislature. In determining the question of reasonableness, it is at liberty to act with reference to the established usages, customs, and traditions of the people, and with a view to the promotion of their comfort, and the preservation of the public peace and good order. Gauged by this standard, we cannot say that a law which authorizes or even requires the separation of the two races in public conveyances*551 is unreasonable, or more obnoxious to the fourteenth amendment than the acts of congress requiring separate schools for colored children in the District of Columbia, the constitutionality of which does not seem to have been questioned, or the corresponding acts of state legislatures.

We consider the underlying fallacy of the plaintiff's argument to consist in the assumption that the enforced separation of the two races stamps the colored race with a badge of inferiority. If this be so, it is not by reason of anything found in the act, but solely because the colored race chooses to put that construction upon it. The argument necessarily assumes that if, as has been more than once the case, and is not unlikely to be so again, the colored race should become the dominant power in the state legislature, and should enact a law in precisely similar terms, it would thereby relegate the white race to an inferior position. We imagine that the white race, at least, would not acquiesce in this assumption. The argument also assumes that social prejudices may be overcome by legislation, and that equal rights cannot be secured to the negro except by an enforced commingling of the two races. We cannot accept this proposition. If the two races are to meet upon terms of social equality, it must be the result of natural affinities, a mutual appreciation of each other's merits, and a voluntary consent of individuals. As was said by the court of appeals of New York in People v. Gallagher, 93 N. Y. 438, 448:

'This end can neither be accomplished nor promoted by laws which conflict with the general sentiment of the community upon whom they are designed to operate. When the government, therefore, has secured to each of its citizens equal rights before the law, and equal opportunities for improvement and progress, it has accomplished the end for which it was organized, and performed all of the functions respecting social advantages with which it is endowed.'

Legislation is powerless to eradicate racial instincts, or to abolish distinctions based upon physical differences, and the attempt to do so can only result in accentuating the difficulties of the present situation. If the civil and political rights of both races be equal, one cannot be inferior to the other civilly*552 or politically. If one race be inferior to the other socially, the constitution of the United States cannot put them upon the same plane.

It is true that the question of the proportion of colored blood necessary to constitute a colored person, as distinguished from a white person, is one upon which there is a difference of opinion in the different states; some holding that any visible admixture of black blood stamps the person as belonging to the colored race (State v. Chavers, 5 Jones [N. C.] 1); others, that it depends upon the preponderance of blood (Gray v. State, 4 Ohio, 354; Monroe v. Collins, 17 Ohio St. 665); and still others, that the predominance of white blood must only be in the proportion of three-fourths (People v. Dean, 14 Mich. 406; Jones v. Com., 80 Va. 544). But these are questions to be determined under the laws of each state, and are not properly put in issue in this case. Under the allegations of his petition, it may undoubtedly become a question of importance whether, under the laws of Louisiana, the petitioner belongs to the white or colored race.

The judgment of the court below is therefore affirmed.

Mr. Justice BREWER did not hear the argument or participate in the decision of this case.

Mr. Justice HARLAN dissenting.

By the Louisiana statute the validity of which is here involved, all railway companies (other than street-railroad companies) carry passengers in that state are required to have separate but equal accommodations for white and colored persons, 'by providing two or more passenger coaches for each passenger train, or by dividing the passenger coaches by a partition so as to secure separate accommodations.' Under this statute, no colored person is permitted to occupy a seat in a coach assigned to white persons; nor any white person to occupy a seat in a coach assigned to colored persons. The managers of the railroad are not allowed to exercise any discretion in the premises, but are required to assign each passenger to some coach or compartment set apart for the exclusive use of is race. If a passenger insists upon going into a coach or compartment not set apart for persons of his race,*553 he is subject to be fined, or to be imprisoned in the parish jail. Penalties are prescribed for the refusal or neglect of the officers, directors, conductors, and employees of railroad companies to comply with the provisions of the act.

Only 'nurses attending children of the other race' are excepted from the operation of the statute. No exception is made of colored attendants traveling with adults. A white man is not permitted to have his colored servant with him in the same coach, even if his condition of health requires the constant personal assistance of such servant. If a colored maid insists upon riding in the same coach with a white woman whom she has been employed to serve, and who may need her personal attention while traveling, she is subject to be fined or imprisoned for such an exhibition of zeal in the discharge of duty.

While there may be in Louisiana persons of different races who are not citizens of the United States, the words in the act 'white and colored races' necessarily include all citizens of the United States of both races residing in that state. So that we have before us a state enactment that compels, under penalties, the separation of the two races in railroad passenger coaches, and makes it a crime for a citizen of either race to enter a coach that has been assigned to citizens of the other race.

Thus, the state regulates the use of a public highway by citizens of the United States solely upon the basis of race.

However apparent the injustice of such legislation may be, we have only to consider whether it is consistent with the constitution of the United States.

That a railroad is a public highway, and that the corporation which owns or operates it is in the exercise of public functions, is not, at this day, to be disputed. Mr. Justice Nelson, speaking for this court in New Jersey Steam Nav. Co. v. Merchants' Bank, 6 How. 344, 382, said that a common carrier was in the exercise 'of a sort of public office, and has public duties to perform, from which he should not be permitted to exonerate himself without the assent of the parties concerned.' Mr. Justice Strong, delivering the judgment of*554 this court in Olcott v. Supervisors, 16 Wall. 678, 694, said: 'That railroads, though constructed by private corporations, and owned by them, are public highways, has been the doctrine of nearly all the courts ever since such conveniences for passage and transportation have had any existence. Very early the question arose whether a state's right of eminent domain could be exercised by a private corporation created for the purpose of constructing a railroad. Clearly, it could not, unless taking land for such a purpose by such an agency is taking land for public use. The right of eminent domain nowhere justifies taking property for a private use. Yet it is a doctrine universally accepted that a state legislature may authorize a private corporation to take land for the construction of such a road, making compensation to the owner. What else does this doctrine mean if not that building a railroad, though it be built by a private corporation, is an act done for a public use?' So, in Township of Pine Grove v. Talcott, 19 Wall. 666, 676: 'Though the corporation [a railroad company] was private, its work was public, as much so as if it were to be constructed by the state.' So, in Inhabitants of Worcester v. Western R. Corp., 4 Metc. (Mass.) 564: 'The establishment of that great thoroughfare is regarded as a public work, established by public authority, intended for the public use and benefit, the use of which is secured to the whole community, and constitutes, therefore, like a canal, turnpike, or highway, a public easement.' 'It is true that the real and personal property, necessary to the establishment and management of the railroad, is vested in the corporation; but it is in trust for the public.'

In respect of civil rights, common to all citizens, the constitution of the United States does not, I think, permit any public authority to know the race of those entitled to be protected in the enjoyment of such rights. Every true man has pride of race, and under appropriate circumstances, when the rights of others, his equals before the law, are not to be affected, it is his privilege to express such pride and to take such action based upon it as to him seems proper. But I deny that any legislative body or judicial tribunal may have regard to the*555 race of citizens when the civil rights of those citizens are involved. Indeed, such legislation as that here in question is inconsistent not only with that equality of rights which pertains to citizenship, national and state, but with the personal liberty enjoyed by every one within the United States.

The thirteenth amendment does not permit the withholding or the deprivation of any right necessarily inhering in freedom. It not only struck down the institution of slavery as previously existing in the United States, but it prevents the imposition of any burdens or disabilities that constitute badges of slavery or servitude. It decreed universal civil freedom in this country. This court has so adjudged. But, that amendment having been found inadequate to the protection of the rights of those who had been in slavery, it was followed by the fourteenth amendment, which added greatly to the dignity and glory of American citizenship, and to the security of personal liberty, by declaring that 'all persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States and of the state wherein they reside,' and that 'no state shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any state deprive any person of life, liberty or property without due process of law, nor deny to any person within its jurisdiction the equal protection of the laws.' These two amendments, if enforced according to their true intent and meaning, will protect all the civil rights that pertain to freedom and citizenship. Finally, and to the end that no citizen should be denied, on account of his race, the privilege of participating in the political control of his country, it was declared by the fifteenth amendment that 'the right of citizens of the United States to vote shall not be denied or abridged by the United States or by any state on account of race, color or previous condition of servitude.'

These notable additions to the fundamental law were welcomed by the friends of liberty throughout the world. They removed the race line from our governmental systems. They had, as this court has said, a common purpose, namely, to secure 'to a race recently emancipated, a race that through*556 many generations have been held in slavery, all the civil rights that the superior race enjoy.' They declared, in legal effect, this court has further said, 'that the law in the states shall be the same for the black as for the white; that all persons, whether colored or white, shall stand equal before the laws of the states; and in regard to the colored race, for whose protection the amendment was primarily designed, that no discrimination shall be made against them by law because of their color.' We also said: 'The words of the amendment, it is true, are prohibitory, but they contain a necessary implication of a positive immunity or right, most valuable to the colored race,—the right to exemption from unfriendly legislation against them distinctively as colored; exemption from legal discriminations, implying inferiority in civil society, lessening the security of their enjoyment of the rights which others enjoy; and discriminations which are steps towards reducing them to the condition of a subject race.' It was, consequently, adjudged that a state law that excluded citizens of the colored race from juries, because of their race, however well qualified in other respects to discharge the duties of jurymen, was repugnant to the fourteenth amendment. Strauder v. West Virginia, 100 U. S. 303, 306, 307; Virginia v. Rives, Id. 313; Ex parte Virginia, Id. 339; Neal v. Delaware, 103 U. S. 370, 386; Bush v. Com., 107 U. S. 110, 116, 1 Sup. Ct. 625. At the present term, referring to the previous adjudications, this court declared that 'underlying all of those decisions is the principle that the constitution of the United States, in its present form, forbids, so far as civil and political rights are concerned, discrimination by the general government or the states against any citizen because of his race. All citizens are equal before the law.' Gibson v. State, 162 U. S. 565, 16 Sup. Ct. 904.

The decisions referred to show the scope of the recent amendments of the constitution. They also show that it is not within the power of a state to prohibit colored citizens, because of their race, from participating as jurors in the administration of justice.

It was said in argument that the statute of Louisiana does *557 not discriminate against either race, but prescribes a rule applicable alike to white and colored citizens. But this argument does not meet the difficulty. Every one knows that the statute in question had its origin in the purpose, not so much to exclude white persons from railroad cars occupied by blacks, as to exclude colored people from coaches occupied by or assigned to white persons. Railroad corporations of Louisiana did not make discrimination among whites in the matter of commodation for travelers. The thing to accomplish was, under the guise of giving equal accommodation for whites and blacks, to compel the latter to keep to themselves while traveling in railroad passenger coaches. No one would be so wanting in candor as to assert the contrary. The fundamental objection, therefore, to the statute, is that it interferes with the personal freedom of citizens. ‘Personal liberty,’ it has been well said, ‘consists in the power of locomotion, of changing situation, or removing one's person to whatsoever places one's own inclination may direct, without imprisonment or restraint, unless by due course of law.’ 1 Bl. Comm. *134. If a white man and a black man choose to occupy the same public conveyance on a public highway, it is their right to do so; and no government, proceeding alone on grounds of race, can prevent it without infringing the personal liberty of each.

It is one thing for railroad carriers to furnish, or to be required by law to furnish, equal accommodations for all whom they are under a legal duty to carry. It is quite another thing for government to forbid citizens of the white and black races from traveling in the same public conveyance, and to punish officers of railroad companies for permitting persons of the two races to occupy the same passenger coach. If a state can prescribe, as a rule of civil conduct, that whites and blacks shall not travel as passengers in the same railroad coach, why may it not so regulate the use of the streets of its cities and towns as to compel white citizens to keep on one side of a street, and black citizens to keep on the other? Why may it not, upon like grounds, punish whites and blacks who ride together in street cars or in open vehicles on a public road*558 or street? Why may it not require sheriffs to assign whites to one side of a court room, and blacks to the other? And why may it not also prohibit the commingling of the two races in the galleries of legislative halls or in public assemblages convened for the consideration of the political questions of the day? Further, if this statute of Louisiana is consistent with the personal liberty of citizens, why may not the state require the separation in railroad coaches of native and naturalized citizens of the United States, or of Protestants and Roman Catholics?

The answer given at the argument to these questions was that regulations of the kind they suggest would be unreasonable, and could not, therefore, stand before the la . Is it meant that the determination of questions of legislative power depends upon the inquiry whether the statute whose validity is questioned is, in the judgment of the courts, a reasonable one, taking all the circumstances into consideration? A statute may be unreasonable merely because a sound public policy forbade its enactment. But I do not understand that the courts have anything to do with the policy or expediency of legislation. A statute may be valid, and yet, upon grounds of public policy, may well be characterized as unreasonable. Mr. Sedgwick correctly states the rule when he says that, the legislative intention being clearly ascertained, 'the courts have no other duty to perform than to execute the legislative will, without any regard to their views as to the wisdom or justice of the particular enactment.' Sedg. St. & Const. Law, 324. There is a dangerous tendency in these latter days to enlarge the functions of the courts, by means of judicial interference with the will of the people as expressed by the legislature. Our institutions have the distinguishing characteristic that the three departments of government are co-ordinate and separate. Each much keep within the limits defined by the constitution. And the courts best discharge their duty by executing the will of the law-making power, constitutionally expressed, leaving the results of legislation to be dealt with by the people through their representatives. Statutes must always have a reasonable construction. Sometimes they are to be construed strictly, sometimes literally, in order to carry out the legislative *559will. But, however construed, the intent of the legislature is to be respected if the particular statute in question is valid, although the courts, looking at the public interests, may conceive the statute to be both unreasonable and impolitic. If the power exists to enact a statute, that ends the matter so far as the courts are concerned. The adjudged cases in which statutes have been held to be void, because unreasonable, are those in which the means employed by the legislature were not at all germane to the end to which the legislature was competent.

The white race deems itself to be the dominant race in this country. And so it is, in prestige, in achievements, in education, in wealth, and in power. So, I doubt not, it will continue to be for all time, if it remains true to its great heritage, and holds fast to the principles of constitutional liberty. But in view of the constitution, in the eye of the law, there is in this country no superior, dominant, ruling class of citizens. There is no caste here. Our constitution is color-blind, and neither knows nor tolerates classes among citizens. In respect of civil rights, all citizens are equal before the law. The humblest is the peer of the most powerful. The law regards man as man, and takes no account of his surroundings or of his color when his civil rights as guarantied by the supreme law of the land are involved. It is therefore to be regretted that this high tribunal, the final expositor of the fundamental law of the land, has reached the conclusion that it is competent for a state to regulate the enjoyment by citizens of their civil rights solely upon the basis of race.

In my opinion, the judgment this day rendered will, in time, prove to be quite as pernicious as the decision made by this tribunal in the Dred Scott Case.

It was adjudged in that case that the descendants of Africans who were imported into this country, and sold as slaves, were not included nor intended to be included under the word 'citizens' in the constitution, and could not claim any of the rights and privileges which that instrument provided for and secured to citizens of the United States; that, at time of the adoption of the constitution, they were 'considered as a subordinate and inferior class of beings, who had been subjugated by the dominant*560 race, and, whether emancipated or not, yet remained subject to their authority, and had no rights or privileges but such as those who held the power and the government might choose to grant them.' 17 How. 393, 404. The recent amendments of the constitution, it was supposed, had eradicated these principles from our institutions. But it seems that we have yet, in some of the states, a dominant race,—a superior class of citizens,—which assumes to regulate the enjoyment of civil rights, common to all citizens, upon the basis of race. The present decision, it may well be apprehended, will not only stimulate aggressions, more or less brutal and irritating, upon the admitted rights of colored citizens, but will encourage the belief that it is possible, by means of state enactments, to defeat the beneficent purposes which the people of the United States had in view when they adopted the recent amendments of the constitution, by one of which the blacks of this country were made citizens of the United States and of the states in which they respectively reside, and whose privileges and immunities, as citizens, the states are forbidden to abridge. Sixty millions of whites are in no danger from the presence here of eight millions of blacks. The destinies of the two races, in this country, are indissolubly linked together, and the interests of both require that the common government of all shall not permit the seeds of race hate to be planted under the sanction of law. What can more certainly arouse race hate, what more certainly create and perpetuate a feeling of distrust between these races, than state enactments which, in fact, proceed on the ground that colored citizens are so inferior and degraded that they cannot be allowed to sit in public coaches occupied by white citizens? That, as all will admit, is the real meaning of such legislation as was enacted in Louisiana.

The sure guaranty of the peace and security of each race is the clear, distinct, unconditional recognition by our governments, national and state, of every right that inheres in civil freedom, and of the equality before the law of all citizens of the United States, without regard to race. State enactments regulating the enjoyment of civil rights upon the basis of race, and cunningly devised to defeat legitimate results of the*561 war, under the pretense of recognizing equality of rights, can have no other result than to render permanent peace impossible, and to keep alive a conflict of races, the continuance of which must do harm to all concerned. This question is not met by the suggestion that social equality cannot exist between the white and black races in this country. That argument, if it can be properly regarded as one, is scarcely worthy of consideration; for social equality no more exists between two races when traveling in a passenger coach or a public highway than when members of the same races sit by each other in a street car or in the jury box, or stand or sit with each other in a political assembly, or when they use in common the streets of a city or town, or when they are in the same room for the purpose of having their names placed on the registry of voters, or when they approach the ballot box in order to exercise the high privilege of voting.

There is a race so different from our own that we do not permit those belonging to it to become citizens of the United States. Persons belonging to it are, with few exceptions, absolutely excluded from our country. I allude to the Chinese race. But, by the statute in question, a Chinaman can ride in the same passenger coach with white citizens of the United States, while citizens of the black race in Louisiana, many of whom, perhaps, risked their lives for the preservation of the Union, who are entitled, by law, to participate in the political control of the state and nation, who are not excluded, by law or by reason of their race, from public stations of any kind, and who have all the legal rights that belong to white citizens, are yet declared to be criminals, liable to imprisonment, if they ride in a public coach occupied by citizens of the white race. It is scarcely just to say that a colored citizen should not object to occupying a public coach assigned to his own race. He does not object, nor, perhaps, would he object to separate coaches for his race if his rights under the law were recognized. But he does object, and he ought never to cease objecting, that citizens of the white and black races can be adjudged criminals because they sit, or claim the right to sit, in the same public coach on a public highway.

*562The arbitrary separation of citizens, on the basis of race, while they are on a public highway, is a badge of servitude wholly inconsistent with the civil freedom and the equality before the law established by the constitution. It cannot be justified upon any legal grounds.

If evils will result from the commingling of the two races upon public highways established for the benefit of all, they will be infinitely less than those that will surely come from state legislation regulating the enjoyment of civil rights upon the basis of race. We boast of the freedom enjoyed by our people above all other peoples. But it is difficult to reconcile that boast with a state of the law which, practically, puts the brand of servitude and degradation upon a large class of our fellow citizens,—our equals before the law. The thin disguise of 'equal' accommodations for passengers in railroad coaches will not mislead any one, nor atone for the wrong this day done.

The result of the whole matter is that while this court has frequently adjudged, and at the present term has recognized the doctrine, that a state cannot, consistently with the constitution of the United States, prevent white and black citizens, having the required qualifications for jury service, from sitting in the same jury box, it is now solemnly held that a state may prohibit white and black citizens from sitting in the same passenger coach on a public highway, or may require that they be separated by a 'partition' when in the same passenger coach. May it not now be reasonably expected that astute men of the dominant race, who affect to be disturbed at the possibility that the integrity of the white race may be corrupted, or that its supremacy will be imperiled, by contact on public highways with black people, will endeavor to procure statutes requiring white and black jurors to be separated in the jury box by a 'partition,' and that, upon retiring from the court room to consult as to their verdict, such partition, if it be a movable one, shall be taken to their consultation room, and set up in such way as to prevent black jurors from coming too close to their brother jurors of the white race. If the 'partition' used in the court room happens to be stationary, provision could be made for screens with openings through*563 which jurors of the two races could confer as to their verdict without coming into personal contact with each other. I cannot see but that, according to the principles this day announced, such state legislation, although conceived in hostility to, and enacted for the purpose of humiliating, citizens of the United States of a particular race, would be held to be consistent with the constitution.

I do not deem it necessary to review the decisions of state courts to which reference was made in argument. Some, and the most important, of them, are wholly inapplicable, because rendered prior to the adoption of the last amendments of the constitution, when colored people had very few rights which the dominant race felt obliged to respect. Others were made at a time when public opinion, in many localities, was dominated by the institution of slavery; when it would not have been safe to do justice to the black man; and when, so far as the rights of blacks were concerned, race prejudice was, practically, the supreme law of the land. Those decisions cannot be guides in the era introduced by the recent amendments of the supreme law, which established universal civil freedom, gave citizenship to all born or naturalized in the United States, and residing ere, obliterated the race line from our systems of governments, national and state, and placed our free institutions upon the broad and sure foundation of the equality of all men before the law.

I am of opinion that the state of Louisiana is inconsistent with the personal liberty of citizens, white and black, in that state, and hostile to both the spirit and letter of the constitution of the United States. If laws of like character should be enacted in the several states of the Union, the effect would be in the highest degree mischievous. Slavery, as an institution tolerated by law, would, it is true, have disappeared from our country; but there would remain a power in the states, by sinister legislation, to interfere with the full enjoyment of the blessings of freedom, to regulate civil rights, common to all citizens, upon the basis of race, and to place in a condition of legal inferiority a large body of American citizens, now constituting a part of the political community, called the*564 'People of the United States,' for whom, and by whom through representatives, our government is administered. Such a system is inconsistent with the guaranty given by the constitution to each state of a republican form of government, and may be stricken down by congressional action, or by the courts in the discharge of their solemn duty to maintain the supreme law of the land, anything in the constitution or laws of any state to the contrary notwithstanding.

For the reason stated, I am constrained to withhold my assent from the opinion and judgment of the majority.

Brown et al. v. Board of Education of Topeka, Shawnee County, KN., et al. Briggs et al. v. Elliott et al. Davis et al. v. County School Board of Prince Edward County, VA., et al. Gebhart et al. v. Belton et al.

Reargued Dec. 7, 8, 9, 1953.

Decided May 17, 1954.

[Syllabus from pages 483-484 intentionally omitted]*484 No. 1:

Mr. Robert L. Carter, New York City, for appellants Brown and others.

Mr. Paul E. Wilson, Topeka, Kan., for appellees Board of Education of Topeka and others.

Nos. 2, 4:

Messrs. Spottswood Robinson III, Thurgood Marshall, New York City, for appellants Briggs and Davis and others.

Messrs. John W. Davis,*485 T. Justin Moore, J. Lindsay Almond, Jr., Richmond, Va., for appellees Elliott and County School Board of Prince Edward County and others.

Asst. Atty. Gen. J. Lee Rankin for United States amicus curiae by special leave of Court.

No. 10:

Mr. H. Albert Young, Wilmington, Del., for petitioners Gebhart et al.

Mr. Jack Greenberg, Thurgood Marshall, New York City, for respondents Belton et al.

*486Mr. Chief Justice WARREN delivered the opinion of the Court.

These cases come to us from the States of Kansas, South Carolina, Virginia, and Delaware. They are premised on different facts and different local conditions, but a common legal question justifies their consideration together in this consolidated opinion.1*487 In each of the cases, minors of the Negro race, through their legal representatives, seek the aid of the courts in obtaining admission to the public schools of their community on a nonsegregated basis. In each instance,*488 they have been denied admission to schools attended by white children under laws requiring or permitting segregation according to race. This segregation was alleged to deprive the plaintiffs of the equal protection of the laws under the Fourteenth Amendment. In each of the cases other than the Delaware case, a three-judge federal district court denied relief to the plaintiffs on the so-called 'separate but equal' doctrine announced by this Court in Plessy v. Ferguson, 163 U.S. 537, 16 S.Ct. 1138, 41 L.Ed. 256. Under that doctrine, equality of treatment is accorded when the races are provided substantially equal facilities, even though these facilities be separate. In the Delaware case, the Supreme Court of Delaware adhered to that doctrine, but ordered that the plaintiffs be admitted to the white schools because of their superiority to the Negro schools.

The plaintiffs contend that segregated public schools are not 'equal' and cannot be made 'equal,' and that hence they are deprived of the equal protection of the laws. Because of the obvious importance of the question presented, the Court took jurisdiction.2 Argument was heard in the 1952 Term, and reargument was heard this Term on certain questions propounded by the Court.3*489 Reargument was largely devoted to the circumstances surrounding the adoption of the Fourteenth Amendment in 1868. It covered exhaustively consideration of the Amendment in Congress, ratification by the states, then existing practices in racial segregation, and the views of proponents and opponents of the Amendment. This discussion and our own investigation convince us that, although these sources cast some light, it is not enough to resolve the problem with which we are faced. At best, they are inconclusive. The most avid proponents of the post-War Amendments undoubtedly intended them to remove all legal distinctions among 'all persons born 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. What others in Congress and the state legislatures had in mind cannot be determined with any degree of certainty.

An additional reason for the inconclusive nature of the Amendment's history, with respect to segregated schools, is the status of public education at that time.4 In the South, the movement toward free common schools, sup*490ported by general taxation, had not yet taken hold. Education of white children was largely in the hands of private groups. Education of Negroes was almost nonexistent, and practically all of the race were illiterate. In fact, any education of Negroes was forbidden by law in some states. Today, in contrast, many Negroes have achieved outstanding success in the arts and sciences as well as in the business and professional world. It is true that public school education at the time of the Amendment had advanced further in the North, but the effect of the Amendment on Northern States was generally ignored in the congressional debates. Even in the North, the conditions of public education did not approximate those existing today. The curriculum was usually rudimentary; ungraded schools were common in rural areas; the school term was but three months a year in many states; and compulsory school attendance was virtually unknown. As a consequence, it is not surprising that there should be so little in the history of the Fourteenth Amendment relating to its intended effect on public education.

In the first cases in this Court construing the Fourteenth Amendment, decided shortly after its adoption, the Court interpreted it as proscribing all state-imposed discriminations against the Negro race.5 The doctrine of*491   "separate but equal" did not make its appearance in this court until 1896 in the case of Plessy v. Ferguson, supra, involving not education but transportation.6 American courts have since labored with the doctrine for over half a century. In this Court, there have been six cases involving the 'separate but equal' doctrine in the field of public education.7 In Cumming v. Board of Education of Richmond County, 175 U.S. 528, 20 S.Ct. 197, 44 L.Ed. 262, and Gong Lum v. Rice, 275 U.S. 78, 48 S.Ct. 91, 72 L.Ed. 172, the validity of the doctrine itself was not challenged. 8 In more recent cases, all on the graduate school*492 level, inequality was found in that specific benefits enjoyed by white students were denied to Negro students of the same educational qualifications. State of Missouri ex rel. Gaines v. Canada, 305 U.S. 337, 59 S.Ct. 232, 83 L.Ed. 208; Sipuel v. Board of Regents of University of Oklahoma, 332 U.S. 631, 68 S.Ct. 299, 92 L.Ed. 247; Sweatt v. Painter, 339 U.S. 629, 70 s.Ct. 848, 94 L.Ed. 1114; McLaurin v. Oklahoma State Regents, 339 U.S. 637, 70 S.Ct. 851, 94 L.Ed. 1149. In none of these cases was it necessary to re-examine the doctrine to grant relief to the Negro plaintiff. And in Sweatt v. Painter, supra, the Court expressly reserved decision on the question whether Plessy v. Ferguson should be held inapplicable to public education.

In the instant cases, that question is directly presented. Here, unlike Sweatt v. Painter, there are findings below that the Negro and white schools involved have been equalized, or are being equalized, with respect to buildings, curricula, qualifications and salaries of teachers, and other 'tangible' factors. 9 Our decision, therefore, cannot turn on merely a comparison of these tangible factors in the Negro and white schools involved in each of the cases. We must look instead to the effect of segregation itself on public education.

In approaching this problem, we cannot turn the clock back to 1868 when the Amendment was adopted, or even to 1896 when Plessy v. Ferguson was written. We must consider public education in the light of its full development and its present place in American life throughout*493 the Nation. Only in this way can it be determined if segregation in public schools deprives these plaintiffs of the equal protection of the laws.

Today, education is perhaps the most important function of state and local governments. Compulsory school attendance laws and the great expenditures for education both demonstrate our recognition of the importance of education to our democratic society. It is required in the performance of our most basic public responsibilities, even service in the armed forces. It is the very foundation of good citizenship. Today it is a principal instrument in awakening the child to cultural values, in preparing him for later professional training, and in helping him to adjust normally to his environment. In these days, it is doubtful that any child may reasonably be expected to succeed in life if he is denied the opportunity of an education. Such an opportunity, where the state has undertaken to provide it, is a right which must be made available to all on equal terms.

We come then to the question presented: Does segregation of children in public schools solely on the basis of race, even though the physical facilities and other 'tangible' factors may be equal, deprive the children of the minority group of equal educational opportunities? We believe that it does.

In Sweatt v. Painter, supra (339 U.S. 629, 70 S.Ct. 850), in finding that a segregated law school for Negroes could not provide them equal educational opportunities, this Court relied in large part on 'those qualities which are incapable of objective measurement but which make for greatness in a law school.' In McLaurin v. Oklahoma State Regents, supra (339 U.S. 637, 70 S.Ct. 853), the Court, in requiring that a Negro admitted to a white graduate school be treated like all other students, again resorted to intangible considerations: '* * * his ability to study, to engage in discussions and exchange views with other students, and, in general, to learn his profession.'*494 Such considerations apply with added force to children in grade and high schools. To separate them from others of similar age and qualifications solely because of their race generates a feeling of inferiority as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone. The effect of this separation on their educational opportunities was well stated by a finding in the Kansas case by a court which nevertheless felt compelled to rule against the Negro plaintiffs:

'Segregation of white and colored children in public schools has a detrimental effect upon the colored children. The impact is greater when it has the sanction of the law; for the policy of separating the races is usually interpreted as denoting the inferiority of the negro group. A sense of inferiority affects the motivation of a child to learn. Segregation with the sanction of law, therefore, has a tendency to (retard) the educational and mental development of Negro children and to deprive them of some of the benefits they would receive in a racial(ly) integrated school system.' 10

Whatever may have been the extent of psychological knowledge at the time of Plessy v. Ferguson, this finding is amply supported by modern authority.11 Any lan*495guage in Plessy v. Ferguson contrary to this finding is rejected.

We conclude that in the field of public education the doctrine of 'separate but equal' has no place. Separate educational facilities are inherently unequal. Therefore, we hold that the plaintiffs and others similarly situated for whom the actions have been brought are, by reason of the segregation complained of, deprived of the equal protection of the laws guaranteed by the Fourteenth Amendment. This disposition makes unnecessary any discussion whether such segregation also violates the Due Process Clause of the Fourteenth Amendment. 12

Because these are class actions, because of the wide applicability of this decision, and because of the great variety of local conditions, the formulation of decrees in these cases presents problems of considerable complexity. On reargument, the consideration of appropriate relief was necessarily subordinated to the primary question—the constitutionality of segregation in public education. We have now announced that such segregation is a denial of the equal protection of the laws. In order that we may have the full assistance of the parties in formulating decrees, the cases will be restored to the docket, and the parties are requested to present further argument on Questions 4 and 5 previously propounded by the Court for the reargument this Term.13 The Attorney General*496 of the United States is again invited to participate. The Attorneys General of the states requiring or permitting segregation in public education will also be permitted to appear as amici curiae upon request to do so by September 15, 1954, and submission of briefs by October 1, 1954.14

It is so ordered.

Cases ordered restored to docket for further argument on question of appropriate decrees.

1

In the Kansas case, Brown v. Board of Education, the plaintiffs are Negro children of elementary school age residing in Topeka. They brought this action in the United States District Court for the District of Kansas to enjoin enforcement of a Kansas statute which permits, but does not require, cities of more than 15,000 population to maintain separate school facilities for Negro and white students. Kan.Gen.Stat.1949, § 72—1724. Pursuant to that authority, the Topeka Board of Education elected to establish segregated elementary schools. Other public schools in the community, however, are operated on a nonsegregated basis. The three-judge District Court, convened under 28 U.S.C. §§ 2281 and 2284, 28 U.S.C.A. §§ 2281, 2284, found that segregation in public education has a detrimental effect upon Negro children, but denied relief on the ground that the Negro and white schools were substantially equal with respect to buildings, transportation, curricula, and educational qualifications of teachers. 98 F.Supp. 797. The case is here on direct appeal under 28 U.S.C. § 1253, 28 U.S.C.A. § 1253.

In the South Carolina case, Briggs v. Elliott, the plaintiffs are Negro children of both elementary and high school age residing in Clarendon County. They brought this action in the United States District Court for the Eastern District of South Carolina to enjoin enforcement of provisions in the state constitution and statutory code which require the segregation of Negroes and whites in public schools. S.C.Const. Art. XI, § 7; S.C.Code 1942, § 5377. The three-judge District Court, convened under 28 U.S.C. §§ 2281 and 2284, 28 U.S.C.A. §§ 2281, 2284, denied the requested relief. The court found that the Negro schools were inferior to the white schools and ordered the defendants to begin immediately to equalize the facilities. But the court sustained the validity of the contested provisions and denied the plaintiffs admis-

sion to the white schools during the equalization program. 98 F.Supp. 529. This Court vacated the District Court's judgment and remanded the case for the purpose of obtaining the court's views on a report filed by the defendants concerning the progress made in the equalization program. 342 U.S. 350, 72 S.Ct. 327, 96 L.Ed. 392. On remand, the District Court found that substantial equality had been achieved except for buildings and that the defendants were proceeding to rectify this inequality as well. 103 F.Supp. 920. The case is again here on direct appeal under 28 U.S.C. § 1253, 28 U.S.C.A. § 1253.

In the Virginia case, Davis v. County School Board, the plaintiffs are Negro children of high school age residing in Prince Edward County. They brought this action in the United States District Court for the Eastern District of Virginia to enjoin enforcement of provisions in the state constitution and statutory code which require the segregation of Negroes and whites in public schools. Va.Const. § 140; Va.Code 1950, § 22—221. The three-judge District Court, convened under 28 U.S.C. §§ 2281 and 2284, 28 U.S.C.A. §§ 2281, 2284, denied the requested relief. The court found the Negro school inferior in physical plant, curricula, and transportation, and ordered the defendants forthwith to provide substantially equal curricula and transportation and to 'proceed with all reasonable diligence and dispatch to remove' the inequality in physical plant. But, as in the South Carolina case, the court sustained the validity of the contested provisions and denied the plaintiffs admission to the white schools during the equalization program. 103 F.Supp. 337. The case is here on direct appeal under 28 U.S.C. § 1253, 28 U.S.C.A. § 1253.

In the Delaware case, Gebhart v. Belton, the plaintiffs are Negro children of both elementary and high school age residing in New Castle County. They brought this action in the Delaware Court of Chancery to enjoin enforcement of provisions in the state constitution and statutory code which require the segregation of Negroes and whites in public schools. Del.Const. Art. X, § 2; Del.Rev.Code, 1935, § 2631, 14 Del.C. § 141. The Chancellor gave judgment for the plaintiffs and ordered their immediate admission to schools previously attended only by white children, on the ground that the Negro schools were inferior with respect to teacher training, pupil-teacher ratio, extracurricular activities, physical plant, and time and distance involved in travel. Del.Ch., 87 A.2d 862. The Chancellor also found that segregation itself results in an inferior education for Negro children (see note 10, infra), but did not rest his decision on that ground. 87 A.2d at page 865. The Chancellor's decree was affirmed by the Supreme Court of Delaware, which intimated, however, that the defendants might be able to obtain a modification of the decree after equalization of the Negro and white schools had been accomplished. 91 A.2d 137, 152. The defendants, contending only that the Delaware courts had erred in ordering the immediate admission of the Negro plaintiffs to the white schools, applied to this Court for certiorari. The writ was granted, 344 U.S. 891, 73 S.Ct. 213, 97 L.Ed. 689. The plaintiffs, who were successful below, did not submit a cross-petition.

2

344 U.S. 1, 73 S.Ct. 1, 97 L.Ed. 3, Id., 344 U.S. 141, 73 S.Ct. 124, 97 L.Ed. 152, Gebhart v. Belton, 344 U.S. 891, 73 S.Ct. 213, 97 L.Ed. 689.

3

345 U.S. 972, 73 S.Ct. 1118, 97 L.Ed. 1388. The Attorney General of the United States participated both Terms as amicus curiae.

4

For a general study of the development of public education prior to the Amendment, see Butts and Cremin, A History of Education in American Culture (1953), Pts. I, II: Cubberley, Public Education in the United States (1934 ed.), cc. II—XII. School practices current at the time of the adoption of the Fourteenth Amendment are described in Butts and Cremin, supra, at 269—275; Cubberley, supra, at 288—339, 408—431; Knight, Public Education in the South (1922), cc. VIII, IX. See also H. Ex. Doc. No. 315, 41st Cong., 2d Sess. (1871). Although the demand for free public schools followed substantially the same pattern in both the North and the South, the development in the South did not begin to gain momentum until about 1850, some twenty years after that in the North. The reasons for the somewhat slower development in the South (e.g., the rural character of the South and the different regional attitudes toward state assistance) are well explained in Cubberley, supra, at 408—423. In the country as a whole, but particularly in the South, the War virtually stopped all progress in public education. Id., at 427—428. The low status of Negro education in all sections of the country, both before and immediately after the War, is described in Beale, A History of Freedom of Teaching in American Schools (1941), 112—132, 175—195. Compulsory school attendance laws were not generally adopted until after the ratification of the Fourteenth Amendment, and it was not until 1918 that such laws were in force in all the states. Cubberley, supra, at 563—565.

5

In re Slaughter-House Cases, 1873, 16 Wall. 36, 67—72, 21 L.Ed. 394; Strauder v. West Virginia, 1880, 100 U.S. 303, 307—308, 25 L.Ed. 664.

'It ordains that no State shall deprive any person of life, liberty, or property, without due process of law, or deny to any person within its jurisdiction the equal protection of the laws. What is this but declaring that the law in the States shall be the same for the black as for the white; that all persons, whether colored or white, shall stand equal before the laws of the States, and, in regard to the colored race, for whose protection the amendment was primarily designed, that no discrimination shall be made against them by law because of their color? The words of the amendment, it is true, are prohibitory, but they contain a necessary implication of a positive immunity, or right, most valuable to the colored race,—the right to exemption from unfriendly legislation against them distinctively as colored, exemption from legal discriminations, implying inferiority in civil society, lessening the security of their enjoyment of the rights which others enjoy, and discriminations which are steps towards reducing them to the condition of a subject race.'

See also State of Virginia v. Rives, 1879, 100 U.S. 313, 318, 25 L.Ed. 667; Ex parte Virginia, 1879, 100 U.S. 339, 344—345, 25 L.Ed. 676.

6

The doctrine apparently originated in Roberts v. City of Boston, 1850, 5 Cush. 198, 59 Mass. 198, 206, upholding school segregation against attack as being violative of a state constitutional guarantee of equality. Segregation in Boston public schools was eliminated in 1855. Mass. Acts 1855, c. 256. But elsewhere in the North segregation in public education has persisted in some communities until recent years. It is apparent that such segregation has long been a nationwide problem, not merely one of sectional concern.

7

See also Berea College v. Kentucky, 1908, 211 U.S. 45, 29 S.Ct. 33, 53 L.Ed. 81.

8

In the Cumming case, Negro taxpayers sought an injunction requiring the defendant school board to discontinue the operation of a high school for white children until the board resumed operation of a high school for Negro children. Similarly, in the Gong Lum case, the plaintiff, a child of Chinese descent, contended only that state authorities had misapplied the doctrine by classifying him with Negro children and requiring him to attend a Negro school.

9

In the Kansas case, the court below found substantial equality as to all such factors. 98 F.Supp. 797, 798. In the South Carolina case, the court below found that the defendants were proceeding 'promptly and in good faith to comply with the court's decree.' 103 F.Supp. 920, 921. In the Virginia case, the court below noted that the equalization program was already 'afoot and progressing,' 103 F.Supp. 337, 341; since then, we have been advised, in the Virginia Attorney General's brief on reargument, that the program has now been completed. In the Delaware case, the court below similarly noted that the state's equalization program was well under way. 91 A.2d 137, 139.

10

A similar finding was made in the Delaware case: 'I conclude from the testimony that in our Delaware society, State-imposed segregation in education itself results in the Negro children, as a class, receiving educational opportunities which are substantially inferior to those available to white children otherwise similarly situated.' 87 A.2d 862, 865.

11

K. B. Clark, Effect of Prejudice and Discrimination on Personality Development (Midcentury White House Conference on Children and Youth, 1950); Witmer and Kotinsky, Personality in the Making (1952), c. VI; Deutscher and Chein, The Psychological Effects of Enforced Segregation: A Survey of Social Science Opinion, 26 J.Psychol. 259 (1948); Chein, What are the Psychological Effects of Segregation Under Conditions of Equal Facilities?, 3 Int. J. Opinion and Attitude Res. 229 (1949); Brameld, Educational Costs, in Discrimination and National Welfare (MacIver, ed., 1949), 44—48; Frazier, The Negro in the United States (1949), 674—681. And see generally Myrdal, An American Dilemma (1944).

12

See Bolling v. Sharpe, 347 U.S. 497, 74 S.Ct. 693, concerning the Due Process Clause of the Fifth Amendment.

13

'4. Assuming it is decided that segregation in public schools violates the Fourteenth Amendment

'(a) would a decree necessarily follow providing that, within the limits set by normal geographic school districting, Negro children should forthwith be admitted to schools of their choice, or

'(b) may this Court, in the exercise of its equity powers, permit an effective gradual adjustment to be brought about from existing segregated systems to a system not based on color distinctions?

'5. On the assumption on which questions 4(a) and (b) are based, and assuming further that this Court will exercise its equity powers to the end described in question 4(b),

'(a) should this Court formulate detailed decrees in these cases;

'(b) if so, what specific issues should the decrees reach;

'(c) should this Court appoint a special master to hear evidence with a view to recommending specific terms for such decrees;

'(d) should this Court remand to the courts of first instance with directions to frame decrees in these cases, and if so what general directions should the decrees of this Court include and what procedures should the courts of first instance follow in arriving at the specific terms of more detailed decrees?'

14

See Rule 42, Revised Rules of this Court, effective July 1, 1954, 28 U.S.C.A.

Richard Perry Loving et ux., Appellants, v. Commonwealth of Virginia.

Argued April 10, 1967.

Decided June 12, 1967.

Philip J. Hirschkop, pro hac vice, by special leave of Court, Bernard S. Cohen, Alexandria, Va., for appellants.

R. D. McIlwaine, III, Richmond, Va., for appellee.

William M. Marutani, Philadelphia, Pa., for Japanese American Citizens League, as amicus curiae, by special leave of Court.

*2Mr. 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*3 of 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*4 modifying the sentence, affirmed the convictions.2 The Lovings appealed this decision, and we noted probable jurisdiction on December 12, 1966, 385 U.S. 986, 87 S.Ct. 595, 17 L.Ed.2d 448.

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,*5 respectively, 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*7 the 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, 8 S.Ct. 723, 31 L.Ed. 654 (1888), the State does not contend in its argument before this Court that its powers to regulate marriage are unlimited notwithstanding the commands of the Fourteenth Amendment. Nor could it do so in light of Meyer v. State of Nebraska, 262 U.S. 390, 43 S.Ct. 625, 67 L.Ed. 1042 (1923), and Skinner v. State of Oklahoma, 316 U.S. 535, 62 S.Ct. 1110, 86 L.Ed. 1655 (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*8 as 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. People of State of New York, 336 U.S. 106, 6 § .Ct. 463, 93 L.Ed. 533 (1949), or an exemption in Ohio's ad valorem tax for merchandise owned by a non-resident in a storage warehouse, Allied Stores of Ohio,*9 Inc. v. Bowers, 358 U.S. 522, 79 S.Ct. 437, 3 L.Ed.2d 480 (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 the 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 born 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 of Topeka, 347 U.S. 483, 489, 74 S.Ct. 686, 689, 98 L.Ed. 873 (1954). See also Strauder*10 v. State of West Virginia, 100 U.S. 303, 310, 25 L.Ed. 664 (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. State of Florida, 379 U.S. 184, 85 S.Ct. 283, 13 L.Ed.2d 222 (1964).

The State finds support for its 'equal application' theory in the decision of the Court in Pace v. State of Alabama, 106 U.S. 583, 1 S.Ct. 637, 27 L.Ed. 207 (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, 379 U.S. at 188, 85 S.Ct. at 286. 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, 21 L.Ed. 394 (1873); Strauder v. State of West Virginia, 100 U.S. 303, 307—308 2 5 L.Ed. 664 (1880); Ex parte Virginia, 100 U.S. 339, 344—345, 26 L.Ed. 676 (1880); Shelley v. Kraemer, 334 U.S. 1, 68 S.Ct. 836, 92 L.Ed. 1161 (1948); Burton v. Wilmington Parking Authority, 365 U.S. 715, 81 S.Ct. 856, 6 L.Ed.2d 45 (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 '(d)istinctions 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, 63 S.Ct. 1375, 1385, 87 L.Ed. 1774 (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, 65 S.Ct. 193, 194, 89 L.Ed. 194 (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, 379 U.S. at 198, 85 S.Ct. at 292, (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*12 the 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. State of Oklahoma, 316 U.S. 535, 541, 62 S.Ct. 1110, 1113, 86 L.Ed. 1655 (1942). See also Maynard v. Hill, 125 U.S. 190, 8 S.Ct. 723, 31 L.Ed. 654 (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.

Reversed.

*13 

1

1. Section 1 of the Fourteenth Amendment provides:

'All persons born or naturalized in the United States and subject to the jurisdiction thereof, are citizens of the United States and of the State wherein they reside. No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.'

2

206 Va. 924, 147 S.E.2d 78 (1966).

3

Section 20—57 of the Virginia Code provides:

'Marriages void without decree.—All marriages between a white person and a colored person shall be absolutely void without any decree of divorce or other legal process.' Va.Code Ann. § 20—57 (1960 Repl.Vol.).

4

Section 20—54 of the Virginia Code provides:

'Intermarriage prohibited; meaning of term 'white persons.' It shall hereafter be unlawful for any white person in this State to marry any save a white person, or a person with no other admixture of blood than white and American Indian. For the purpose of this chapter, the term 'white person' shall apply only to such person as has no trace whatever of any blood other than Caucasian; but persons who have one-sixteenth or less of the blood of the American Indian and have no other non-Caucasic blood shall be deemed to be white persons. All laws heretofore passed and now in effect regarding the intermarriage of white and colored persons shall apply to marriages prohibited by this chaper.' Va.Code Ann. § 20—54 (1960 Repl.Vol.).

The exception for persons with less than one-sixteenth 'of the blood of the American Indian' is apparently accounted for, in the words of a tract issued by the Registrar of the State Bureau of Vital Statistics, by 'the desire of all to recognize as an integral and honored part of the white race the descendants of John Rolfe and Pocahontas * * *.' Plecker, The New Family and Race Improvement, 17 Va.Health Bull., Extra No. 12, at 25—26 (New Family Series No. 5, 1925), cited in Wadlington, The Loving Case; Virginia's Anti-Miscegenation Statute in Historical Perspective, 52 Va.L.Rev. 1189, 1202, n. 93 (1966).

Section 1—14 of the Virginia Code provides:

Colored persons and Indians defined.—Every person in whom there is ascertainable any Negro blood shall be deemed and taken to be a colored person, and every person not a colored person having one fourth or more of American Indian blood shall be deemed an American Indian; except that members of Indian tribes existing in this Commonwealth having one fourth or more of Indian blood and less than one sixteenth of Negro blood shall be deemed tribal Indians.' Va.Code Ann. § 1—14 (1960 Repl.Vol.).

5

After the initiation of this litigation, Maryland repealed its prohibitions against interracial marriage, Md.Laws 1967, c. 6, leaving Virginia and 15 other States with statutes outlawing interracial marriage: Alabama, Ala.Const., Art. 4, § 102, Ala.Code, Tit. 14, § 360 (1958); Arkansas, Ark.Stat.Ann. § 55—104 (1947); Delaware, Del.Code Ann., Tit. 13, § 101 (1953); Florida, Fla.Const., Art. 16, § 24, F.S.A., Fla.Stat. § 741.11 (1965) F.S.A.; Georgia, Ga.Code Ann. § 53—106 (1961); Kentucky, Ky.Rev.Stat.Ann. § 402.020 (Supp.1966); Louisiana, La.Rev.Stat. § 14:79 (1950); Mississippi, Miss.Const., Art. 14, § 263, Miss.Code Ann. § 459 (1956); Missouri, Mo.Rev.Stat. § 451.020 (Supp.1966), V.A.M.S.; North Carolina, N.C.Const., Art. XIV, § 8, N.C.Gen.Stat. § 14—181 (1953); Oklahoma, Okla.Stat., Tit. 43, § 12 (Supp.1965); South Carolina, S.C.Const., Art. 3, § 33, S.C.Code Ann. § 20—7 (1962); Tennessee, Tenn.Const., Art. 11, § 14, Tenn.Code Ann. § 36 402 (1955); Vernon's Ann.Texas, Tex.Pen.Code, Art. 492 (1952); West Virginia, W.Va.Code Ann. § 4697 (1961).

Over the past 15 years, 14 States have repealed laws outlawing interracial marriages: Arizona, California, Colorado, Idaho, Indiana, Maryland, Montana, Nebraska, Nevada, North Dakota, Oregon, South Dakota, Utah, and Wyoming.

The first state court to recognize that miscegenation statutes violate the Equal Protection Clause was the Supreme Court of California. Perez v. Sharp, 32 Cal.2d 711, 198 P.2d 17 (1948).

6

For a historical discussion of Virginia's miscegenation statutes, see Wadlington, supra, n. 4.

7

Va.Code Ann. § 20—54 (1960 Repl.Vol.).

8

Va.Code Ann. § 20—53 (1960 Repl.Vol.).

9

Va.Code Ann. § 20—50 (1960 Repl.Vol.).

10

a.Code Ann. § 20—54 (1960 Repl.Vol.).

11

Appellants point out that the State's concern in these statutes, as expressed in the words of the 1924 Act's title, 'An Act to Preserve Racial Integrity,' extends only to the integrity of the white race. While Virginia prohibits whites from marrying any nonwhite (subject to the exception for the descendants of Pocahontas), Negroes, Orientals, and any other racial class may intermarry without statutory interference. Appellants contend that this distinction renders Virginia's miscegenation statutes arbitrary and unreasonable even assuming the constitutional validity of an official purpose to preserve 'racial integrity.' We need not reach this contention because we find the racial classifications in these statutes repugnant to the Fourteenth Amendment, even assuming an even-handed state purpose to protect the 'integrity' of all races.

Mr. 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. State of Florida, 379 U.S. 184, 198, 85 S.Ct. 283, 292, 13 L.Ed.2d 222 (concurring opinion). Because I adhere to that belief, I concur in the judgment of the Court.

JOHNSON v. CALIFORNIA et al.

No. 03-636.

Argued November 2, 2004

Decided February 23, 2005

O’Connor, J., delivered the opinion of the Court, in which Kennedy, Souter, Ginsburg, and Breyer, JJ., joined. Ginsburg, J., filed a concurring opinion, in which Souter and Breyer, JJ., joined, post, p. 516. Stevens, J., filed a dissenting opinion, post, p. 517. Thomas, J., filed a dissenting opinion, in which SCALIA, J., joined, post, p. 524. REHNQUIST, C. J., took no part in the decision of the case.

Bert H. Deixler argued the cause for petitioner. With him on the briefs were Charles S. Sims, Lois D. Thompson, and Tanya L. Forsheit.

Acting Solicitor General Clement argued the cause for the United States as amicus curiae urging reversal. With him on the brief were former Solicitor General Olson, Assistant Attorney General Acosta, David B. Salmons, David K. Flynn, and Tovah R. Calderon.

Frances T Grander, Senior Assistant Attorney General of California, argued the cause for respondents. With her on the brief were Bill Lockyer, Attorney General, Manuel M. Medeiros, Solicitor General, Robert R. Anderson, Chief Assistant Attorney General, and Sara Turner, Supervising Deputy Attorney General.*

*

Briefs of amici curiae urging reversal were filed for the American Civil Liberties Union et al. by Elizabeth Alexander, David C. Fathi, Steven R. Shapiro, Jordan C. Budd, Alan Schlosser, and Mark D. Rosen-baum; and for Former State Corrections Officials by Michael C. Small.

Briefs of amici curiae urging affirmance were filed for the State of Utah et al. by Mark L. Shurtleff, Attorney General of Utah, and Gene C. Schaerr, and by the Attorneys General for their respective States as follows: Troy King of Alabama, Gregg D. Renkes of Alaska, M. Jane Brady of Delaware, Lawrence G. Wasden of Idaho, Brian Sandoval of Nevada, Kelly A. Ayotte of New Hampshire, and Wayne Stenehjem of North Dakota; and for the National Association of Black Law Enforcement Officers, Inc., by David T. Goldberg.

John H. Findley filed a brief for the Pacific Legal Foundation as ami-cus curiae.

Justice O’Connor

delivered the opinion of the Court.

The California Department of Corrections (CDC) has an unwritten policy of racially segregating prisoners in double cells in reception centers for up to 60 days each time they enter a new correctional facility. We consider whether strict scrutiny is the proper standard of review for an equal protection challenge to that policy.

I

A

CDC institutions house all new male inmates and all male inmates transferred from other state facilities in reception centers for up to 60 days upon their arrival. During that time, prison officials evaluate the inmates to determine their ultimate placement. Double-cell assignments in the reception centers are based on a number of factors, predominantly race. In fact, the CDC has admitted that the chances of an inmate being assigned a cellmate of another race are “ ‘[p]retty close’ ” to zero percent. App. to Pet. for Cert. 3a. The CDC further subdivides prisoners within each racial group. Thus, Japanese-Americans are housed separately from Chinese-Americans, and northern California Hispanics are separated from southern California Hispanics.

The CDC’s asserted rationale for this practice is that it is necessary to prevent violence caused by racial gangs. Brief for Respondents 1-6. It cites numerous incidents of racial violence in CDC facilities and identifies five major prison gangs in the State: Mexican Mafia, Nuestra Familia, Black Guerilla Family, Aryan Brotherhood, and Nazi Low Riders. Id., at 2. The CDC also notes that prison-gang culture is violent and murderous. Id., at 3. An associate warden testified that if race were not considered in making initial housing assignments, she is certain there would be racial conflict in the cells and in the yard. App. 215a. Other prison officials also expressed their belief that violence and conflict would result if prisoners were not segregated. See, e. g., id., at 305a-306a. The CDC claims that it must therefore segregate all inmates while it determines whether they pose a danger to others. See Brief for Respondents 29.

With the exception of the double cells in reception areas, the rest of the state prison facilities — dining areas, yards, and cells — are fully integrated. After the initial 60-day period^ prisoners are allowed to choose their own cellmates. The CDC usually grants inmate requests to be housed together, unless there are security reasons for denying them.

B

Garrison Johnson is an African-American inmate in the custody of the CDC. He has been incarcerated since 1987 and, during that time, has been housed at a number of California prison facilities. Fourth Amended Complaint 3, Record, Doc. No. 78. Upon his arrival at Folsom prison in 1987, and each time he was transferred to a new facility thereafter, Johnson was double-celled with another African-American inmate. See ibid.

Johnson filed a complaint pro se in the United States District Court for the Central District of California on February 24, 1995, alleging that the CDC’s reception-center housing policy violated his right to equal protection under the Fourteenth Amendment by assigning him cellmates on the basis of his race. He alleged that, from 1987 to 1991, former CDC Director James Rowland instituted and enforced an unconstitutional policy of housing inmates according to race. Second Amended Complaint 2-4, Record, Doc. No. 21. Johnson made the same allegations against former Director James Gomez for the period from 1991 until the filing of his complaint. Ibid. The District Court dismissed his complaint for failure to state a claim. The Court of Appeals for the Ninth Circuit reversed and remanded, holding that Johnson had stated a claim for racial discrimination in violation of the Equal Protection Clause of the Fourteenth Amendment. Johnson v. California, 207 F. 3d 650, 655 (2000).

On remand, Johnson was appointed counsel and granted leave to amend his complaint. On July 5, 2000, he filed his Fourth Amended Complaint. Record, Doc. No. 81. Johnson claimed that the CDC’s policy of racially segregating all inmates in reception-center cells violated his rights under the Equal Protection Clause. Johnson sought damages, alleging that former CDC Directors Rowland and Gomez, in their individual capacities, violated his constitutional rights by formulating and implementing the CDC’s housing policy. He also sought injunctive relief against former CDC Director Stephen Cambra.

Johnson has consistently challenged, and the CDC has consistently defended, the policy as a whole — as it relates to both new inmates and. inmates transferred from other facilities. Johnson was first segregated in 1987 as a new inmate when he entered the CDC facility at Folsom. Since 1987, he has been segregated each time he has been transferred to a new facility. Thus, he has been subject to the CDC’s policy both as a new inmate and as an inmate transferred from one facility to another.

After discovery, the parties moved for summary judgment. The District Court granted summary judgment to the defendants on grounds that they were entitled to qualified immunity because their conduct was not clearly unconstitutional. The Court of Appeals for the Ninth Circuit affirmed. 321 F. 3d 791 (2003). It held that the constitutionality of the CDC’s policy should be reviewed under the deferential standard we articulated in Turner v. Safley, 482 U. S. 78 (1987)—not strict scrutiny. 321 F. 3d, at 798-799. Applying Turner, it held that Johnson had the burden of refuting the “common-sense connection” between the policy and prison violence. 321 F. 3d, at 802. Though it believed this was a “close case,” id., at 798, the Court of Appeals concluded that the policy survived Turner’s deferential standard, 321 F. 3d, at 807.

The Court of Appeals denied Johnson’s petition for rehearing en banc. Judge Ferguson, joined by three others, dissented on grounds that “[t]he panel’s decision ignore[d] the Supreme Court’s repeated and unequivocal command that all racial classifications imposed by the government must be analyzed by a reviewing court under strict scrutiny, and fail[ed] to recognize that [the] Turner analysis is inapplicable in cases, such as this one, in which the right asserted is not inconsistent with legitimate penological objectives.” 336 F. 3d 1117 (2003) (internal quotation marks and citations omitted). We granted certiorari to decide which standard of review applies. 540 U. S. 1217 (2004).

II

A

We have held that “all racial classifications [imposed by government] . . . must be analyzed by a reviewing court under strict scrutiny.” Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 227 (1995) (emphasis added). Under strict scrutiny, the government has the burden of proving that racial classifications “are narrowly tailored measures that further compelling governmental interests.” Ibid. We have insisted on strict scrutiny in every context, even for so-called “benign” racial classifications, such as race-conscious university admissions policies, see Grutter v. Bollinger, 539 U. S. 306, 326 (2003), race-based preferences in government contracts, see Adarand, supra, at 226, and race-based districting intended to improve minority representation, see Shaw v. Reno, 509 U. S. 630, 650 (1993).

The reasons for strict scrutiny are familiar. Racial classifications raise special fears that they are motivated by an invidious purpose. Thus, we have admonished time and again that, “[ajbsent searching judicial inquiry into the justification for such race-based measures, there is simply no way of determining ... what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.” Richmond v. J. A. Croson Co., 488 U. S. 469, 493 (1989) (plurality opinion). We therefore apply strict scrutiny to all racial classifications to “‘smoke out’ illegitimate uses of race by assuring that [government] is pursuing a goal important enough to warrant use of a highly suspect tool.” Ibid.1

The CDC claims that its policy should be exempt from our categorical rule because it is “neutral” — that is, it “neither benefits nor burdens one group or individual more than any other group or individual.” Brief for Respondents 16. In other words, strict scrutiny should not apply because all prisoners are “equally” segregated. The CDC’s argument ignores our repeated command that “racial classifications receive close scrutiny even when they may be said to burden or benefit the races equally.” Shaw, supra, at 651. Indeed, we rejected the notion that separate can ever be equal — or “neutral” — 50 years ago in Brown v. Board of Education, 347 U. S. 483 (1954), and we refuse to resurrect it today. See also Powers v. Ohio, 499 U. S. 400, 410 (1991) (rejecting the argument that race-based peremptory challenges were permissible because they applied equally to white and black jurors and holding that “[i]t is axiomatic that racial classifications do not become legitimate on the assumption that all persons suffer them in equal degree”).

We have previously applied a heightened standard of review in evaluating racial segregation in prisons. In Lee v. Washington, 390 U. S. 333 (1968) (per curiam), we upheld a three-judge court’s decision striking down Alabama’s policy of segregation in its prisons. Id., at 333-334. Alabama had argued that desegregation would undermine prison security and discipline, id., at 334, but we rejected that contention. Three Justices concurred “to make explicit something that is left to be gathered only by implication from the Court’s opinion” — “that prison authorities have the right, acting in good faith and in particularized circumstances, to take into account racial tensions in maintaining security, discipline, and good order in prisons and jails.” Ibid, (emphasis added). The concurring Justices emphasized that they were “unwilling to assume that state or local prison authorities might mistakenly regard such an explicit pronouncement as evincing any dilution of this Court’s firm commitment to the Fourteenth Amendment’s prohibition of racial discrimination.” Ibid.

The need for strict scrutiny is no less important here, where prison officials cite racial violence as the reason for their policy. As we have recognized in the past, racial classifications “threaten to stigmatize individuals by reason of their membership in a racial group and to incite racial hostility.” Shaw, supra, at 643 (citing J. A. Croson Co., supra, at 493 (plurality opinion); emphasis added). Indeed, by insisting that inmates be housed only with other inmates of the same race, it is possible that prison officials will breed further hostility among prisoners and reinforce racial and ethnic divisions. By perpetuating the notion that race matters most, racial segregation of inmates “may exacerbate the very patterns of [violence that it is] said to counteract.” Shaw, supra, at 648; see also Trulson & Marquart, The Caged Melting Pot: Toward an Understanding of the Consequences of Desegregation in Prisons, 36 Law & Soc. Rev. 743, 774 (2002) (in a study of prison desegregation, finding that “over [10 years] the rate of violence between inmates segregated by race in double cells surpassed the rate among those racially integrated”). See also Brief for Former State Corrections Officials as Amici Curiae 19 (opinion of former corrections officials from six States that “racial integration of cells tends to diffuse racial tensions and thus diminish interracial violence” and that “a blanket policy of racial segregation of inmates is contrary to sound prison management”).

The CBC’s policy is unwritten. Although California claimed at oral argument that two other States follow a similar policy, see Tr. of Oral Arg. 30-31, this assertion was unsubstantiated, and we are unable to confirm or deny its accuracy.2 Virtually all other States and the Federal Government manage their prison systems without reliance on racial segregation. See Brief for United States as Amicus Curiae 24. Federal regulations governing the Federal Bureau of Prisons (BOP) expressly prohibit racial segregation. 28 CFR § 551.90 (2004) (“[BOP] staff shall not discriminate against inmates on the basis of race, religion, national origin, sex, disability, or political belief. This includes the making of administrative decisions and providing access to work, housing and programs”). The United States contends that racial integration actually “leads to less violence in BOP’s institutions and better prepares inmates for re-entry into society.” Brief for United States as Amicus Curiae 25. Indeed, the United States argues, based on its experience with the BOP, that it is possible to address “concerns of prison security through individualized consideration without the use of racial segregation, unless warranted as a necessary and temporary response to a race riot or other serious threat of race-related violence.” Id., at 24. As to transferees, in particular, whom the CDC has already evaluated at least once, it is not clear why more individualized determinations are not possible.

Because the CDC’s policy is an express racial classification, it is “immediately suspect.” Shaw, 509 U. S., at 642; see also Washington v. Seattle School Dist. No. 1, 458 U. S. 457, 485 (1982). We therefore hold that the Court of Appeals erred when it failed to apply strict scrutiny to the CDC’s policy and to require the CDC to demonstrate that its policy is narrowly tailored to serve a compelling state interest.

B

The CDC invites us to make an exception to the rule that strict scrutiny applies to all racial classifications, and instead to apply the deferential standard of review articulated in Turner v. Safley, 482 U. S. 78 (1987), because its segregation policy applies only in the prison context. We decline the invitation. In Turner, we considered a claim by Missouri prisoners that regulations restricting inmate marriages and inmate-to-inmate correspondence were unconstitutional. Id., at 81. We rejected the prisoners’ argument that the regulations should be subject to . strict scrutiny, asking instead whether the regulation that burdened the prisoners’ fundamental rights was “reasonably related” to “legitimate penological interests.” Id., at 89.

We have never applied Turner to racial classifications. Turner itself did not involve any racial classification, and it cast no doubt on Lee. We think this unsurprising, as we have applied Turner’s reasonable-relationship test only to rights that are “inconsistent with proper incarceration.” Overton v. Bazzetta, 539 U. S. 126, 131 (2003); see also Pell v. Procunier, 417 U. S. 817, 822 (1974) (“[A] prison inmate retains those First Amendment rights that are not inconsistent with his status as a prisoner or with the legitimate penological objectives of the corrections system”). This is because certain privileges and rights must necessarily be limited in the prison context. See O’Lone v. Estate of Shabazz, 482 U. S. 342, 348 (1987) (“ ‘[Ljawful incarceration brings about the necessary withdrawal or limitation of many privileges and rights, a retraction justified by the considerations underlying our penal system’ ” (quoting Price v. Johnston, 334 U. S. 266, 285 (1948))). Thus, for example, we have relied on Turner in addressing First Amendment challenges to prison regulations, including restrictions on freedom of association, Overton, supra; limits on inmate correspondence, Shaw v. Murphy, 532 U. S. 223 (2001); restrictions on inmates’ access to courts, Lewis v. Casey, 518 U. S. 343 (1996); restrictions on receipt of subscription publications, Thornburgh v. Abbott, 490 U. S. 401 (1989); and work rules limiting prisoners’ attendance at religious services, Shabazz, supra. We have also applied Turner to some due process claims, such as involuntary medication of mentally ill prisoners, Washington v. Harper, 494 U. S. 210 (1990); and restrictions on the right to marry, Turner, supra.

The right not to be discriminated against based on one’s race is not susceptible to the logic of Turner. It is not a right that need necessarily be compromised for the sake of proper prison administration. On the contrary, compliance with the Fourteenth Amendment’s ban on racial discrimination is not only consistent with proper prison administration, but also bolsters the legitimacy of the entire criminal justice system. Race discrimination is “especially pernicious in the administration of justice.” Rose v. Mitchell, 443 U. S. 545, 555 (1979). And public respect for our system of justice is undermined when the system discriminates based on race. Cf. Batson v. Kentucky, 476 U. S. 79, 99 (1986) (“[Pjublic respect for our criminal justice system and the rule of law will be strengthened if we ensure that no citizen is disqualified from jury service because of his race”). When government officials are permitted to use race as a proxy for gang membership and violence without demonstrating a compelling government interest and proving that their means are narrowly tailored, society as a whole suffers. For similar reasons, we have not used Turner to evaluate Eighth Amendment claims of cruel and unusual punishment in prison. We judge violations of that Amendment under the “deliberate indifference” standard, rather than Turner’s “reasonably related” standard. See Hope v. Pelzer, 536 U. S. 730, 738 (2002) (asking whether prison officials displayed “ ‘deliberate indifference’ to the inmates’ health or safety” where an inmate claimed that they violated his rights under the Eighth Amendment (quoting Hudson v. McMillian, 503 U. S. 1, 8 (1992))). This is because the integrity of the criminal justice system depends on full compliance with the Eighth Amendment. See Spain v. Procunier, 600 F. 2d 189, 193-194 (CA9 1979) (Kennedy, J.) (“[T]he full protections of the eighth amendment most certainly remain in force [in prison]. The whole point of the amendment is to protect persons convicted of crimes. . . . Mechanical deference to the findings of state prison officials in the context of the eighth amendment would reduce that provision to a nullity in precisely the context where it is most necessary”).

In the prison context, when the government’s power is at its apex, we think that searching judicial review of racial classifications is necessary to guard against invidious discrimination. Granting the CDC an exemption from the rule that strict scrutiny applies to all racial classifications would undermine our “unceasing efforts to eradicate racial prejudice from our criminal justice system.” McCleskey v. Kemp, 481 U. S. 279, 309 (1987) (internal quotation marks omitted).

The CDC argues that “[djeference to the particular expertise of prison officials in the difficult task of managing daily prison operations” requires a more relaxed standard of review for its segregation policy. Brief for Respondents 18. But we have refused to defer to state officials’ judgments on race in other areas where those officials traditionally exercise substantial discretion. For example, we have held that, despite the broad discretion given to prosecutors when they use their peremptory challenges, using those challenges to strike jurors on the basis of their race is impermissible. See Batson, supra, at 89-96. Similarly, in the redistricting context, despite the traditional deference given to States when they design their electoral districts, we have subjected redistricting plans to strict scrutiny when States draw district lines based predominantly on race. Compare generally Vieth v. Jubelirer, 541 U. S. 267 (2004) (partisan gerrymandering), with Shaw v. Reno, 509 U. S. 630 (1993) (racial gerrymandering).

We did not relax the standard of review for racial classifications in prison in Lee, and we refuse to do so today. Rather, we explicitly reaffirm what we implicitly held in Lee: The “necessities of prison security and discipline,” 390 U. S., at 334, are a compelling government interest justifying only those uses of race that are narrowly tailored to address those necessities. See Grutter, 539 U. S., at 353 (Thomas, J., concurring in part and dissenting in part) (citing Lee for the principle that “protecting prisoners from violence might justify narrowly tailored racial discrimination”); J. A. Croson Co., 488 U. S., at 521 (Scalia, J., concurring in judgment) (citing Lee for the proposition that “only a social emergency rising to the level of imminent danger to life and limb — for example, a prison race riot, requiring temporary segregation of inmates — can justify an exception to the principle embodied in the Fourteenth Amendment that ‘[o]ur Constitution is color-blind, and neither knows nor tolerates classes among citizens’ ” (quoting Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (Harlan, J., dissenting))); see also Pell, 417 U. S., at 823 (“[C]entral to all other corrections goals is the institutional consideration of internal security within the corrections facilities themselves”).

. Justice Thomas would subject race-based policies in prisons to Turner’s deferential standard of review because, in his view, judgments about whether race-based policies are necessary “are better left in the first instance to the officials who run our Nation’s prisons.” Post, at 542. But Turner is too lenient a standard to ferret out invidious uses of race. Turner requires only that the policy be “reasonably related” to “legitimate penological interests.” 482 U. S., at 89. Turner would allow prison officials to use race-based policies even when there are race-neutral means to accomplish the same goal, and even when the race-based policy does not in practice advance that goal. See, e. g., 321 F. 3d, at 803 (case below) (reasoning that, under Turner, the Court of Appeals did “not have to agree that the policy actually advances the CDC’s legitimate interest, but only [that] ‘defendants might reasonably have thought that the policy would advance its interests’”). See also Turner, supra, at 90 (warning that Turner is not a “least restrictive alternative test” (internal quotation marks omitted)).

For example, in Justice Thomas’ world, prison officials could segregate visiting areas on the ground that racial mixing would cause unrest in the racially charged prison atmosphere. Under Turner, “[t]he prisoner would have to prove that there would not be a riot[.] [But] [i]t is certainly ‘plausible’ that such a riot could ensue: our society, as well as our prisons, contains enough racists that almost any interracial interaction could potentially lead to conflict.” 336 F. 3d, at 1120 (case below) (Ferguson, J., dissenting from denial of rehearing en banc). Indeed, under Justice Thomas’ view, there is no obvious limit to permissible segregation in prisons. It is sot readily apparent why, if segregation in reception centers is justified, segregation in the dining halls, yards, and general housing areas is not also permissible. Any of these areas could be the potential site of racial violence. If Justice Thomas’ approach were to carry the day, even the blanket segregation policy struck down in Lee might, stand a chance of survival if prison officials simply asserted that it was necessary to prison management. We therefore reject the Turner standard for racial classifications in prisons because it would make rank discrimination too easy to defend.

The CDC protests that strict scrutiny will handcuff prison administrators and render them unable to address legitimate problems of race-based violence in prisons. See also post, at 531-532, 546-547 (Thomas, J., dissenting). Not so. Strict scrutiny is not “strict in theory, but fatal in fact.” Adarand, 515 U. S., at 237 (internal quotation marks omitted); Grutter, 539 U. S., at 326-327 (“Although all governmental uses of race are subject to strict scrutiny, not all are invalidated by it”). Strict scrutiny does not preclude the ability of prison officials to address the compelling interest in prison safety. Prison administrators, however, will have to demonstrate that any race-based policies are narrowly tailored to that end. See id., at 327 (“When race-based action is necessary to further a compelling governmental interest, such action does not violate the constitutional guarantee of equal protection so long as the narrow-tailoring requirement is also satisfied”).3

The fact that strict scrutiny applies “says nothing about the ultimate validity of any particular law; that determination is the job of the court applying strict scrutiny.” Adarand, supra, at 229-230. At this juncture, no such determination has been made. On remand, the CDC will have the burden of demonstrating that its policy is narrowly tailored with regard to new inmates as well as transferees. Prisons are dangerous places, and the special circumstances they present may justify racial classifications in some contexts. Such circumstances can be considered in applying strict scrutiny, which is designed to take relevant differences into account.

Ill

We do not decide whether the CDC’s policy violates the Equal Protection Clause. We hold only that strict scrutiny is the proper standard of review and remand the case to allow the Court of Appeals for the Ninth Circuit, or the District Court, to apply it in the first instance. See Consolidated Rail Corporation v. Gottshall, 512 U. S. 532, 557-558 (1994) (reversing and remanding for the lower court to apply the correct legal standard in the first instance); Lucas v. South Carolina Coastal Council, 505 U. S. 1003, 1031-1032 (1992) (same). The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

The Chief Justice took no part in the decision of this case.

1

Justice Thomas takes a hands-off approach to racial classifications in prisons, suggesting that a “compelling showing [is] needed to overcome the deference we owe to prison administrators.” Post, at 543 (dissenting opinion). But such deference is fundamentally at odds with our equal protection jurisprudence. We put the burden on state actors to demonstrate that their race-based policies are justified.

2

Though, as Justice Thomas points out, see post, at 544-545, and n. 12, inmates in reception centers in Oklahoma and Texas “‘are not generally assigned randomly to racially integrated cells,’” it is also the case that “these inmates are not precluded from integrated cell assignments,” Oklahoma Dept, of Corrections, Policies and Procedures, Operations Memorandum No. OP-030102, Inmate Housing (Sept. 16, 2004), available at http://www.doc.state.ok.us/docs/policies.htm (as visited Jan. 21, 2005, and available in Clerk of Court’s case file); Texas Dept, of Criminal Justice, Security Memorandum No. SM-01.28, Assignment to General Population Two-Person Cells (June 15, 2002). See also Brief for Former State Corrections Officials as Amici Curiae 20, n. 10 (“To the extent that race is considered in the assignment calculus in Oklahoma, it appears to be one factor among many, and as a result, individualized consideration is given to all inmates”). We therefore have no way of knowing whether, in practice, inmates in Oklahoma and Texas, like those in California, have close to no chance, App. to Pet. for Cert. 3a, of being celled with a person of a different race. See also Brief for Former State Corrections Officials as Amici Curiae 19-20 (“[W]e are aware of no state other than California that assumes that every incoming prisoner is incapable of getting along with a cell mate of a different race. And we are aware of no state other than California that has acted on such an assumption by adopting an inflexible and absolute policy of racial segregation of double cells in reception centers”).

3

Justice Thomas characterizes the CDC’s policy as a “limited” one, see post, at 525, but the CDC’s policy is in fact sweeping in its application.. It applies to all prisoners housed in double cells in reception centers, whether newly admitted or transferred from one facility to another. Moreover, despite Justice Thomas’ suggestion that the CDC considers other nonracial factors in determining housing placements, the CDC itself has admitted that, in practice, there is a “ ‘[p'lretty close’ ” to zero percent chance that an inmate will be housed with a person of a different race. App. to Pet. for Cert. 3a. See also generally post, at 517-518, and n. 1 (Stevens, J., dissenting). Thus, despite an inmate's “age, physical size, mental health, medical needs, [and] criminal history,” post, at 536 (Thomas, J., dissenting), the fact that he is black categorically precludes him from being celled with a white inmate. As we explain, see infra this page, we do not decide whether the threat of violence in California prisons is sufficient to justify such a broad policy.

Justice Ginsburg,

with whom Justice Souter and Justice Breyer join,

concurring.

I join the Court’s opinion, subject to the reservation expressed in Grutter v. Bollinger, 539 U. S. 306, 344-346 (2003) (Ginsburg, J., concurring).

The Court today resoundingly reaffirms the principle that state-imposed racial segregation is highly suspect and cannot, be justified on the ground that “ ‘all persons suffer [the separation] in equal degree.’” Ante, at 506 (quoting Powers v. Ohio, 499 U. S. 400, 410 (1991)). While I join that declaration without reservation, I write separately to express again my conviction that the same standard of review ought not control judicial inspection of every official race classification. As I stated most recently in Gratz v. Bollinger, 539 U. S. 244, 301 (2003) (dissenting opinion): “Actions designed to burden groups long denied full citizenship stature are not sensibly ranked with measures taken to hasten the day when entrenched discrimination and its aftereffects have been extirpated.” See also Grutter, 539 U. S., at 344-346 (Ginsburg, J., concurring); Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 271-276 (1995) (Ginsburg, J., dissenting).

There is no pretense here, however, that the California Department of Corrections (CDC) installed its segregation policy to “correct inequalities.” See Wechsler, The Nationalization of Civil Liberties and Civil Rights, Supp. to 12 Tex. Q. 10, 23 (1968). Experience in other States and in federal prisons, see ante, at 508-509; post, at 519-520 (Stevens, J., dissenting), strongly suggests that CDC’s race-based assignment of new inmates and transferees, administratively convenient as it may be, is not necessary to the safe management of a penal institution.

Disagreeing with the Court that “strict scrutiny” properly applies to any and all racial classifications, see ante, at 505-509, 511-513, 514, but agreeing that the stereotypical classification at hand warrants rigorous scrutiny, I join the Court’s opinion.

Justice Stevens,

dissenting.

In my judgment a state policy of segregating prisoners by race during the first 60 days of their incarceration, as well as the first 60 days after their transfer from one facility to another, violates the Equal Protection Clause of the Fourteenth Amendment. The California Department of Corrections (CDC) has had an ample opportunity to justify its policy during the course of this litigation, but has utterly failed to do so whether judged under strict scrutiny or the more deferential standard set out in Turner v. Safley, 482 U. S. 78 (1987). The CDC had no incentive in the proceedings below to withhold evidence supporting its policy; nor has the CDC made any offer of proof to suggest that a remand for further factual development would serve any purpose other than to postpone the inevitable. I therefore agree with the submission of the United States as amicus curiae that the Court should hold the policy unconstitutional on the current record.

The CDC’s segregation policy1 is based on a conclusive presumption that housing inmates of different races together creates an unacceptable risk of racial violence. Under the policy’s logic, an inmate’s race is a proxy for gang membership, and gang membership is a proxy for violence. The CDC, however, has offered scant empirical evidence or expert opinion to justify this use of race under even a minimal level of constitutional scrutiny. The presumption underlying the policy is undoubtedly overbroad. The CDC has made no effort to prove what fraction of new or transferred inmates are members of race-based gangs, nor has it shown more generally that interracial violence is disproportionately greater than intraracial violence in its prisons. Proclivity toward racial violence unquestionably varies from inmate to inmate, yet the CDC applies its blunderbuss policy to all new and transferred inmates housed in double cells regardless of their criminal histories or records of previous incarceration. Under the CDC’s policy, for example, two car thieves of different races — neither of whom has any history of gang involvement, or of violence, for that matter — would be barred from being housed together during their first two months of prison. This result derives from the CDC’s inflexible judgment that such integrated living conditions are simply too dangerous. This Court has never countenanced such racial prophylaxis.

To establish a link between integrated cells and violence, the CDC relies on the views of two state corrections officials. They attested to their belief that double-celling members of different races would lead to violence and that this violence would spill out into the prison yards. One of these officials, an associate warden, testified as follows:

“[W]ith the Asian population, the control sergeants have to be more careful than they do with Blacks, Whites, and Hispanics because, for example, you cannot house a Japanese inmate with a Chinese inmate. You cannot. They will kill each other. They won’t even tell you about it. They will just do it. The same with Laotians, Vietnamese, Cambodians, Filipinos. You have to be very careful about housing other Asians with other Asians. It’s very culturally heavy.” App. 189a.

Such musings inspire little confidence. Indeed, this comment supports the suspicion that the policy is based on racial stereotypes and outmoded fears about the dangers of racial integration. This Court should give no credence to such cynical, reflexive conclusions about race. See, e. g., Palmore v. Sidoti, 466 U. S. 429, 432 (1984) (“Classifying persons according to their race is more likely to reflect racial prejudice than legitimate public concerns; the race, not the person, dictates the category”); Watson v. Memphis, 373 U. S. 526, 536 (1963) (rejecting the city’s plea for delay in desegregating public facilities when “neither the asserted fears of violence and tumult nor the asserted inability to preserve the peace was demonstrated at trial to be anything more than personal speculations or vague disquietudes of city officials”).

The very real risk,that prejudice (whether conscious or not) partly underlies the CDC’s policy counsels in favor of relaxing the usual deference we pay to corrections officials in these matters. We should instead insist on hard evidence, especially given that California’s policy is an outlier when compared to nationwide practice. The Federal Bureau of Prisons administers 104 institutions; no similar policy is applied in any of them. Countless state penal institutions are operated without such a policy. An amici brief filed by six former state corrections officials with an aggregate of over 120 years of experience managing prison systems in Wisconsin, Georgia, Oklahoma, Kansas, Alaska, and Washington makes clear that a blanket policy of even temporary segregation runs counter to the great weight of professional opinion on sound prison management. See Brief for Former State Corrections Officials as Amici Curiae 19. Tellingly, the CDC can only point to two other States, Texas and Oklahoma, that use racial status in assigning inmates in prison reception areas. It is doubtful from the record that these States’ policies have the same broad and inflexible sweep as California’s, and this is ultimately beside the point. What is important is that the Federal Government and the vast majority of States address the threat of interracial violence in prisons without resorting to the expedient of segregation.

In support of its policy, the CDC offers poignant evidence that its prisons are infested with violent race-based gangs. The most striking of this evidence involves a series of riots that took place between 1998 and 2001 at Pelican Bay State Prison. That prison houses some of the State’s most violent criminal offenders, including “validated” gang members who have been transferred from other prisons. The riots involved both interracial and intraracial violence. In the most serious incident, involving 250-300 inmates, “Southern Hispanic” gang members, joined by some white inmates, attacked a number of black inmates.

Our judicial role, however, requires that we scratch below the surface of this evidence, lest the sheer gravity of a threat be allowed to authorize any policy justified in its name. Upon inspection, the CDC’s post hoc, generalized evidence of gang violence is only tenuously related to its segregation policy. Significantly, the CDC has not cited a single specific incident of interracial violence between cellmates — much less a pattern of such violence — that prompted the adoption of its unique policy years ago. Nor is there any indication that antagonism between cellmates played any role in the more recent riots' the CDC mentions. And despite the CDC’s focus on prison gangs and its suggestion that such gangs will recruit new inmates into committing racial violence during their 60-day stays in the reception centers, the CDC has cited no evidence of such recruitment, nor has it identified any instances in which new inmates committed racial violence against other new inmates in the common areas, such as the yard or the cafeteria. Perhaps the CDC’s evidence might provide a basis for arguing that at Pelican Bay and other facilities that have experienced similar riots, some race-conscious measures are justified if properly tailored. See Lee v. Washington, 390 U. S. 333, 334 (1968) (Black, J., concurring). But even if the incidents cited by the CDC, which occurred in the general prison population, were relevant to the conditions in the reception centers, they provide no support for the CDC’s decision to apply its segregation policy to all of its reception centers, without regard for each center’s security level or history of racial violence. Nor do the incidents provide any support for a policy applicable only to cellmates, while the common areas of the prison in which the disturbances occurred remain fully integrated.

Given the inherent indignity of segregation and its shameful historical connotations, one might assume that the CDC came to its policy only as a last resort. Distressingly, this is not so: There is no evidence that the CDC has ever experimented with, or even carefully considered, race-neutral methods of achieving its goals. That the policy is unwritten reflects, I think, the evident lack of deliberation that preceded its creation.

Specifically, the CDC has failed to explain why it could not, as an alternative to automatic segregation, rely on an individualized assessment of each inmate’s risk of violence when assigning him to a cell in a reception center. The Federal Bureau of Prisons and other state systems do so without any apparent difficulty. For inmates who are being transferred from one facility to another — who represent approximately 85% of those subject to the segregation policy — the CDC can simply examine their prison records to determine if they have any known gang affiliations or if they have ever engaged in or threatened racial violence. For example, the CDC has had an opportunity to observe petitioner for almost 20 years; surely the CDC could have determined his placement without subjecting him to a period of segregation.2 For new inmates, assignments can be based on their presentence reports, which contain information about offense conduct, criminal record, and personal history — including any available information about gang affiliations. In fact, state law requires the county probation officer to transmit a presentence report to the CDC along with an inmate’s commitment papers. See Cal. Penal Code Ann. § 1203c (West 2004); Cal. Rule of Court 4.411(d) (Criminal Cases) (West Supp. 2004).

Despite the rich information available in these records, the CDC considers these records only rarely in assigning inmates to cells in the reception centers. The CDC’s primary explanation for this is administrative inefficiency — the records, it says, simply do not arrive in time. The CDC’s counsel conceded at oral argument that presentence reports “have a fair amount of information,” but she stated that, “in California, the presentence report does not always accompany the inmate and frequently does not. It follows some period of time later from the county.” Tr. of Oral Arg. 33. Despite the state-law requirement to the contrary, counsel informed the Court that the counties are not preparing the presentenee reports “in a timely fashion.” Ibid. Similarly, with regard to transferees, counsel stated that their prison records do not arrive at the reception centers in time to make cell assignments. Id., at 28. Even if such inefficiencies might explain a temporary expedient in some cases, they surely do not justify a systemwide policy. When the State’s interest in administrative convenience is pitted against the Fourteenth Amendment’s ban on racial segregation, the latter must prevail. When there has been no “serious, good faith consideration of workable race-neutral alternatives that will achieve the [desired goal],” Grutter v. Bollinger, 539 U. S. 306, 339 (2003), and when “obvious, easy alternatives” are available, Turner, 482 U. S., at 90, the conclusion that CDC’s policy is unconstitutional is inescapable regardless of the standard of review that the Court chooses to apply.3

In fact, the CDC’s failure to demand timely presentence reports and prison records undercuts the sincerity of its concern for inmate security during the reception process. Race is an unreliable and necessarily underinclusive predictor of violence. Without the inmate-specific information found in the records, there is a risk that corrections officials will, for example, house together inmates of the same race who are nevertheless members of rival gangs, such as the Bloods and Crips.4

Accordingly, while I agree that a remand is appropriate for a resolution of the issue of qualified immunity, I respectfully dissent from the Court’s refusal to decide, on the basis of the record before us, that the CDC’s policy is unconstitutional.

1

The CDC operates 32 prisons, 7 of which house reception centers. All new inmates and all inmates transferring between prisons are funneled through one of these reception centers before they are permanently placed. At the centers, inmates are housed either in dormitories, double cells, or single cells (of which there are few). Under the CDC’s segregation policy, race is a determinative factor in placing inmates in double cells, regardless of the other factors considered in such decisions. While a corrections official with 24 years of experience testified that an exception to this policy was once granted to a Hispanic inmate who had been “raised with Crips,” App. 184a, the CDC’s suggestion that its policy is therefore flexible, see Brief for Respondents 9, strains credulity. There is no evidence that the CDC routinely allows inmates to opt out of segregation, much less evidence that the CDC informs inmates of their supposed right to do so.

2

In explaining why it cannot prescreen new inmates, the CDC’s brief all but concedes that segregating transferred inmates is unnecessary. See Brief for Respondents 42 (“If the officials had all of the necessary information to assess the inmates’ violence potential when the inmates arrived, perhaps a different practice could be used. But unlike the federal system, where the inmates generally are in federal custody from the moment they are arrested, state inmates are in county custody until they are convicted and later transferred to the custody of the CDC”).

3

Because the Turner factors boil down to a tailoring test, and I conclude that the CDC’s policy is, at best, an “ ‘exaggerated response’ ” to its asserted security concerns, see Turner v. Safley, 482 U. S. 78, 90 (1987), I find it unnecessary to address specifically the other factors, such as whether new and transferred inmates have “alternative means” of exercising their right to equal protection during their period of housing segregation, id., at 89. Indeed, this case demonstrates once again that “[h]ow a court describes its standard of review when a prison regulation infringes fundamental constitutional rights often has far less consequence[s] for the inmates than the actual showing that the court demands of the State in order to uphold the regulation.” Id., at 100 (Stevens, J., concurring in part and dissenting in part).

4

The CDC’s policy may be counterproductive in other ways. For example, an official policy of segregation may initiate new arrivals into a corrosive culture of prison racial segregation, lending credence to the view that members of other races are to be feared and that racial alliances are necessary. While integrated cells encourage inmates to gain valuable cross-racial experiences, segregated cells may well facilitate the formation of race-based gangs. See Brief for Former State Corrections Officials as Amici Curiae 19 (citing evidence and experience suggesting that the racial integration of cells on balance decreases interracial violence).

Justice Thomas,

with whom Justice Scalia joins,

dissenting.

The questions presented in this case require us to resolve two conflicting lines of precedent. On the one hand, as the Court stresses, this Court has said that “ ‘all racial classifications reviewable under the Equal Protection Clause must be strictly scrutinized.’” Gratz v. Bollinger, 539 U. S. 244, 270 (2003) (quoting Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 224 (1995); emphasis added). On the other, this Court has no less categorically said that “the [relaxed] standard of review we adopted in Turner [v. Safley, 482 U. S. 78 (1987),] applies to all circumstances in which the needs of prison administration implicate constitutional rights." Washington v. Harper, 494 U. S. 210, 224 (1990) (emphasis added).

Emphasizing the former line of cases, the majority resolves the conflict in favor , of strict scrutiny. I disagree. The Constitution has always demanded less within the prison walls. Time and again, even when faced with constitutional rights no less “fundamental” than the right to be free from state-sponsored racial discrimination, we have deferred to the reasonable judgments of officials experienced in running this Nation’s prisons. There is good reason for such deference in this case. California oversees roughly 160,000 inmates in prisons that have been a breeding ground for some of the most violent prison gangs in America — all of them organized along racial lines. In that atmosphere, California racially segregates a portion of its inmates, in a part of its prisons, for brief periods of up to 60 days, until the State can arrange permanent housing. The majority is concerned with sparing inmates the indignity and stigma of racial discrimination. Ante, at 507-508. California is concerned with their safety and saving their lives. I respectfully dissent.

I

To understand this case, one must understand just how limited the policy at issue is. That requires more factual background than the Court’s opinion provides. Petitioner Garrison Johnson is a black inmate in the California Department of Corrections (CDC), currently serving his sentence for murder, robbery, and assault with a deadly weapon. App. 255a-256a, 259a. Johnson began serving his sentence in June 1987 at the California Institution for Men in Chino, California. Id., at 79a, 264a. Since that time he has been transferred to a number of other facilities within the CDC. Id., at 79a-82a.

When an inmate like Johnson is admitted into the California prison system or transferred between the - CDC’s institutions, he is housed initially for a brief period — usually no more than 60 days — in one of California’s prison reception centers for men. Id., at 303a-305a. CDC, Department Operations Manual §61010.3 (2004) (hereinafter CDC Operations Manual), available at http://www.corr.ca.gov/ RegulationsPolicies/PDF/DOM/00_dept_ops_maunal.pdf (all Internet materials as visited Feb. 18, 2005, and available in Clerk of Court’s case file). In 2003, the centers processed more than 40,000 newly admitted inmates, almost 72,000 inmates returned from parole, over 14,000 inmates admitted for other reasons, and some portion of the 254,000 inmates who were transferred from one prison to another. CDC, Movement of Prison Population 3 (2003).

At the reception center, prison officials have limited information about an inmate, “particularly if he has never been housed in any CDC facility.” App. 303a. The inmate therefore is classified so that prison officials can place the inmate in appropriate permanent housing. During this process, the CDC evaluates the inmate’s “physical, mental and emotional health.” Ibid. The CDC also reviews the inmate’s criminal history and record in jail to assess his security needs and classification level. Id., at 304a. Finally, the CBC investigates whether the inmate has any enemies in prison. Ibid. This process determines the inmate’s ultimate housing placement and has nothing to do with race.

While the process is underway, the CDC houses the inmate in a one-person cell, a two-person cell, or a dormitory. Id., at 305a. The few single cells available at reception centers are reserved for inmates who present special security problems, including those convicted of especially heinous crimes or those in need of protective custody. See, e.g., CDC Operations Manual §61010.11.3. At the other end of the spectrum, lower risk inmates are assigned to dormitories. App. 189a-190a. Placement in either a single cell or a dormitory has nothing to do with race, except that prison officials attempt to maintain a racial balance within each dormitory. Id., at 250a. Inmates placed in single cells or dormitories lead fully integrated lives: The CBC does not distinguish based on race at any of its facilities when it comes to jobs, meals, yard and recreation time, or vocational and educational assignments. Ibid.

Yet some prisoners, like Johnson, neither require confinement in a single cell nor may be safely housed in a dormitory. The CBC houses these prisoners in double cells during the 60-day period. In pairing cellmates, race is indisputably the predominant factor. Id., at 305a, 309a. California’s reason is simple: Its prisons are dominated by violent gangs. Brief for Respondents 1-5. And as the largest gangs’ names indicate — the Aryan Brotherhood, the Black Guerrilla Family, the Mexican Mafia, the Nazi Low Riders, and La Nuestra Familia — they are organized along racial lines. See Part II-B, infra.

According to the State, housing inmates in double cells without regard to race threatens not only prison discipline, but also the physical safety of inmates and staff. App. 305a-306a, 310a-311a. That is because double cells are especially dangerous. The risk of racial violence in public areas of prisons is high, and the tightly confined, private conditions of cells hazard even more violence. Prison staff cannot see into the cells without going up to them, and inmates can cover the windows to prevent the staff from seeing inside the cells. Id., at 306a. The risk of violence caused by this privacy is grave, for inmates are confined to their cells for much of the day. Ibid.; id., at 187a-188a.

Nevertheless, while race is the predominant factor in pairing cellmates, it is hardly the only one. After dividing this subset of inmates based on race, the CDC further divides them based on geographic or national origin. As an example, Hispanics from northern and southern California are not housed together in reception centers because they often belong to rival gangs — La Nuestra Familia and the Mexican Mafia, respectively. Id., at 185a. Likewise, Chinese and Japanese inmates are not housed together, nor are Cambodians, Filipinos, Laotians, or Vietnamese. Id., at 189a. In addition to geographic and national origin, prison officials consider a host of other factors, including inmates’ age, mental health, medical needs, criminal history, and gang affiliation. Id., at 304a, 309a. For instance, when Johnson was admitted in 1987, he was a member of the Crips, a black street gang. Id., at 93a. He was therefore ineligible to be housed with nonblack inmates. Id., at 183a; Brief for Respondents 12, n. 9.

Moreover, while prison officials consider race in assigning inmates to double cells, the record shows that inmates are not necessarily housed with other inmates of the same race during that 60-day period. When a Hispanic inmate affiliated with the Crips asked to be housed at the reception center with a black inmate, for example, prison administrators granted his request. App. 183a-184a, 199a. Such requests are routinely granted after the 60-day period, when prison officials complete the classification process and transfer an inmate from the reception center to a permanent placement at that prison or another one.1 Id., at 311a-312a.

II

Traditionally, federal courts rarely involved themselves in the administration of state prisons, “adopting] a broad hands-off attitude toward problems of prison administration.”2 Procunier v. Martinez, 416 U. S. 396, 404 (1974). For most of this Nation’s history, only law-abiding citizens could claim the cover of the Constitution: Upon conviction and incarceration, defendants forfeited their constitutional rights and possessed instead only those rights that the State chose to extend them. See, e. g., Shaw v. Murphy, 532 U. S. 223, 228 (2001); Ruffin v. Commonwealth, 62 Va. 790, 796 (1871). In recent decades, however, this Court has decided that incarceration does not divest prisoners of all constitutional protections. See, e. g., Wolff v. McDonnell, 418 U. S. 539, 555-556 (1974) (the right to due process); Cruz v. Beto, 405 U. S. 319, 322 (1972) (per curiam) (the right to free exercise of religion).3

At the same time, this Court quickly recognized that the extension of the Constitution’s demands behind prison walls had to accommodate the needs of prison administration. This Court reached that accommodation in Turner v. Safley, 482 U. S. 78 (1987), which “adopted a unitary, deferential standard for reviewing prisoners’ constitutional claims,” Shaw, supra, at 229. That standard should govern Johm son’s claims, as it has governed a host of other claims challenging conditions of confinement, even when restricting the rights at issue would otherwise have occasioned strict scrutiny. Under the Turner standard, the CDC’s policy passes constitutional muster because it is reasonably related to legitimate penological interests.

A

Well before Turner, this Court recognized that experienced prison administrators, and not judges, are in the best position to supervise the daily operations of prisons across this country. See, e. g., Jones v. North Carolina Prisoners' Labor Union, Inc., 433 U. S. 119, 125 (1977) (courts must give “appropriate deference to the decisions of prison administrators”); Procunier, supra, at 405 (“[C]ourts are ill equipped to deal with the increasingly urgent problems of prison administration and reform”). Turner made clear that a deferential standard of review would apply across the board to inmates’ constitutional challenges to prison policies.

At issue in Turner was the constitutionality of a pair of Missouri prison regulations limiting inmate-to-inmate correspondence and inmate marriages. The Court’s analysis proceeded in two steps. First, the Court recognized that prisoners are not entirely without constitutional rights. As proof, it listed certain constitutional rights retained by prisoners, including the right to be “protected against invidious racial discrimination . . . , Lee v. Washington, 390 U. S. 333 (1968).” Turner, 482 U. S., at 84. Second, the Court concluded that for prison administrators rather than courts to “ ‘make the difficult judgments concerning institutional operations,”’ id., at 89 (quoting Jones, supra, at 128), courts should uphold prison regulations that impinge on those constitutional rights if they reasonably relate to legitimate pe-nological interests, 482 U. S., at 89. Nowhere did the Court suggest that Lee's right to be free from racial discrimination was immune from Turner’s deferential standard of review. To the contrary, “[w]e made quite clear that the standard of review we adopted in Turner applies to all circumstances in which the needs of prison administration implicate constitutional rights.” Harper, 494 U. S., at 224 (emphasis added).

Consistent with that understanding, this Court has applied Turner’s standard to a host of constitutional claims by prisoners, regardless of the standard of review that would apply outside prison walls.4 And this Court has adhered to Turner despite being, urged to adopt different standards of review based on the constitutional provision at issue. See Harper, supra, at 224 (Turner's standard of review “applies] in all cases in which a prisoner asserts that a prison regulation violates the Constitution, not just those in which the. prisoner invokes the First Amendment” (emphasis added)); O’Lone v. Estate of Shabazz, 482 U. S. 342, 353 (1987) (“We take this opportunity to reaffirm our refusal, even where claims are made under the First Amendment, to substitute our judgment on ... difficult and sensitive matters of institutional administration for the determinations of those charged with the formidable task of running a prison” (internal quotation marks and citation omitted; emphasis added)). Our steadfast adherence makes sense: If Turner is our accommodation of the Constitution’s demands to those of prison administration, see supra, at 530, we should apply it uniformly to prisoners’ challenges to their conditions of confinement.

After all, Johnson’s claims, even more than other claims to which we have applied Turner’s test, implicate Turner’s rationale. In fact, in a passage that bears repeating, the Turner Court explained precisely why deference to the judgments of California’s prison officials is necessary:

“Subjecting the day-to-day judgments of prison officials to an inflexible strict scrutiny analysis would seriously hamper their ability to anticipate security problems and to adopt innovative solutions to the intractable problems of prison administration. The rule would also distort the decisionmaking process, for every administrative judgment would be subject to the possibility that some court somewhere would conclude that it had a less restrictive way of solving the problem at hand. Courts inevitably would become the primary arbiters of what constitutes the best solution to every administrative problem, thereby unnecessarily perpetuating the involvement of the federal courts in affairs of prison administration.” 482 U. S., at 89 (internal quotation marks and alteration omitted).

The majority’s failure to heed that advice is inexplicable, especially since Turner itself recognized the “growing problem with prison gangs.” Id., at 91. In fact, there is no more “intractable problem” inside America’s prisons than racial violence, which is driven by race-based prison gangs. See, e. g., Dawson v. Delaware, 503 U. S. 159, 172-173, and n. 1 (1992) (Thomas, J., dissenting); Stefanow v. McFadden, 103 F. 3d 1466, 1472 (CA9 1996) (“Anyone familiar with prisons understands the seriousness of the problems caused by prison gangs that are fueled by actively virulent racism and religious bigotry”).

B

The majority decides this case without addressing the problems that racial violence poses for wardens, guards, and inmates throughout the federal and state prison systems. But that is the core of California’s justification for its policy: It maintains that, if it does not racially separate new cellmates thrown together in close confines during their initial admission or transfer, violence will erupt.

The dangers California seeks to prevent are real. See Brief for National Association of Black Law Enforcement Officers, Inc., as Amicus Curiae 12. Controlling prison gangs is the central challenge facing correctional officers and administrators. Carlson, Prison Interventions: Evolving Strategies to Control Security Threat Groups, 5 Corrections Mgmt. Q. 10 (Winter 2001) (hereinafter Carlson). The worst gangs are highly regimented and sophisticated organizations that commit crimes ranging from drug trafficking to theft and murder. Id., at 12; Cal. Dept. of Justice, Division of Law Enforcement, Organized Crime in California Annual Report to the California Legislature 2003, p. 15, available at http://caag.state.ca.us/publications/org_crime.pdf. In fact, street gangs are often just an extension of prison gangs, their “ ‘foot soldiers’ ” on the outside. Ibid.; Widens, Structure, Content and the Exigencies of War: American Prison Law After Twenty-Five Years 1962-1987, 37 Am. U. L. Rev. 41, 55-56 (1987). And with gang membership on the rise, the percentage of prisoners affiliated with prison gangs more than doubled in the 1990’s.5

The problem of prison gangs is not unique to California,6 but California has a history like no other. There are at least five major gangs in this country — the Aryan Brotherhood, the Black Guerrilla Family, the Mexican Mafia, La Nuestra Familia, and the Texas Syndicate — all of which originated in California’s prisons.7 Unsurprisingly, then, California has the largest number of gang-related inmates of any correctional system in the country, including the Federal Government. Carlson 16.

As their very names suggest, prison gangs like the Aryan Brotherhood and the Black Guerrilla Family organize themselves along racial lines, and these gangs perpetuate hate and violence. Irwin 182, 184. Interracial murders and assaults among inmates perpetrated by these gangs are common.8 And, again, that brutality is particularly severe in California’s prisons. See, e. g., Walker v. Gomez, 370 F. 3d 969, 971 (CA9 2004) (describing “history of significant racial tension and violence” at Calipatria State Prison); id., at 979-980 (Rymer, J., dissenting) (same); App. 297a-299a (describing 2-year span at Pelican Bay Prison, during which there were no fewer than nine major riots that left at least one inmate dead and many more wounded).

C

It is against this backdrop of pervasive racial violence that California racially segregates inmates in the reception centers’ double cells, for brief periods of up to 60 days, until such time as the State can assign permanent housing. Viewed in that context and in light of the four factors enunciated in Turner, California’s policy is constitutional: The CDC’s policy is reasonably related to a legitimate penological interest; alternative means of exercising the restricted right remain open to inmates; racially integrating double cells might negatively impact prison inmates, staff, and administrators; and there are no obvious, easy alternatives to the CDC’s policy.

1

First, the policy is reasonably related to a legitimate penological interest. Turner, supra, at 89. The protection of inmates and staff is undeniably a legitimate penological interest. See Bell v. Wolfish, 441 U. S. 520, 546-547 (1979). The evidence shows, and Johnson has never contested, that the objective of California’s policy is reducing violence among the inmates and against the staff. No cells are designated for, nor are special privileges afforded to, any racial group. App. 188a, 305a. Because prison administrators use race as a factor in making initial housing assignments “solely on the basis of [its] potential implications for prison security,” the CDC’s cell assignment practice is neutral. Thornburgh v. Abbott, 490 U. S. 401, 415 (1989); Turner, 482 U. S., at 90.

California’s policy bears a valid, rational connection to this interest. The racial component to prison violence is impossible for prison administrators to ignore. Johnson himself testified that he is afraid of violence — based solely on the color of his skin.9 In combating that violence, an inmate’s arrival or transfer into a new prison setting is a critical time for inmate and staff alike. The policy protects an inmate from other prisoners, and they from him, while prison officials gather more information, including his gang affiliation, about his compatibility with other inmates. App. 249a. This connection between racial violence and the policy makes it far from “arbitrary or irrational.” Turner, supra, at 89-90.

Indeed, Johnson concedes that it would be perfectly constitutional for California to take account of race “as part of an overall analysis of proclivity to violence based upon a series of facts existing in that prison.” Tr. of Oral Arg. 15. But that is precisely what California does. It takes into account a host of factors in addition to race: geographic or national origin, age, physical size, mental health, medical needs, criminal history, and, of course, gang affiliation. Supra, at 527. California does not simply assign inmates to double cells in the reception centers based on race — it also separates intra-racially (for example, northern from southern Hispanies or violent from nonviolent offenders).

2

Second, alternative means of exercising the restricted right remain open to inmates like Johnson. Turner, supra, at 90. The CDC submits, and Johnson does not contest, that all other facets of prison life are fully integrated: work, vocational, and educational assignments; dining halls; and exercise yards and recreational facilities. App. 250a. And after a brief detention period at the reception center, inmates may select their own cellmates regardless of race in the absence of overriding security concerns. Id., at 311a-312a. Simply put, Johnson has spent, and will continue to spend, the vast bulk of his sentence free from any limitation on the race of his cellmate.

3

Third, Johnson fails to establish that the accommodation he seeks — i e., assigning inmates to double cells without regard to race — would not significantly impact prison personnel, other inmates, and the allocation of prison resources. Harper, 494 U. S., at 226-227; Turner, supra, at 90. Prison staff cannot see into the double cells without going up to them, and inmates can cover the windows so that staff cannot see inside the cells at all. App. 306a. Because of the limited number of staff to oversee the many cells, it “would be very difficult to assist inmates if the staff were needed in several places at one time.” Ibid. Coordinated gang attacks against nongang cellmates could leave prison officials unable to respond effectively. In any event, diverting prison resources to monitor cells disrupts services elsewhere.

Then, too, fights in the cells are likely to spill over to the exercise yards and common areas. Ibid.; see also id., at 187a. As Turner made clear: “When accommodation of an asserted right will have a significant ‘ripple effect’ on fellow inmates or on prison staff, courts should be particularly deferential to the informed discretion of corrections officials.” 482 U. S., at 90; see also White v. Morris, 832 F. Supp. 1129, 1130 (SD Ohio 1993) (racially integrated double-celling contributed to a race riot in which 10 people were murdered). California prison officials are united in the view that racially integrating double cells in the reception centers would lead to serious violence.10 This is precisely the sort of testimony that the Court found persuasive in Turner itself. 482 U. S., at 92.

4

Finally, Johnson has not shown that there are “obvious, easy alternatives” to the CDC’s policy. Id., at 90. Johnson contends that, for newly admitted inmates, prison officials need only look to the information available in the presen-tence report that must accompany a convict to prison. See Cal. Penal Code Ann. § 1203(c) (West 2004); Cal. Rules of Ct., Crim., Rule 4.411(d) (West Supp. 2004). But prison officials already do this to the extent that they can. Indeed, gang affiliation, not race, is the first factor in determining initial housing assignments. App. 315a. Race becomes the predominant factor only because gang affiliation is often not known, especially with regard to newly admitted inmates. As the Court of Appeals pointed out: “There is little chance that inmates will be forthcoming about their past violent episodes or criminal gang activity so as to provide an accurate and dependable picture of the inmate.” 321 F. 3d 791, 806 (CA9 2003); see also App. 185a, 189a. Even if the CDC had the manpower and resources to prescreen the more than 40,000 new inmates it receives yearly, leafing through pre-sentence reports would not tell prison officials what they need to know. See ante, at 521-523 (Stevens, J., dissenting).

Johnson presents a closer case with regard to the segregation of prisoners whom the CDC transfers between facilities. As I understand it, California has less need to segregate prisoners about whom it already knows a great deal (since they have undergone the initial classification process and been housed for some period of time). However, this does not inevitably mean that racially integrating transferred inmates, while obvious and easy, is a true alternative. For instance, an inmate may have affiliated with a gang since the CDC’s last official assessment, or his past lack of racial violence may have been due to the absence of close confinement with members of other races. The CDC’s policy , does not appear to arise from laziness or neglect; California is a leader in institutional intelligence gathering. See Carlson 16 (“The CDC devotes 75 intelligence staff to gathering and verifying inmate-related information,” both in prisons and on the streets). In short, applying the policy to transfers is not “arbitrary or irrational,” requiring that we set aside the considered contrary judgment of prison administrators. Turner, supra, at 89-90.

Ill

The majority claims that strict scrutiny is the applicable standard of review based on this Court’s precedents and its general skepticism of racial classifications. It is wrong on both scores.

A

Only once before, in Lee v. Washington, 390 U. S. 333 (1968) (per curiam), has this Court considered the constitutionality of racial classifications in prisons. The majority claims that Lee applied “a heightened standard of review.” Ante, at 506. But Lee did not address the applicable standard of review. And even if it bore on the standard of review, Lee would support the State here.

In Lee, a three-judge District Court ordered Alabama to desegregate its prisons under Brown v. Board of Education, 347 U. S. 483 (1954). Washington v. Lee, 263 F. Supp. 327, 331-332 (MD Ala. 1966). In so doing, the District Court rejected any notion that “consideration^] of prison security or discipline” justified the “complete and permanent segregation of the races in all the Alabama penal facilities.” Id., at 331. However, the District Court noted “that in some isolated instances prison security and discipline necessitates segregation of the races for a limited period.” Ibid, (footnote omitted). It provided only one example — “the ‘tank’ used in . . . large municipal jails where intoxicated persons are placed upon their initial incarceration and kept until they become sober,” id., at 331, n. 6 — and the court left unmentioned why it would have-been necessary to separate drunk whites from blacks on a Birmingham Saturday night.

This Court, in a per curiam, one-paragraph opinion, affirmed the District Court’s order. It found “unexceptionable” not only the District Court’s general rule that wholesale segregation of penal facilities was unconstitutional, but also the District Court’s “allowance for the necessities of prison security and discipline.” Lee, 390 U. S., at 334. Indeed, Justices Black, Harlan, and Stewart concurred

“to make explicit something that is left to be gathered only by implication from the Court’s opinion. This is that prison authorities have the right, acting in good faith and in particularized circumstances, to take into account racial tensions in maintaining security, discipline, and good order in prisons and jails.” Ibid.

Those Justices were “unwilling to assume” that such an “explicit pronouncement [would] evinc[e] any dilution of this Court’s firm commitment to the Fourteenth Amendment’s prohibition of racial discrimination.” Ibid.

Lee said nothing about the applicable standard of review, for there was no need. Surely Alabama’s wholesale segregation of its prisons was unconstitutional even under the more deferential standard of review, that applies within prisons. This Court’s brief, per curiam opinion in Lee simply cannot bear the weight or interpretation the majority places on it. See U. S. Bancorp Mortgage Co. v. Bonner Mall Partnership, 513 U. S. 18, 24 (1994) (noting “our customary skepticism toward per curiam dispositions that lack the reasoned consideration of a full opinion”); Edelman v. Jordan, 415 U. S. 651, 670-671 (1974).

Yet even if Lee had announced a heightened standard of review for prison policies that pertain to race, Lee also carved out an exception to the standard that California’s policy would certainly satisfy. As the Lee concurrence explained without objection, the Court’s exception for “the necessities of prison security and discipline” meant that “prison authorities háve the right, acting in good faith and in particularized circumstances, to take into account racial tensions in maintaining security, discipline, and good order in prisons and jails.” 390 U. S., at 334 (opinion of Black, Harlan, and Stewart, JJ., concurring) (emphasis added).

California’s policy — which is a far cry from the wholesale segregation at issue in Lee — would fall squarely within Lee’s exception. Johnson has never argued that California’s policy is motivated by anything other than a desire to protect inmates and staff. And the “particularized” nature of the policy is evident: It applies only to new inmates and transfers, only in a handful of prisons, only to double cells, and only then for a period of no more than two months. In the name of following a test that Lee did not create, the majority opts for a more demanding standard of. review than Lee's language even arguably supports.

The majority heavily relies on this Court’s statement that “ ‘all racial classifications [imposed by government]... must be analyzed by a reviewing court under strict scrutiny.’” Ante, at 505 (emphasis deleted) (quoting Adarand Constrctors, Inc., 515 U. S., at 227). Adamnd has nothing to do with this case. Adamnd's statement that “all racial classifications” are subject to strict scrutiny addréssed the contention that classifications favoring rather than disfavoring blacks are exempt. Id., at 226-227; accord, Grutter v. Bollinger, 539 U. S. 306, 353 (2003) (Thomas, J., concurring in part and dissenting in part). None of these statements overruled, sub silentio, Turner and its progeny, especially since the Court has repeatedly held that constitutional demands are diminished in the unique context of prisons. See, e. g., Harper, 494 U. S., at 224; Abbott, 490 U. S., at 407; Turner, 482 U. S., at 85; see also Webster v. Fall, 266 U. S. 507, 511 (1925) (“Questions which merely lurk in the record, neither brought to the attention of the court nor ruled upon, are not to be considered as having been so decided as to constitute precedents”).

B

The majority offers various other reasons for applying strict scrutiny. None is persuasive. The majority’s main reason is that “Turner’s, reasonable-relationship test [applies] only to rights that are ‘inconsistent with proper incarceration.’” Ante, at 510 (quoting Overton v. Bazzetta, 539 U. S. 126, 131 (2003)). According to the majority, the question is thus whether a right “need necessarily be compromised for the sake of proper prison administration.” Ante, at 510. This inconsistency-with-proper-prison-administration test begs the question at the heart of this case. For a court to know whether any particular right is inconsistent with proper prison administration, it must have some implicit notion of what a proper prison ought to look like and how it ought to be administered. Overton, supra, at 139 (Thomas, J., concurring in judgment). But the very issue in this case is whether such second-guessing is permissible.

The majority’s test eviscerates Turner. Inquiring whether a given right is consistent with “proper prison administration” calls for precisely the sort of judgments that Turner said courts were ill equipped to make. In none of the cases in which the Court deferred to the judgments of prison officials under Turner did it examine whether “proper” prison security and discipline permitted greater speech or associational rights (Abbott, supra; Shaw, 532 U. S. 223; and Overton, supra); expanded access to the courts (Lewis v. Casey, 518 U. S. 343 (1996)); broader freedom from bodily restraint (Harper, supra); or additional free exercise rights (O’Lone, 482 U. S. 342). The Court has steadfastly refused to undertake the threshold standard-of-review inquiry that Turner settled, and that the majority today resurrects. And with good reason: As Turner pointed out, these judgments are better left in the first instance to the officials who run our Nation’s prisons, not to the judges who run its courts.

In place of the Court’s usual deference, the majority gives conclusive force to its own guesswork about “proper” prison administration. It hypothesizes that California’s policy might incite, rather than diminish, racial hostility.11 Ante, at 506-508. The majority's speculations are implausible. New arrivals have a strong interest in promptly convincing other inmates of their willingness to use violent force. See Brief for National Association of Black Law Enforcement Officers, Inc., as Amicus Curiae 13-14 (citing commentary and congressional findings); cf. United States v. Santiago, 46 F. 3d 885, 888 (CA9 1995) (describing one Hispanic inmate's murder of another in order to join the Mexican Mafia); United States v. Silverstein, 732 F. 2d 1338, 1341 (CA7 1984) (prospective members of the Aryan Brotherhood must “make bones,” or commit a murder, to be eligible for membership). In any event, the majority’s guesswork falls far short of the compelling showing needed to overcome the deference we owe to prison administrators.

The majority contends that the Court “[has] put the burden on state actors to demonstrate that their race-based policies are justified,” ante, at 506, n. 1, and “[has] refused to defer to state officials’ judgments on race in other areas where those officials traditionally exercise substantial discretion,” ante, at 512. Yet two Terms ago, in upholding the University of Michigan Law School’s affirmative-action program, this Court deferred to the judgment by the law school’s faculty and administrators on their need for diversity in the student body. See Grutter, supra, at 328 (“The Law School’s educational judgment that... diversity is essential to its educational mission is one to which we defer”). Deference would seem all the more warranted in the prison context, for whatever the Court knows of administering educational institutions, it knows much less about administering penal ones. The potential consequences of second-guessing the judgments of prison administrators are also much more severe. See White v. Morris, 832 F. Supp. 1129, 1130 (SD Ohio 1993) (racially integrated double-celling that resulted from federal consent decree was a factor in the worst prison riot in Ohio history). More importantly, as I have explained, the Court has recognized that the typically exacting review it applies to restrictions on fundamental rights must be relaxed in the unique context of prisons. See, e. g., Harper, 494 U. S., at 224; Abbott, 490 U. S., at 407; Turner, 482 U. S., at 85. The majority cannot fall back on the Constitution’s usual demands, because those demands have always been lessened inside the prison walls. See supra, at 529.

The majority also mentions that California’s policy may be the only one of its kind, as virtually all other States and the Federal Government manage their prison systems without racially segregating inmates. Ante, at 508-509. This is both irrelevant and doubtful. It is irrelevant because the number of States that have followed California’s lead matters not to the applicable standard of review (the only issue the Court today decides), but to whether California satisfies whatever standard applies, a question the majority leaves to be addressed on remand. In other words, the uniqueness of California’s policy might show whether the policy is reasonable or narrowly tailored — but deciding whether to apply Turner or strict scrutiny in the first instance must depend on something else, like the majority’s inconsistency-with-proper-prison-administration test. The commonness of California’s housing policy is further irrelevant because strict scrutiny now applies to all claims of racial discrimination in prisons, regardless of whether the policies being challenged are unusual.

The majority’s assertion is doubtful, because at least two other States apply similar policies to newly admitted inmates. Both Oklahoma and Texas, like California, assign newly admitted inmates to racially segregated cells in their prison reception centers.12 The similarity is not surprising: States like California and Texas have historically had the most severe problems with prison gangs. However, even States with less severe problems maintain that policies like California’s are necessary to deal with race-related prison violence. See Brief for States of Utah, Alabama, AJaska, Delaware, Idaho, Nevada, New Hampshire and North Dakota as Amici Curiae 16. Relatedly, 10.3% of all wardens at maximum security facilities in the United States report that their inmates are assigned to racially segregated cells — apr parently on a permanent basis. Henderson, Cullen, Carroll, & Feinberg, Race, Rights, and Order in Prison: A National Survey of Wardens on the Racial Integration of Prison Cells, 80 Prison J. 295, 304 (Sept. 2000). In the same survey, 4.3% of the wardens report that their States have an official policy against racially integrating male inmates in cells. Id., at 302. Presumably, for the remainder of prisons in which inmates are assigned to racially segregated cells, that policy is the result of discretionary decisions by wardens rather than of official state directives. Ibid. In any event, the ongoing debate about the best way to reduce racial violence in prisons should not be resolved by judicial decree: It is the job “of prison administrators ... and not the courts, to make the difficult judgments concerning institutional operations.” Jones, 433 U. S., at 128.

The majority also observes that we have already carved out an exception to Turner for Eighth Amendment claims of cruel and unusual punishment in prison. See Hope v. Pelzer, 536 U. S. 730, 738 (2002). In that context, we have held that “[a] prison official’s ‘deliberate indifference’ to a substantial risk of serious harm to an inmate violates the Eighth Amendment.” Farmer v. Brennan, 511 U. S. 825, 828 (1994). Setting aside whether claims challenging inmates’ conditions of confinement should be cognizable under the Eighth Amendment at all, see Hudson v. McMillian, 503 U. S. 1, 18-19 (1992) (Thomas, J., dissenting), the “deliberate indifference” standard does not bolster the majority’s argument. If anything, that standard is more deferential to the judgments of prison administrators than Turner's reasonable-relationship test: It subjects prison officials to liability only when they are subjectively aware of the risk to the inmate, and they fail to take reasonable measures to abate the risk. Farmer, supra, at 847. It certainly does not demonstrate the wisdom of an exception that imposes a heightened standard of review , on the actions of prison officials.

Moreover, the majority’s decision subjects prison officials to competing and perhaps conflicting demands. In this case, California prison officials have uniformly averred that random double-celling poses a substantial risk of serious harm to the celled inmates. App. 245a-246a, 251a. If California assigned inmates to double cells without regard to race, knowing full well that violence might result, that would seem the very definition of deliberate indifference. See Robinson v. Prunty, 249 F. 3d 862, 864-865 (CA9 2001) (prisoner alleged an Eighth Amendment violation because administrators had failed to consider race when releasing inmates into the yards); Jensen v. Clarke, 94 F. 3d 1191, 1201, 1204 (CA8 1996) (court held that random double-celling by prison officials constituted deliberate indifference, and affirmed an injunction and attorney’s fees awarded against the officials). Nor would a victimized inmate need to prove that prison officials had anticipated any particular attack; it would be sufficient that prison officials had ignored a dangerous condition that was chronic and ongoing — like interracial housing in closely confined quarters within prisons dominated by racial gangs. Farmer, supra, at 843-844. Under Farmer, prison officials could have been ordered to take account of the very thing to which they may now have to turn a blind eye: inmates’ race.

Finally, the majority presents a parade of horribles designed to show that applying the Turner standard would grant prison officials unbounded discretion to segregate inmates throughout prisons. See ante, at 513-514. But we have never treated Turner as a blank check to prison officials. Quite to the contrary, this Court has long had “confidence that ... a reasonableness standard is not toothless.” Abbott, 490 U. S., at 414 (internal quotation marks omitted). California prison officials segregate only double cells, because only those cells are particularly difficult to monitor— unlike “dining halls, yards, and general housing areas.” Ante, at 514. Were California’s policy not so narrow, the State might well have race-neutral means at its disposal capable of accommodating prisoners’ rights without sacrificing their safety. See Turner, 482 U. S., at 90-91. The majority does not say why Turner’s standard ably polices all other constitutional infirmities, just not racial discrimination. In any event, it is not the refusal to apply — for the first time ever — a strict standard of review in the prison context that is “fundamentally at odds” with our constitutional jurisprudence. Ante, at 506, n. 1. Instead, it is the majority’s refusal — for the first time ever — to defer to the expert judgment of prison officials.

IV

Even under strict scrutiny analysis, “it is possible, even likely, that prison officials could show that the current policy meets the test.” 336 F. 3d 1117, 1121 (CA9 2003) (Ferguson, J., joined by Pregerson, Nelson, and Reinhardt, JJ., dissenting from denial of rehearing en banc). As Johnson concedes, all States have a compelling interest in maintaining order and internal security within their prisons. See Reply Brief for Petitioner 18; see also Procunier, 416 U. S., at 404. Thus the question on remand will be whether the CDC’s policy is narrowly tailored to serve California’s compelling interest.13 The other dissent notes the absence of evidence on that question, see ante, at 518-621 (opinion of Stevens, J.), but that is hardly California’s fault.

. From the outset, Johnson himself has alleged, in terms taken from Turner, that the CDC’s policy is “not related to a legitimate penological interest.” Johnson v. California, 207 F. 3d 650, 655 (CA9 2000) (per curiam) (discussing Johnson’s Third Amended Complaint). In reinstating Johnson’s equal protection claim following the District Court’s dismissal, the Court of Appeals repeated Johnson’s allegation, without indicating that strict scrutiny should apply on remand before the District Court.14 Ibid. And on remand, again Johnson alleged only that the CDC’s policy “is not reasonably related to the legitimate penological interests of the CBC.” App. 51a (Fourth Amended Complaint ¶ 23).

After the District Court granted qualified immunity to some of the defendants, Johnson once again appealed. In his brief before the Court of Appeals, Johnson assumed that both Lee and Turner applied, without arguing that there was any tension between them; indeed, nowhere in his brief did Johnson even mention the words “strict scrutiny.” Brief for Appellant in No. 01-56436 (CA9), pp. 20, 26, 2001 WL 34091249. Perhaps as a result, the Court of Appeals did not discuss strict scrutiny in its second' decision, the one currently before this Court. The Court of Appeals did find tension between Lee and Turner; however, it resolved this tension in Turner’s favor. 321 F. 3d, at 799. Yet the Court of Appeals accepted Lee’s test at face value: Prison officials may only make racial classifications “ ‘in good faith and in particularized circumstances.’” 321 F. 3d, at 797. The Court of Appeals, like Johnson, did not equate Lee’s test with strict scrutiny, and in fact it mentioned strict scrutiny only when it quoted the portion of Turner that rejects strict scrutiny as the proper standard of review in the prison context. 321 F. 3d, at 798. Even Johnson did not make the leap equating Lee with strict scrutiny when he requested that the Court of Appeals rehear his case. Appellant’s Petition for Panel Rehearing with Suggestion for Rehearing En Banc in No. 01-56436 (CA9), pp. 4-5. That leap was first made by the judges who dissented from the Court of Appeals’ denial of rehearing en banc. 336 F. 3d, at 1118 (Ferguson, J., joined by Pregerson, Nelson, and Reinhardt, JJ., dissenting from denial of rehearing en banc).

Thus, California is now, after the close of discovery, subject to a more stringent standard than it had any reason to anticipate from Johnson’s pleadings, the Court of Appeals’ initial decision, or even the Court of Appeals’ decision below. In such circumstances, California should be allowed to present evidence of narrow tailoring, evidence it was never obligated to present in either appearance before the District Court. See Lucas v. South Carolina Coastal Council, 505 U. S. 1003, 1031-1032 (1992) (remanding for consideration under the correct legal standard); id., at 1033 (Kennedy, J., concurring in judgment) (“Although we establish a framework for remand,... we do not decide the ultimate [constitutional] question [because] [t]he facts necessary to the determination have not been developed in the record”).

* * *

Petitioner Garrison Johnson challenges not permanent, but temporary, segregation of only a portion of California’s prisons. Of the 17 years Johnson has been incarcerated, California has assigned him a cellmate of the same race for no more than a year (and probably more like four months); Johnson has had black cellmates during the other 16 years, but by his own choice. Nothing in the record demonstrates that if Johnson (or any other prisoner) requested to be housed with a person of a different race, it would be denied (though Johnson’s gang affiliation with the Crips might stand in his way). Moreover, Johnson concedes that California’s prisons are racially violent places, and that he lives in fear of being attacked because of his race. Perhaps on remand the CDC’s policy will survive strict scrutiny, but in the event that it does not, Johnson may well have won a Pyrrhic victory.

1

Johnson has never requested — not during his initial admittance, nor his subsequent transfers, nor his present incarceration — that he be housed with a person of a different race. App. 106a, 112a-113a, 175a. According to Johnson, he considered the policy a barrier to any such request; however, Johnson has also testified that he never filed a grievance with prison officials about the segregation policy. Id., at 112a-113a, 124a-125a. Neither the parties nor the majority discusses whether Johnson has exhausted his action under Rev. Stat. § 1979, 42 U. S. C. § 1983, as required by the Prison Litigation Reform Act of 1995, 110 Stat. 1321-66, as amended, 42 U. S. C. § 1997e(a). See Booth v. Churner, 532 U. S. 731, 734 (2001). The majority thus assumes that statutorily mandated exhaustion is not jurisdictional, and that California has waived the issue by failing to raise it. See, e. g., Richardson v. Goord, 347 F. 3d 431, 433-434 (CA2 2003); Perez v. Wisconsin Dept. of Corrections, 182 F. 3d 532, 536 (CA7 1999).

2

The majority refers to my approach as a “hands-off” one because I would accord deference to the judgments of the State’s prison officials. See ante, at 506, n. 1. Its label is historically inaccurate. The “hands-off” approach was that taken prior to the 1960’s by federal courts, which generally declined to consider the merits of prisoners’ claims. See, e. g., J. Fliter, Prisoners’ Rights: The Supreme Court and Evolving Standards of Decency 64-65 (2001); M. Feeley & E. Rubin, Judicial Policy Making and the Modern State 30-34 (2000); S. Krantz & L. Branham, Cases and Materials on the Law of Sentencing, Corrections and Prisoners’ Rights 264-265 (4th ed. 1991).

3

A prisoner may not entirely surrender his constitutional rights at the prison gates, Bell v. Wolfish, 441 U. S. 520, 545 (1979); Jones v. North Carolina Prisoners’ Labor Union, Inc., 433 U. S. 119, 129 (1977), but certainly he leaves some of his liberties behind him. When a prisoner makes a constitutional claim, the initial question should be whether the prisoner possesses the right at issue at all, or whether instead the prisoner has been divested of the right as a condition of his conviction and confinement. See Overton v. Bazzetta, 539 U. S. 126, 140 (2003) (Thomas, J., concurring in judgment); Coffin v. Reichard, 143 F. 2d 443, 445 (CA6 1944).

4

See, e. g., Overton, supra, at 132 (the right to association under the First and Fourteenth Amendments); Shaw v. Murphy, 532 U. S. 223, 228-229 (2001) (the right to communicate with fellow inmates under the First Amendment); Lewis v. Casey, 518 U. S. 343, 361 (1996) (the right of access to the courts under the Due Process and Equal Protection Clauses); Washington v. Harper, 494 U. S. 210, 223-225 (1990) (the right to refuse forced medication under the Due Process Clause); Thornburgh v. Abbott, 490 U. S. 401, 413-414 (1989) (the right to receive correspondence under the First Amendment); O’Lone v. Estate of Shabazz, 482 U. S. 342, 349-350 (1987) (the right to free exercise of religion under the First Amendment).

5

See National Gang Crime Research Center, A National Assessment of Gangs and Security Threat Groups (STGs) in Adult Correctional Institutions: Results of the 1999 Adult Corrections Survey, p. 5, http://www. ngcrc.com/ngcrc/page7.htm.

6

See, e. g., Fraise v. Terhune, 283 F. 3d 506, 512-513 (CA3 2002) (describing violence caused by a single black prison gang, the Five Percent Nation, in various New Jersey correctional facilities); Conroy v. Dingle, No. Civ. 01-1626 (RHK/RLE), 2002 WL 31357055, *1-*2 (D. Minn., Oct. 11, 2002) (describing rival racial gangs at Minnesota’s Moose Lake facility, a medium security prison).

7

See D. Orlando-Morningstar, Prison Gangs, Special Needs Offenders Bulletin, Federal Judicial Center 4 (Oct. 1997); see also J. Irwin, Prisons 'in Turmoil 189 (1980) (hereinafter Irwin) (describing the establishment and rise of gangs inside the California prison system, first the Mexican Mafia, followed by La Nuestra Familia, the Aryan Brotherhood, and the Black Guerrilla Family); United States v. Shryock, 342 F. 3d 948, 961 (CA9 2003) (detailing rise of Mexican Mafia inside the California prison system).

8

See, e. g., id., at 962-969 (describing a host of murders and attempted murders by a handful of Mexican Mafia members); United States v. Silverstein, 732 F. 2d 1338, 1341-1342 (CA7 1984) (describing murder of a black inmate by members of the Aryan Brotherhood); State v. Kell, 61 P. 3d 1019, 1024-1025 (Utah 2002) (describing fatal stabbing of a black inmate by two white supremacists); State v. Farmer, 126 Ariz. 569, 570-571, 617 P. 2d 521, 522-523 (1980) (en banc) (describing murder of a black inmate by members and recruits of the Aryan Brotherhood).

9

Specifically, Johnson testified:

“I was incarcerated at Calipatria before the major riot broke out there with Mexican and black inmates.... If I would have stayed there, I would have been involved in that because you have four facilities there and each facility went on a major riot and a lot of people got hurt and injured just based on your skin color. I’m black, and if I was there I would have been hurt.” App. 102a (emphasis added).

10

See id., at 245a-246a (Cambra declaration) (“If race were to be disregarded entirely, however, I am certain, based upon my experience with CDC prisoners, that... there will be fights in the cells and the problems will emanate onto the prison yards”); id., at 250a-251a (Schulteis declaration) (“At CSP-Lancaster, if we were to disregard the initial housing placement [according to race], then I am certain there would be serious violence among inmates. I have worked in five different CDC institutions and this would be true for all of them”),

11

The majority’s sole empirical support for its speculation is a study of Texas prison desegregation that found the rate of violence higher in racially segregated double cells. Ante, at 507-508 (citing Trulson & Mar-quart, The Caged Melting Pot: Toward an Understanding of the Consequences of Desegregation in Prisons, 36 Law & Soc. Rev. 743, 774 (2002)). However, the study’s authors specifically note that Texas — like California — does not integrate its “initial diagnostic facilities” or its “transfer facilities.” See id., at 753, n. 13. Thus the study says nothing about the violence likely to result from integrating cells when inmates are thrown together for brief periods during admittance or transfer. What the study does say is that, once Texas has had the time to gather inmate-related information and make more permanent housing assignments, racially integrated cells may be the preferred option. But California leaves open that door: Inmates are generally free to room with whomever they like on a permanent basis.

12

See Oklahoma Dept, of Corrections, Policies and Procedures, Operations Memorandum No. OP-030102, Inmate Housing (Sept. 16, 2004) (“Upon arrival at the assessment and reception center ... [f]or reasons of safety and security, newly received inmates are not generally assigned randomly to racially integrated cells”) (available at http://www.doc.state. ok.us/docs/policies.htm); Texas Dept, of Criminal Justice, Security Memorandum No. SM-01.28, Assignment to General Population Two-Person Cells (June 15, 2002) (“Upon arrival at a reception and diagnostic center . . . [f]or reasons of safety and security, newly-received offenders are not generally assigned randomly to racially integrated cells due to the fact that the specific information needed to assess an offender’s criminal and victimization history is not available until after diagnostic processing has been completed”).

13

On the majority's account, deference to the judgments of prison officials in the application of strict scrutiny is presumably warranted to account for “the special circumstances [that prisons] present,” ante, at 515. See Grutter v. Bollinger, 539 U. S. 306, 328 (2003). Although I disagree that deference is normally appropriate when scrutinizing racial classifications, there is some logic to the majority’s qualification in this case because the Constitution’s demands have always been diminished in the prison context. See, e. g., Harper, 494 U. S., at 224; Abbott, 490 U. S., at 407; Turner v. Safley, 482 U. S. 78, 85 (1987).

14

The Court of Appeals cited both Turner and Lee v. Washington, 390 U. S. 333 (1968) (per curiam), for the proposition that certain constitutional protections, among them the protection against state-sponsored racial discrimination, extend to the prison setting. However, the Court of Appeals did not discuss the applicable standard of review, nor did it attempt to resolve the tension between Turner and Lee that the majority finds.

YICK WO v. HOPKINS, SHERIFF. WO LEE v. HOPKINS, SHERIFF.

ERROR TO THE SUPREME COURT OF THE STATE OF CALIFORNIA.

APPEAL FROM THE CIRCUIT COURT OF THE UNITED STATES FOE THE DISTRICT OF CALIFORNIA.

Submitted April 14, 1886. —

Decided May 10, 1886.

In a suit brought to this court from a State court which involves the constitutionality of ordinances made hy a municipal corporation in the State, this court will, when necessary, put its own independent construction upon the ordinances.

A municipal ordinance to regulate the carrying on of public laundries within the limits of the municipality violates the provisions of the Constitution of the United States, if it confers upon the municipal authorities arbitrary power, at their own will, and without regard to discretion in the legal sense of the term, to give or withhold consent as to persons or places, without regard to the competency of the persons-applying, or the propriety of the place selected, for the carrying on of the business.

An administration of a municipal ordinance for the carrying on of a lawful business within the corporate limits violates the provisions of the Constitution of the United States, if it makes arbitrary and unjust discriminations, founded on differences of race, between persons otherwise in similar circumstances.

The guarantees of protection contained in the Fourteenth Amendment to the Constitution extend to all persons within the territorial jurisdiction of the United States, without regard to differences of race, of color, or of nationality.

Thosé subjects of the Emperor of China who have the right to temporarily or permanently reside within the United States, are entitled to enjoy the protection guaranteed by the Constitution and afforded by the laws.

These two cases were argued as one and depended upon precisely the same state of facts; tbe first coming here upon a writ of error to the Supreme Court of the State of California, the second on appeal from tbe Circuit Court of the United States for that district.

The plaintiff in error, Yick Wo, on August 24, 1885, petitioned tbe Supreme Court of California for a writ of habeas corpus, alleging that be was illegally deprived of his personal liberty by the defendant as sheriff of the city and county of San Francisco.

The sheriff made return to the writ that he held the petitioner in custody by virtue of a sentence of the Police Judges Court, No. 2, of the city and county of San Francisco, whereby he was found guilty of a violation of certain ordinances of the board of supervisors of that county, and adjudged to pay a fine of $10, and, in default of payment, be imprisoned in the county jail at the rate of one day for each dollar of fine until said fine should be satisfied, and a commitment in consequence of non-payment of said fine.

The ordinances for the violation of which he had been found guilty were set out as follows:

Order No. 1569, passed May 26, 1880, prescribing the kind of buildings in which laundries may be- located.

“ The people of the city and county of San Francisco do ordain as follows:

“ Sec. 1. It shall be unlawful, from and after the passage of this order, for any person or persons to establish, maintain, or carry on a laundry within the corporate limits of the city and county of San Francisco without having first obtained the consent of the board of supervisors, except the same be located in a building constructed either of brick or stone.

“ Sec. 2. It shall be unlawful for any person to erect, build, or maintain, or cause to be erected, built, or maintained, over or. upon the roof of any building now erected or which may hereafter be erected within the limits of said city and county, any scaffolding, without first obtaining the written permission of the board of supervisors, which permit shall state fully for what purpose said scaffolding is to be erected and used, and such scaffolding shall not be used for any other purpose than that designated in such permit.

“ Sec. 3. Any person who shall violate any of the provisions of this order shall be deemed guilty of a misdemeanor, and upon conviction thereof shall be punished by a fine of not more than one thousand dollars, or by imprisonment in the county jail not more than six months, or by both such fine and imprisonment.”

Order No. 1587, passed July 28,1880, the following section: “ Sec. 68. It shall be unlawful, from and after the passage of this order, for any person or persons to establish, maintain, or carry on a laundry within the corporate limits of the city and county of San Francisco without having first obtained the consent of the board of supervisors, except the same be located in a building constructed either of brick or stone.”

The following facts were also admitted on the record: That petitioner is a native of China and came to California in 1861, and is still a subject of the Emperor of China; that he has been engaged in the laundry business in the same premises and building for twenty-two years last past; that he had a license from the board of fire wardens, dated March 3, 1884, from which it appeared “that the above described premises have been inspected by the board of fire wardens, and upon such inspection said board found all proper arrangements for carrying on the business; that the stoves, washing and drying apparatus, and the appliances for heating smoothing irons are in good condition, and that their use is not dangerous to the surrounding property from fire, and that all proper precautions have been taken to comply with the provisions of order No. 1617, defining ‘ the fire limits of the city and county of San Francisco and making regulations concerning the erection and use of buildings in said city and county,’ and of order No. 1670, ‘prohibiting the kindling, maintenance, and use of open fires in houses; ’ that' he had a certificate from the health officer that the same premises had been inspected by him, and that he found that they were properly and sufficiently drained, and that all proper arrangements for carrying on the business of a laundry, without injury to the sanitary condition of the neighborhood, had been complied with; that the city license of the petitioner was in force and expired October 1st, 1885; and that the petitioner applied to the board of supervisors, June 1st, 1885, for consent of said board to maintain and carry on his laundry, but that said board, on July 1st, 1885, refuged said consent.” It is also admitted to be true, as alleged in the petition, that, on February 24, 1880, “ there were about 320 laundries in the city and county of San Francisco, of which about 240' were owned and conducted by subjects of China, and of the whole number, viz., 320, about 310 were constructed of wood, the same material that constitutes nine-tenths of the houses in the city of San Francisco.'- The capital thus invested by the subjects of China was not less than two hundred thousand dollars, and they paid annually for rent, license, taxes, gas, and water about one hundred and eighty thousand dollars.”

It was alleged in the petition, that “ your petitioner and more than one hundred and fifty of his countrymen have been arrested upon the charge of carrying on business without having such special consent, while those who are not subjects of China, and who are conducting, eighty odd laundries under similar conditions, are left unmolested and free to enjoy the enhanced trade and profits arising from this hurtful and unfair discrimination. The business of your petitioner, and of those of his countrymen similarly .situated, is greatly impaired, and in many cases practically ruined by this system of oppression to one kind of men and favoritism to all others.”

The statement therein contained as to the arrest, &c., was admitted to be true, with the qualification only, that the eighty odd laundries referred to are in wooden buildings without scaffolds on the roofs. \ '

It was also admitted “ that petitioner and 200 of his countrymen similarly situated petitipne'd the board of supervisors for permission to continue their business in the various houses which they had been occupying and using for laundries for more than twenty years, and such petitions -were denied, and all the petitions of those who were not Chinese, with one exception of Mrs. Mary Meagles, were granted.”

By section 2 of article XI of the Constitution of California it is provided that “ any county, city, town, or township may make and enforce within its limits all such local, police, sanitary, and other regulations as are not in conflict with general laws.”

By section 14 of the Act of April 19,1856, usually known as the consolidation act, the board of supervisors is empowered, among other things, “ to provide by regulation for the prevention and summary removal of nuisances to public health, the prevention of contagious diseases; . .- . to prohibit the erection of wooden buildings within any fixed limits where the streets shall have been established and graded; ... to regulate the sale, storage, and use of gunpowder or other explosive or combustible materials and substances, and make all needful regulations for protection against fire; to make such regulations concerning the erection and use of buildings as may be necessary for the safety of the inhabitants.”

The Supreme Court of California, in the opinion pronouncing the judgment in this case, said: “ The board of supervisors, under the several statutes conferring authority upon them, has the power to prohibit or regulate all occupations which are against good morals, contrary to public order and decency, or dangerous to the public safety. Clothes washing is certainly not opposed to good morals or subversive of public order or decency, but when conducted in given localities it may be highly dangerous to the public safety. Of this fact the supervisors are made the judges, and, having taken action in the premises, we do not find that they have prohibited the establishment of laundries, but that they have, as they well might do, regulated the places at which they should be established, the character of .the buildings in which they are to be maintained, etc. The process of washing is not prohibited by thus regulating the places at which and the surroundings by which it must be exercised. The order .No. 1569 and section 68 of order No. 1587 are not in contravention of common right or unjust, unequal, partial, or oppressive, in such sense as authorizes us in this proceeding to pronounce them invalid.”

• After answering the position taken in behalf of the petitioner, that the ordinances in question had been repealed, the court added : “We have not deemed it necessary to discuss the question in the light of supposed infringement of petitioner’s rights under the Constitution of the United States, for the reason that we think the principles upon Avhich contention on that head can be based have in effect- been set at rest by the cases of Barbier v. Connolly, 113 U. S. 27, and Soon Hing v. Crowley, 113 U. S. 703.” The writ was accordingly discharged and the prisoner remanded.

In the other case the appellant, Wo Lee, petitioned for his discharge from an alleged illegal imprisonment, upon a state of facts shown upon the record, precisely similar to that in the case of Tick Wo. In disposing-of the application, the learned Circuit Judge, Sawyer, in his opinion, 26 Fed. Rep. 471, after quoting the ordinance in question, proceeded at length as follows:

“ Thus, in a territory some ten miles wide by fifteen or more miles long, much of it still occupied as mere farming and pasturage lands, and much of it unoccupied sand banks, in many places without a building within a quarter or half a mile of each other, including the isolated and almost wholly unoccupied Goat Island, the right to carry on this, when properly guarded, harmless and necessary occupation, in a wooden building, is not made to depend upon any prescribed conditions giving a right to anybody complying with them, but upon the consent or arbitrary will of the board of supervisors. In three-fourths of the territory covered by the ordinance there is no more need of prohibiting or regulating laundries than if they were located in any portion of the farming regions of the State. Hitherto the regulation of laundries has been limited to the thickly settled portions of the city. . Why this unnecessary extension of the limits affected, if not designed to prevent the establishment of laundries, after a compulsory removal from their present locations, within practicable reach of the customers or their proprietors? And the uncontradicted petition shows that all Chinese applications are, in fact, denied, and 'those of Caucasians granted — thus, in fact, making the discriminations, in the administration of the ordinance, which its terms permit. The fact that the right to give consent is reserved in the ordinance shows that carrying on the laundry business in wooden buildings is not deemed of itself necessarily dangerous. It must be apparent to every well-informed mind that a fire, properly guarded, for laundry purposes, in a wooden building, is just as necessary, and no more dangerous, than a fire for cooking purposes or for warming a house. If the ordinance under consideration is valid, then the board of supervisors can pass a valid ordinance preventing the maintenance, in a wooden building, of a cooking stove, heating apparatus, or a restaurant, within the boundaries of the city and county of San Francisco, without the consent of that body, arbitrarily given or withheld, as their prejudices or other motives may dictate. If it is competent for the board of supervisors to pass a valid ordinance prohibiting the inhabitants of San Francisco from following any ordinary, proper, and necessary calling within the limits of the city and county, except at its arbitrary and unregulated discretion and special consent, and it can do so if this ordinance is valid, then it seems to us that there has been a wide departure from the principles that have heretofore been supposed to guard and protect the rights, property, and liberties of the American people. And if, by an ordinance, general in its terms and form, like the one in question, by reserving an arbitrary discretion in the enacting^ body to grant or deny permission to engage in a proper and necessary calling, a discrimination against any class can be made in its execution, thereby evading and, in effect, nullifying the provisions of the National Constitution, then the insertion of provisions to guard the rights of every cláss and person in that instrument was a vain and futile act. The effect of the execution of this ordinance in the manner indicated in the record would seem to be necessarily to close up the many Chinese laundries now existing, or compel their owners to pull down their present buildings and reconstruct of brick or stone, or to drive them outside the city and county of San Francisco, to the adjoining counties, beyond the convenient reach of customers, either of which results would be little short of absolute confiscation of the large amount of property shown to be now, and to have been for a long time, invested in these occupations. If this would not be depriving such parties of their property without due process of law, it would be difficult to say what would effect that prohibited result. The necessary tendency, if not the specific purpose, of this ordinance, and of enforcing it in the manner indicated in the record, is to drive out of business all the numerous small laundries, especially those owned by Chinese, and give a monopoly of the business to the large institutions established and carried on by means of large associated Caucasian capital. If the facts appearing on the face of the ordinance, on the petition and return, and admitted in the case, and shown by the notorious public and municipal history of the times, indicate a purpose to drive out the Chinese laundrymen, and not merely to regulate the business for the public safety, does it not disclose a case of violation of the provisions of the Fourteenth Amendment to the National Constitution, and of the treaty between the United States and China, in more than one particular ? . . .If this means prohibition of the occupation, and destruction of the business and property of the Chinese laundrymen in San Francisco — and it seems to us this must be the effect of executing the ordinance — and not merely the proper regulation of the business, then there is discrimination and a violation of other highly important rights secured by the Fourteenth Amendment and the treaty. That it does mean prohibition, as to the Chinese, it seems to us must be apparent to every citizen of San Francisco who has been here long enough to be familiar with the cause of an active and aggressive branch of public opinion and of public notorious events. Can a court be blind to what must be necessarily known- to every intelligent person in the State? See Ah Kow v. Nunan, 5 Sawyer, 552, 560: Sparrow v. Strong, 3 Wall, 97, 104; Brown v. Piper, 91 U. S. 37, 42.”

But, in deference to the decision of the Supreme Court of California in the case of Tick'Wo, and contrary to his own .opinion as thus expressed, the circuit judge discharged the writ and remanded the prisoner.

Mr. Hall McAllister, Mr. L. H. Van Schaick, and Mr. D. L. Smoot for plaintiffs in errorv

Mr. Alfred OlcwTce and Mr. M. G. Sieberst for defendant in error.

We claim that the city has power to adopt the section we are examining under article XI, section 11 of the Constitution “ to make and enforce all such local police, sanitary and other regulations as are not in conflict with general laws.” The police power of the State does extend to the regulation of this business by excluding it from certain limits, as shown by In re McClain, 61 Cal. 436; In re Chin Yan, 60 Cal. 78; In re Ah Sing, 59 Cal. 404; The Slaughter-House Cases, 16 Wall. 36, 62, et seq.; Ailstock v. Paige, 77 Va. 386; In re Lester, 77 Va. 663; Commonwealth v. Merriam, 136 Mass. 433; Muller v. Commissioners, 89 N. C. 171; State v. Mayor, 15 Vroom (44 N. J. Law), 114; State v. Fay, 15 Vroom (44 N. J. Law), 474; Commonwealth v. Whelan, 134 Mass. 206; In re Liquor Locations, 13 R. I. 733; State v. Tarver, 11 Lea, 658.

Under our State constitution, the legislature is prohibited by art. IY., sec. 25, sub. 2, from exercising the local police power; but the power which is denied to the legislature is vested by art. XI., sec. 11, in the municipal corporations throughout the State. In re Stewart, 61 Cal. 374; In re Moynier, 65 Cal. 33; In re Soon Hing, March 13, 1884, by Supreme Court in bank, not reported; In re Walters, 65 Cal. 269; Barbier v. Connolly, 113 U. S. 27; Soon Hing v. Crowley, 113 U. S. 703.

The police power is indestructible and inalienable, and being (so far as the regulation of local matters) denied to the legislature, it must reside in the municipalities. The sovereign people have located this power in the municipalities, and it is now too late to question its existence. See observations by Taney, C. J., in Ohio Life Ins. Co. v. Debolt, 16 How. 416, 428.

In addition to the cases heretofore cited, we refer to the following as recent illustrations of the extent of the police power: Butchers’ Union Co. v. Crescent City Co., 111 U. S. 746; Foster v. Kansas, 112 U. S. 201; Missouri Pacific Railway v. Humes, 115 U. S. 512.

Admitting for the sake of argument that the laundry of petitioner was not a fully developed common-law nuisance, we say the State has power to regulate it, as was shown in Barbier v. Connolly, 113 U. S. 27. The washing of Mr. Barbier was not a nuisance, but it was regulated. See also, In re Delaney, 43 Cal. 478.

It has been held that “ the State may construe her own laws.” Hall v. De Cuir, 95 U. S. 504, 515. This is what the State has done. And because some other State (Maryland for instance, in Baltimore v. Radecke, 49 Maryland, 217) has taken a different view, it does not follow that the construction by the California Courts of their laws should be reversed. We do not question the right of Maryland to make or administer her laws. This decision was presented to the Supreme Court of California in Yick Wo, the case at bar, and our court declined to follow the Supreme Court of Maryland, and adhered to the contrary rule which had long been in force in our State. Ought we to disregard the Supreme Court of California, and follow the Supreme Court of Maryland ? Can this court reverse the Supreme Court of California because it refuses to follow the Supreme Court of Maryland and adheres to its own decisions? In re Frazer, 54 Cal. 94; In re Johnson, 62 Cal. 263.

No disguise will conceal the fact that there is a conflict of authority upon the question we are examining, as will be seen on inspection of a few of the decisions which treat the question at bar.

Decisions restraining the police power of the State. — (1878). Baltimore v. Radecke, 49 Maryland, 217; (1882). July, In re Quong Wo, 7 Sawyer, 526, 531.

Decisions asserting the police power of the State. — (1871), In re Ruth, 32 Iowa, 250; (1871), Whitten v. Covington, 43. Geo. 421; (1872), State v. Court, etc., 7 Vroom (36 N. J. Law), 72; (1873), Groesch v. State, 42 Ind. 547; (1873), State v. Ludington, 33 Wis. 107; (1875), Rohrbacker v. Jackson, 51 Mississippi, 735; (1876), Kansas Pacific Railroad Co. v. Riley, 16 Kansas, 573; (1879), Eureka v. Davis, 21 Kansas, 578; (1881), Pleuler v. State, 11 Neb. 547; (1883), State v. Brown, 19 Fla. 563.

The Fourteenth Amendment bécame a part of the Constitution July 28, 1868, and yet we find thé States from that time to this asserting and exercising this power.

Mr. Justice Matthews

delivered the opinion of the court.

In the case of the petitioner, brought here by writ of error to the Supreme Court of California, our jurisdiction is limited to the question, whether the plaintiff in error has been denied a right in violation of the Constitution, laws, or treaties of the United States. The question whether his imprisonment is illegal, under the constitution and laws of the State, is not open to us. And although that question might have been considered in the Circuit Court in the application made to it, and by this court on appeal from its order, yet judicial propriety is best consulted by accepting the judgment of the State court upon the points involved in that inquiry.

That, however, does not preclude this court from putting upon the ordinances of the supervisors of the county and city of San Francisco an independent construction; for the determination of the question whether the proceedings under these ordinances and in enforcement of them are in conflict with the Constitution and laws of the United States, necessarily involves the meaning of the ordinances, which, for that purpose, we are required to ascertain and adjudge.

We are consequently constrained, at the outset, to differ from the Supreme Court of California upon the real meaning of the ordinances in question. That court considered these ordinances as vesting in the board of supervisors a not unusual discretion in granting or withholding their assent to the use of wooden buildings as laundries, to be exercised in reference to the circumstances of each case, with a view to the protection of the public against the dangers of fire. We are not able to concur in that interpretation of the power conferred upon the supervisors. There is nothing in the ordinances which points to such a regulation of the business of keeping and conducting laundries. They seem intended to confer, and actually do confer, not a discretion to be exercised upon a consideration of the circumstances of each case, but a naked and arbitrary power to give or withhold consent, not only as to places, but as to persons. So that, if an applicant for such consent, being in every way a competent and qualified person, and having complied with every reasonable condition demanded by any public ^interest, should, failing to obtain the requisite consent of the supervisors to the prosecution of his business, apply for redress by the judicial process of mandamus, to require the supervisors to consider and act upon his case, it would be a sufficient answer for them to say that the law had conferred upon them authority to withhold their assent, without reason and without responsibility. The power given to them is not confided to their discretion in the legal sense of that term, but is granted to their mere will. It is purely arbitrary, and acknowledges neither guidance nor restraint.

This erroneous view of the ordinances in question led the Supreme Court of California into the further error of holding that they were justified by the decisions of this court in the cases of Barbier v. Connolly, 113 U. S. 27, and Soon Hing v. Crowley, 113 U. S. 703. In both of these cases the ordinance involved was simply a prohibition to carry on the washing and ironing of clothes in public laundries and washhouses, within certain prescribed limits of the city and county of San Francisco, from ten o’clock at night until six o’clock in the morning of the following day. This provision was held to be purely a police regulation, within the competency of any municipality possessed of the ordinary powers belonging to such bodies ; a necessary measure of precaution in a city composed largely of wooden buildings like San Francisco, in the application of which there was no invidious discrimination against any one within the prescribed limits, all- persons engaged in the same business being treated alike, and subject to the same restrictions, and entitled to the same privileges, under similar conditions.

For these reasons, that ordinance was adjudged not to be within the prohibitions of the Fourteenth Amendment to the Constitution of the United States, which, it was said, in the first case cited, “undoubtedly intended not only that there should be no arbitrary deprivation of life or liberty, or arbitrary spoliation of property, but that equal protection and security should be given to all under like circumstances in the enjoyment of their personal and civil rights; that all persons should be equally entitled to pursue their happiness and acquire and enjoy property; that they should have like access to the courts of the country for the protection of their persons and property, the prevention and redress of wrongs, and the enforcement of contracts; that no impediment should be interposed to the pursuits .of any one, except as applied to the same pursuits by others under like circumstances; that no greater burdens should be laid upon one than are laid upon others in the same calling and condition; and that in the administration of criminal justice no different or higher punishment should be imposed upon one than such as is prescribed to all for like offences.” “ Class legislation, discriminating against some and favoring others, is prohibited, but legislation which, in carrying out a public purpose, is limited in its application, if within the sphere of its operation it affects alike all persons similarly situated, is not within the amendment.”

The ordinance drawn in question in the present case is of a very different character'. It does not prescribe a rule and conditions for the regulation of the use of property for laundry purposes, to which all similarly situated may conform. It allows without restriction the use for such purposes of buildings of .brick or stone; but, as to wooden, buildings, constituting nearly all those in previous use, it divides the owners or occupiers into two classes, not having respect to their .personal character and qualifications for the business, nor the situation and nature and adaptation of the buildings themselves, but merely by an arbitrary line, on one side of which are those who are permitted to pursue their industry by the mere will and consent of the supervisors, and on the other those from whom that consent is withheld, at their mere will and pleasure. And both classes are alike only in this, that they are tenants at will, under the supervisors, of their means of living. The ordinance, therefore, also differs from the not unusual case, where discretion is lodged by law in public officers or bodies to grant or withhold licenses to keep taverns, or places for the sale of spirituous liquors, and the like, when one of the conditions is that the applicant shall be a fit person for the exercise of the privilege, because in such cases the fact of fitness is submitted to the judgment of the officer, and calls for the exercise of a discretion of a judicial nature.

The rights of the petitioners, as affected by the proceedings of which they complain, are not less, because they are aliens and subjects of the Emperor of China. By the third article of the treaty between this Government and that of China, concluded November 17, 1880, 22 Stat. 827, it is stipulated: “ If Chinese laborers, or Chinese of any other class, now either permanently or temporarily residing in the territory of the United States, meet with ill treatment at the hands of any other persons, the Government of the United States will exert all its powers to devise measures for their protection, and to secure to them the same rights, privileges, immunities and exemptions as may be enjoyed by the citizens or subjects of the most favored nation, and to which they are entitled by treaty.”

The Fourteenth Amendment to the Constitution is not confined to the protection of citizens. It says : “ Nor shall any State deprive any person of life, liberty, or property without due process of law ; nor deny to any person within its jurisdiction the equal protection of the laws.” These provisions are universal in their application, to all persons within the territorial jurisdiction, without regard to any differences of race, of color, or of nationality ; and the equal protection of the laws is a pledge of the protection of equal laws. It is accordingly enacted by § 1977 of the Kevised Statutes, that “ all persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts, to sue, be parties, give evidence, and to the full and equal benefit of all laws and proceedings .for the security of persons and property as is enjoyed by white citizens and shall be subject to like punishment, pains, penalties, taxes, licenses, and exactions of every kind, and to no other.” The questions we have to consider and decide in these cases, therefore, are to be treated as involving the rights .of every citizen of the United States equally with those of the strangers and aliens who now invoke the jurisdiction of the court.

It is contended on the part of the petitioners, that the ordinances for violations of which they are severally sentenced to imprisonment, are void on their face, as being within the prohibitions of the Fourteenth Amendment; and, in the alternative, if not so, that they are void by reason of their administration, operating unequally, so as to punish in the present petitioners what is permitted to others as lawful, without any distinction of circumstances — an unjust and illegal discrimination, it is claimed, which, though not made expressly by the ordinances is made possible by them.

When we consider the nature and the theory of our institutions of government, the principles upon which they are sup.posed to rest, and review the history of their development, we are constrained to conclude that they do not mean to leave room for the play and action of purely personal and arbitrary power. Sovereignty itself is, of course, not subject to law, for it is the author and source of law; but in our system, while sovereign powers are delegated to the agencies of government, sovereignty itself remains with the people, by whom and for whom all government exists and acts. And the law is the definition and limitation of power. It is, indeed, quite true, that there must always be lodged somewhere, and in some person or body, the authority of final decision; and in many cases of mere administration the responsibility is purely'political, no appeal lying except to the ultimate tribunal of the public judgment, exercised either in the pressure of opinion or by means of the suffrage. But the fundamental rights to life, liberty, and the pursuit of happiness, considered as individual possessions, are secured by those maxims of constitutional law which are the monuments showing the victorious progress of the race in securing to men the blessings of civilization under the reign of just and equal laws, so that, in the famous language of the Massachusetts Bill of Rights, the government of the commonwealth “may be a government of laws and not of men.” For, the very idea that one man may be compelled to hold his life, or the means of living, or any material right essential to the enjoyment of life, at the mere will of another, seems to be intolerable in any country where freedom prevails, as being the essence of slavery itself.

There are many illustrations that might be given of this truth, which would make manifest that it was self-evident in the light of our system of jurisprudence. The case of the political franchise of voting is one. Though not regarded strictly as a natural right, but as a privilege merely conceded by society according to its will, under certain conditions, nevertheless it is regarded as a fundamental political right, because preservative of all rights.

In reference to that right, it was declared by the Supreme Judicial Court of Massachusetts, in Capen v. Foster, 12 Pick. 485, 489, in the words of Chief Justice Shaw, “that in all cases where the constitution has conferred a political right or privilege, and where the constitution has not particularly designated the manner in which that right is to be exercised, it is clearly within the just and constitutional limits of the legislative power, to adopt any reasonable and uniform regulations, in regard to the time and mode of exercising that right, which are designed to secure and, facilitate the exercise of such right, in a prompt, orderly, and convenient manner; ” nevertheless, “ such a construction would afford no warrant for such an exercise of legislative power, as, under the pretence and color of regulating, should subvert or injuriously restrain the right itself.” It has accordingly been held generally in the States, that, whether the particular provisions of an act of legislation, establishing means for ascertaining the qualifications of those entitled to vote, and making previous registration in lists of such, a condition precedent to the exercise of the right, were or were not reasonable regulations, and accordingly valid or void, was always open to inquiry, as a judicial question. See Daggett v. Hudson, 1 Western Reporter, 789, decided by the Supreme Court of Ohio, where many of the cases are collected; Monroe v. Collins, 17 Ohio St. 665.

The same principle has beeh more freely extended to the quasi-legislative acts of inferior municipal bodies, in respect to which it is an ancient jurisdiction of judicial tribunals to pronounce upon the reasonableness and consequent validity of their by-laws. In respect to these, it was the doctrine, that every by-law must be reasonajffe, not inconsistent with the charter of the corporation, nor with any statute of Parliament, nor with the general principles of the common law of the land, particularly those having relation to the liberty of the subject or the rights of private property. Dillon on Municipal Corporations, 3d ed., § 319, and cases cited in notes. Accordingly, in the case of The State of Ohio ex rel. &c. v. The Cincinnati Gas-Light and Coke Company, 18 Ohio St. 262, 300, an ordinance of the city council purporting to fix the price to be charged for gas, under an authority of law giving discretionary power to do so, was held to be bad, if passed in bad faith, fixing an unreasonable price, for the fraudulent purpose of compelling the gas company to submit to an unfair appraisement of their works. And a similar question, very pertinent to the one in the present cases, was decided by the Court of Appeals of Maryland, in the case of the City of Baltimore v. Radecke, 49 Maryland, 217. In that case the defendant had erected and used a steam engine, in the prosecution of his business as a carpenter and box-maker in the city of Baltimore, under a permit from the mayor and city council, which contained a condition that the engine was “to be removed after six months’ notice to that effect from the mayor.” After such notice and refusal to conform to it, a suit was instituted to recover the penalty provided by the ordinance, to restrain the prosecution of which a bill in equity was filed. The court holding the opinion that “ there may be a case in which an ordinance, passed under grants of power like those we have cited, is so clearly unreasonable, so arbitrary, oppressive, or partial, as to raise the presumption that the legislature never intended to confer the power to pass it, and to justify the courts in interfering and setting it aside as a' plain abuse of authority,” it proceeds to speak, with regard to the ordinance in question, in relation to the use of steam engines, as follows: “It does not profess to prescribe regulations for their construction, location, or use, nor require such precautions and safeguards to be provided by those who own and use them as are best calculated to render them less dangerous to life and propei’ty, nor does it restrain their use in box factories and other similar establishments within certain defined limits, nor in any other wray attempt to promote their safety and security without destroying their usefulness. But it commits to the unrestrained will of a single public officer the power to notify every person who now employs a steam engine in the prosecution of any business in the city of Baltimore, to cease to do so, and, by providing compulsory fines for every day’s disobedience of such notice and order of removal, renders his power over the use of steam in that city practically absolute, so that he may prohibit its use altogether. But if he should not choose to do this, but only to act in particular cases, there is nothing in the ordinance to guide or control his action. It lays down no mies by which its impartial execution can be secured or partiality and oppression prevented. It is clear that giving and enforcing these notices may, and quite likely will, bring ruin to the business of those, against whom they are directed, while others, from whom they are withheld, may be actually benefited by what is thus done to their neighbors; and, when we remember that this action or non-action may proceed from enmity or prejudice, from partisan zeal or animosity, from favoritism and other improper influences and motives easy of concealment and difficult to be detected and exposed, it becomes unnecessary to suggest or to comment upon the injustice capable of being brought under cover of such a power, for that becomes apparent to every one who gives to the subject a moment’s consideration. In- fact, an ordinance which clothes a single individual with such power hardly falls within the domain of law, and we are constrained to pronounce it inoperative and void.”

This conclusion, and the reasoning on which it is based, are deductions from the face of the ordinance, as to its necessary tendency and ultimate actual operation. In the present cases we are not obliged to reason from the probable to the actual, and pass upon the validity of the ordinances complained of, as tried merely by the opportunities which their terms afford, of unequal and unjust discrimination in their administration. For the cases present the ordinances in actual operation, and the facts shown establish an administration directed so exclusively against a particular class1 of persons as to warrant and require the conclusion, that, whatever may have been the intent of the ordinances as adopted, they are applied by the public authorities charged with their administration, and thus representing the State itself, wfth a mind so unequal and oppressive as to amount to a practical denial by the State of that equal protection of the laws which is secured to the petitioners, as to all other persons, by the broad and benign provisions of the Fourteenth Amendment to the Constitution of the United States. Though the law itself be fair on its face and impartial in appearance, yet, if it is applied and administered by public authority with an evil eye and an unequal hand, so as practically to make unjust and illegal discriminations between persons in similar circumstances, material to their rights, the denial of equal justice is still within the prohibition of the Constitution. This principle of interpretation has been sanctioned by this court in Henderson v. Mayor of New York, 92 U. S. 259; Chy Lung v. Freeman, 92 U. S. 275; Ex parte Virginia, 100 U. S. 339; Neal v. Delaware, 103 U. S. 370; and Soon Hing v. Crowley, 113 U. S. 703.

The present cases, as shown by the facts disclosed in the record, are within this class. It appears that both petitioners have complied with every requisite, deemed by the law or by the public officers charged with its administration, necessary for the protection of neighboring property from fire, or as a precaution against injury to the public health. No reason whatever, except the will of the supervisors, is assigned why they should not be permitted to carry on, in the accustomed manner, their harmless and useful occupation, on which they depend for a livelihood. And while this consent of the supervisors is withheld from them and from two hundred others who have also petitioned, all of whom happen to be Chinese subjects, eighty others, not Chinese subjects, are permitted to carry on the same business under similiar conditions. The fact of this discrimination is admitted. No reason for it is shown, and the conclusion cannot be resisted, that no reason for it exists except hostility to the race and nationality to which the petitioners belong, and which in the eye of the law is not justified. The discrimination is, therefore, illegal, and the public administration which enforces it is a denial of the equal protection of the laws and a violation of the Fourteenth Amendment of the Constitution. The imprisonment of the petitioners is, therefore, illegal, and they must be discharged. To this end,

The judgment of the Supreme Court of California in the case of Yick Wo, and that of the Girouit Court of the United States for the District of Odlifornia ini the case of Wo Lee, . are severally reversed, and the cases remanded, each to the pi'oper court, with directions to discharge the petitioners from custody cmd imprisonment.

          Respondents Harley and Sellers, both Negroes (hereinafter respondents), whose applications to become police officers in the District of Columbia had been rejected, in an action against District of Columbia officials (petitioners) and others, claimed that the Police Department's recruiting procedures, including a written personnel test (Test 21), were racially discriminatory and violated the Due Process Clause of the Fifth Amendment, 42 U.S.C. § 1981, and D.C.Code § 1-320. Test 21 is administered generally to prospective Government employees to determine whether applicants have acquired a particular level of verbal skill. Respondents contended that the test bore no relationship to job performance and excluded a disproportionately high number of Negro applicants. Focusing solely on Test 21, the parties filed cross-motions for summary judgment. The District Court, noting the absence of any claim of intentional discrimination, found that respondents' evidence supporting their motion warranted the conclusions that (a) the number of black police officers, while substantial, is not proportionate to the city's population mix; (b) a higher percentage of blacks fail the test than whites; and (c) the test has not been validated to establish its reliability for measuring subsequent job performance. While that showing sufficed to shift the burden of proof to the defendants in the action, the court concluded that respondents were not entitled to relief, and granted petitioners' motion for summary judgment, in view of the facts that 44% Of new police recruits were black, a figure proportionate to the blacks on the total force and equal to the number of 20- to 29-year-old blacks in the recruiting area; that the Police Department had affirmatively sought to recruit blacks, many of whom passed the test but failed to report for duty; and that the test was a useful indicator of training school performance (precluding the need to show validation in terms of job performance) and was not designed to, and did not, discriminate against otherwise qualified blacks. Respondents on

Page 230

appeal contended that their summary judgment motion (which was based solely on the contention that Test 21 invidiously discriminated against Negroes in violation of the Fifth Amendment) should have been granted. The Court of Appeals reversed, and directed summary judgment in favor of respondents, having applied to the constitutional issue the statutory standards enunciated in Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158, which held that Title VII of the Civil Rights Act of 1964, as amended, prohibits the use ofests that operate to exclude members of minority groups, unless the employer demonstrates that the procedures are substantially related to job performance. The court held that the lack of discriminatory intent in the enactment and administration of Test 21 was irrelevant; that the critical fact was that four times as many blacks as whites failed the test; and that such disproportionate impact sufficed to establish a constitutional violation, absent any proof by petitioners that the test adequately measured job performance. Held:

          1. The Court of Appeals erred in resolving the Fifth Amendment issue by applying standards applicable to Title VII cases. Pp. 238-248.

          (a) Though the Due Process Clause of the Fifth Amendment contains an equal protection component prohibiting the Government from invidious discrimination, it does not follow that a law or other official act is unconstitutional Solely because it has a racially disproportionate impact regardless of whether it reflects a racially discriminatory purpose. Pp. 239-245.

          (b) The Constitution does not prevent the Government from seeking through Test 21 modestly to upgrade the communicative abilities of its employees rather than to be satisfied with some lower level of competence, particularly where the job requires special abilities to communicate orally and in writing; and respondents, as Negroes, could no more ascribe their failure to pass the test to denial of equal protection than could whites who also failed. Pp. 245-246.

          (c) The disproportionate impact of Test 21, which is neutral on its face, does not warrant the conclusion that the test was a purposely discriminatory device, and on the facts before it the District Court properly held that any inference of discrimination was unwarranted. P. 246.

          (d) The rigorous statutory standard of Title VII involves a more probing judicial review of, and less deference to, the seemingly reasonable acts of administrators and executives than is

Page 231

appropriate under the Constitution where, as in this case, special racial impact but no discriminatory purpose is claimed. Any extension of that statutory standard should await legislative prescription. Pp. 246-248.

          2. Statutory standards similar to those obtaining under Title VII were also satisfied here. The District Court's conclusion that Test 21 was directly related to the requirements of the police training program and that a positive relationship between the test and that program was sufficient to validate the test (wholly aside from its possible relationship to actual performance as a police officer) is fully supported on the record in this case, and no remand to establish further validation is appropriate. Pp. 248-252.

          168 U.S.App.D.C. 42, 512 F.2d 956, reversed.

          David P. Sutton, Washington, D. C., for petitioners.

          Mark L. Evans, Washington, D. C., for the federal respondents.

          Richard B. Sobol, Washington, D. C., for respondents Davis et al.

Page 232

           Mr. Justice WHITE delivered the opinion of the Court.

          This case involves the validity of a qualifying test administered to applicants for positions as police officers in the District of Columbia Metropolitan Police Department. The test was sustained by the District Court but invalidated by the Court of Appeals. We are in agreement with the District Court and hence reverse the judgment of the Court of Appeals.

          This action began on April 10, 1970, when two Negro police officers filed suit against the then Commissioner of the District of Columbia, the Chief of the District's Metropolitan Police Department, and the Commissioners of the United States Civil Service Commission.¹ An amended complaint, filed December 10, alleged that the promotion policies of the Department were racially discriminatory and sought a declaratory judgment and an injunction. The respondents Harley and Sellers were permitted to intervene, their amended complaint assert-

Page 233

ing that their applications to become officers in the Department had been rejected, and that the Department's recruiting procedures discriminated on the basis of race against black applicants by a series of practices including, but not limited to, a written personnel test which excluded a disproportionately high number of Negro applicants. These practices were asserted to violate respondents' rights "under the due process clause of the Fifth Amendment to the United States Constitution, under 42 U.S.C. § 1981 and under D.C.Code § 1-320."² Defendants answered, and discovery and

Page 234

various other proceedings followed.³Respondents then filed a motion for partial summary judgment with respect to the recruiting phase of the case, seeking a declaration that the test administered to those applying to become police officers is "unlawfully discriminatory and thereby in violation of the due process clause of the Fifth Amendment . . . ." No issue under any statute or regulation was raised by the motion. The District of Columbia defendants, petitioners here, and the federal parties also filed motions for summary judgment with respect to the recruiting aspects of the case, asserting that respondents were entitled to relief on neither constitutional nor statutory grounds.⁴ The District Court granted petitioners' and denied respondents' motions. 348 F.Supp. 15 (DC1972).

          According to the findings and conclusions of the District Court, to be accepted by the Department and to enter an intensive 17-week training program, the police recruit was required to satisfy certain physical and character standards, to be a high school graduate or its equivalent, and to receive a grade of at least 40 out of 80 on "Test 21," which is "an examination that is used generally throughout the federal service," which "was developed by the Civil Service Commission, not the Police Department,"

Page 235

and which was "designed to test verbal ability, vocabulary, reading and comprehension." Id., at 16.

          The validity of Test 21 was the sole issue before the court on the motions for summary judgment. The District Court noted that there was no claim of "an intentional discrimination or purposeful discriminatory acts" but only a claim that Test 21 bore no relationship to job performance and "has a highly discriminatory impact in screening out black candidates." Ibid. Respondents' evidence, the District Court said, warranted three conclusions: "(a) The number of black police officers, while substantial, is not proportionate to the population mix of the city. (b) A higher percentage of blacks fail the Test than whites. (c) The Test has not been validated to establish its reliability for measuring subsequent job performance." Ibid. This showing was deemed sufficient to shift the burden of proof to the defendants in the action, petitioners here; but the court nevertheless concluded that on the undisputed facts respondents were not entitled to relief. The District Court relied on several factors. Since August 1969, 44% Of new police force recruits had been black; that figure also represented the proportion of blacks on the total force and was roughly equivalent to 20- to 29-year-old blacks in the 50-mile radius in which the recruiting efforts of the Police Department had been concentrated. It was undisputed that the Department had systematically and affirmatively sought to enroll black officers many of whom passed the test but failed to report for duty. The District Court rejected the assertion that Test 21 was culturally slanted to favor whites and was "satisfied that the undisputable facts prove the test to be reasonably and directly related to the requirements of the police recruit training program and that it is neither so designed nor operates (Sic ) to discriminate

Page 236

against otherwise qualified blacks' Id., at 17. It was thus not necessary to show that Test 21 was not only a useful indicator of training school performance but had also been validated in terms of job performance "The lack of job performance validation does not defeat the Test, given its direct relationship to recruiting and the valid part it plays in this process." Ibid. The District Court ultimately concluded that "(t)he proof is wholly lacking that a police officer qualifies on the color of his skin rather than ability" and that the Department "should not be required on this showing to lower standards or to abandon efforts to achieve excellence." ⁵ Id., at 18.

          Having lost on both constitutional and statutory issues in the District Court, respondents brought the case to the Court of Appeals claiming that their summary judgment motion, which rested on purely constitutional grounds, should have been granted. The tendered constitutional issue was whether the use of Test 21 invidiously discriminated against Negroes and hence denied them due process of law contrary to the commands of the Fifth Amendment. The Court of Appeals, addressing that issue, announced that it would be guided by Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), a case involving the interpretation and application of Title VII of the Civil Rights Act of 1964, and held that the statutory standards elucidated in that case were to govern the due process question tendered in this one.⁶ 168 U.S.App.D.C. 42,

Page 237

512 F.2d 956 (1975). e court went on to declare that lack of discriminatory intent in designing and administering Test 21 was irrelevant; the critical fact was rather that a far greater proportion of blacks four times as many failed the test than did whites. This disproportionate impact, standing alone and without regard to whether it indicated a discriminatory purpose, was held sufficient to establish a constitutional violation, absent proof by petitioners that the test was an adequate measure of job performance in addition to being an indicator of probable success in the training program, a burden which the court ruled petitioners had failed to discharge. That the Department had made substantial efforts to recruit blacks was held beside the point and the fact that the racial distribution of recent hirings and of the Department itself might be roughly equivalent to the racial makeup of the surrounding community, broadly conceived, was put aside as a "comparison (not) material to this appeal." Id., at 46 n. 24, 512 F.2d, at 960 n. 24. The Court of Appeals, over a dissent, accordingly reversed the judgment of the District Court and directed that respondents' motion for partial summary judgment be granted. We granted the petition for certiorari, 423 U.S. 820, 96 S.Ct. 33, 46 L.Ed.2d 37 (1975), filed by the District of Columbia officials.⁷

Page 238

          Because the Court of Appeals erroneously applied the legal standards applicable to Title VII cases in resolving the constitutional issue before it, we reverse its judgment in respondents' favor. Although the petition for certiorari did not present this ground for reversal,⁸ our Rule 40(1)(d)(2) provides that we "may notice a plain error not presented"; ⁹ and this is an appropriate occasion to invoke the Rule.

          As the Court of Appeals understood Title VII,¹⁰ employees or applicants proceeding under it need not concern themselves with the employer's possibly discriminatory purpose but instead may focus solely on the racially differential impact of the challenged hiring or promotion

Page 239

practices. This is not the constitutional rule. We have never held that the constitutional standard for adjudicating claims of invidious racial discrimination is identical to the standards applicable under Title VII, and we decline to do so today.

          The central purpose of the Equal Protection Clause of the Fourteenth Amendment is the prevention of official conduct discriminating on the basis of race. It is also true that the Due Process Clause of the Fifth Amendment contains an equal protection component prohibiting the United States from invidiously discriminating between individuals or groups. Bolling v. Sharpe, 347 U.S. 497, 74 S.Ct. 693, 98 L.Ed. 884 (1954). But our cases have not embraced the proposition that a law or other official act, without regard to whether it reflects a racially discriminatory purpose, is unconstitutional Solely because it has a racially disproportionate impact.

          Almost 100 years ago, Strauder v. West Virginia, 100 U.S. 303, 25 L.Ed. 664 (1880), established that the exclusion of Negroes from grand and petit juries in criminal proceedings violated the Equal Protection Clause, but the fact that a particular jury or a series of juries does not statistically reflect the racial composition of the community does not in itself make out an invidious discrimination forbidden by the Clause. "A purpose to discriminate must be present which may be proven by systematic exclusion of eligible jurymen of the proscribed race or by unequal application of the law to such an extent as to show intentional discrimination." Akins v. Texas, 325 U.S. 398, 403-404, 65 S.Ct. 1276, 1279, 89 L.Ed. 1692, 1696 (1945). A defendant in a criminal case is entitled "to require that the State not deliberately and systematically deny to members of his race the right to participate as jurors in the administration of justice." Alexander v. Louisiana, 405 U.S. 625, 628-629, 92 S.Ct. 1221, 1224, 31 L.Ed.2d 536 (1972). See also Carter v. Jury Comm'n, 396 U.S. 320, 335-

Page 240

337, 339, 90 S.Ct. 5, 526-528, 529, 24 L.Ed.2d 549, 560-561, 562 (1970); Cassell v. Texas, 339 U.S. 282, 287-290, 70 S.Ct. 629, 631-633, 94 L.Ed. 839, 847-849 (1950); Patton v. Mississippi, 332 U.S. 463, 468-469, 68 S.Ct. 184, 187, 92 L.Ed. 76, 80 (1947).

          The rule is the same in other contexts. Wright v. Rockefeller, 376 U.S. 52, 84 S.Ct. 603, 11 L.Ed.2d 512 (1964), upheld a New York congressional apportionment statute against claims that district lines had been racially gerrymandered. The challenged districts were made up predominantly of whites or of minority races, and their boundaries were irregularly drawn. The challengers did not prevail because they failed to prove that the New York Legislature "was either motivated by racial considerations or in fact drew the districts on racial lines"; the plaintiffs had not shown that the statute "was the product of a state contrivance to segregate on the basis of race or place of origin." Id., at 56, 58, 84 S.Ct., at 605, 11 L.Ed.2d, at 515. The dissenters were in agreement that the issue was whether the "boundaries . . . were purposefully drawn on racial lines." Id., at 67, 84 S.Ct., at 611, 11 L.Ed.2d, at 522.

          The school desegregation cases have also adhered to the basic equal protection principle that the invidious quality of a law claimed to be racially discriminatory must ultimately be traced to a racially discriminatory purpose. That there are both predominantly black and predominantly white schools in a community is not alone violative of the Equal Protection Clause. The essential element of De jure segregation is "a current condition of segregation resulting from intentional state action. Keyes v. School Dist. No. 1, 413 U.S. 189, 205, 93 S.Ct. 2686, 2696, 37 L.Ed.2d 548 (1973). The differentiating factor between De jure segregation and so-called De facto segregation . . . is Purpose or Intent to segregate." Id., at 208, 93 S.Ct., at 2696, 37 L.Ed.2d, at 561. See also Id., at 199, 211, 213, 93 S.Ct. at 2692, 2698, 2699, 37 L.Ed.2d, at 558, 564, 566. The Court has also recently rejected allegations of racial discrimination based solely on the statistically disproportionate racial impact of various provisions of the Social Security Act because "(t)he acceptance of appellants'

Page 241

constitutional theory would render suspect each difference in treatment among the grant classes, however lacking in racial motivation and however otherwise rational the treatment might be." Jefferson v. Hackney, 406 U.S. 535, 548, 92 S.Ct. 1724, 1732, 32 L.Ed.2d 285, 297 (1972). And compare Hunter v. Erickson, 393 U.S. 385, 89 S.Ct. 557, 21 L.Ed.2d 616 (1969), with James v. Valtierra, 402 U.S. 137, 91 S.Ct. 1331, 28 L.Ed.2d 678 (1971).

          This is not to say that the necessary discriminatory racial purpose must be express or appear on the face of the statute, or that a law's disproportionate impact is irrelevant in cases involving Constitution-based claims of racial discrimination. A statute, otherwise neutral on its face, must not be applied so as invidiously to discriminate on the basis of race. Yick Wo v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064, 30 L.Ed. 220 (1886). It is also clear from the cases dealing with racial discrimination in the selection of juries that the systematic exclusion of Negroes is itself such an "unequal application of the law . . . as to show intentional discrimination." Akins v. Texas, supra, 325 U.S., at 404, 65 S.Ct., at 1279, 89 L.Ed., at 1696. Smith v. Texas, 311 U.S. 128, 61 S.Ct. 164, 85 L.Ed. 84 (1940); Pierre v. Louisiana, 306 U.S. 354, 59 S.Ct. 536, 83 L.Ed. 757 (1939); Neal v. Delaware, 103 U.S. 370, 26 L.Ed. 567 (1881). A prima facie case of discriminatory purpose may be proved as well by the absence of Negroes on a particular jury combined with the failure of the jury commissioners to be informed of eligible Negro jurors in a community, Hill v. Texas, 316 U.S. 400, 404, 62 S.Ct. 1159, 1161, 86 L.Ed. 1559, 1562 (1942), or with racially non-neutral selection procedures, Alexander v. Louisiana, supra ; Avery v. Georgia, 345 U.S. 559, 73 S.Ct. 891, 97 L.Ed. 1244 (1953); Whitus v. Georgia, 385 U.S. 545, 87 S.Ct. 643, 17 L.Ed.2d 599 (1967). With a prima facie case made out, "the burden of proof shifts to the State to rebut the presumption of unconstitutional action by showing that permissible racially neutral selection criteria and procedures have produced the monochromatic result." Alexander, supra, 405 U.S., at 632, 92 S.Ct., at 1226, 31 L.Ed.2d, at 542. See also Turner v. Fouche, 396 U.S. 346, 361, 90 S.Ct. 532, 540, 24 L.Ed.2d 567, 579 (1970); Eubanks v. Louisiana, 356 U.S. 584, 587, 78 S.Ct. 970, 973, 2 L.Ed.2d 991, 994 (1958).

Page 242

          Necessarily, an invidious discriminatory purpose may often be inferred from the totality of the relevant facts, including the fact, if it is true, that the law bears more heavily on one race than another. It is also not infrequently true that the discriminatory impact in the jury cases for example, the total or seriously disproportionate exclusion of Negroes from jury venires may for all practical purposes demonstrate unconstitutionality because in various circumstances the discrimination is very difficult to explain on nonracial grounds. Nevertheless, we have not held that a law, neutral on its face and serving ends otherwise within the power of government to pursue, is invalid under the Equal Protection Clause simply because it may affect a greater proportion of one race than of another. Disproportionate impact is not irrelevant, but it is not the sole touchstone of an invidious racial discrimination forbidden by the Constitution. Standing alone, it does not trigger the rule, McLaughlin v. Florida, 379 U.S. 184, 85 S.Ct. 283, 13 L.Ed.2d 222 (1964), that racial classifications are to be subjected to the strictest scrutiny and are justifiable only by the weightiest of considerations.

          There are some indications to the contrary in our cases. In Palmer v. Thompson, 403 U.S. 217, 91 S.Ct. 1940, 29 L.Ed.2d 438 (1971), the city of Jackson, Miss., following a court decree to this effect, desegregated all of its public facilities save five swimming pools which had been operated by the city and which, following the decree, were closed by ordinance pursuant to a determination by the city council that closure was necessary to preserve peace and order and that integrated pools could not be economically operated. Accepting the finding that the pools were closed to avoid violence and economic loss, this Court rejected the argument that the abandonment of this service was inconsistent with the outstanding desegregation decree and that the otherwise seemingly permissible ends served by the ordinance could be impeached by demonstrating that

Page 243

racially invidious motivations had prompted the city council's action. The holding was that the city was not overtly or covertly operating segregated pools and was extending identical treatment to both whites and Negroes. The opinion warned against grounding decision on legislative purpose or motivation, thereby lending support for the proposition that the operative effect of the law rather than its purpose is the paramount factor. But the holding of the case was that the legitimate purposes of the ordinance to preserve peace and avoid deficits were not open to impeachment by evidence that the councilmen were actually motivated by racial considerations. Whatever dicta the opinion may contain, the decision did not involve, much less invalidate, a statute or ordinary having neutral purposes but disproportionate racial consequences.

          Wright v. Council of City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972), also indicates that in proper circumstances, the racial impact of a law, rather than its discriminatory purpose, is the critical factor. That case involved the division of a school district. The issue was whether the division was consistent with an outstanding order of a federal court to desegregate the dual school system found to have existed in the area. The constitutional predicate for the District Court's invalidation of the divided district was "the enforcement until 1969 of racial segregation in a public school system of which Emporia had always been a part." Id., at 459, 92 S.Ct., at 2202, 33 L.Ed.2d, at 60. There was thus no need to find "an independent constitutional violation." Ibid. Citing Palmer v. Thompson, we agreed with the District Court that the division of the district had the effect of interfering with the federal decree and should be set aside.

          That neither Palmer Nor Wright was understood to have changed the prevailing rule is apparent from Keyes v. School Dist. No. 1, supra, where the principal issue

Page 244

in litigation was whether to what extent there had been purposeful discrimination resulting in a partially or wholly segregated school system. Nor did other later cases, Alexander v. Louisiana, supra, and Jefferson v. Hackney, supra, indicate that either Palmer or Wright had worked a fundamental change in equal protection law.¹¹

          Both before and after Palmer v. Thompson, however, various Courts of Appeals have held in several contexts, including public employment, that the substantially disproportionate racial impact of a statute or official practice standing alone and without regard to discriminatory purpose, suffices to prove racial discrimination violating the Equal Protection Clause absent some justification going substantially beyond what would be necessary to validate most other legislative classifications.¹² The

Page 245

cases impressively demonstrate that there is another side to the issue; but, with all due respect, to the extent that those cases rested on or expressed the view that proof of discriminatory racial purpose is unnecessary in making out an equal protection violation, we are in disagreement.

          As an initial matter, we have difficulty understanding how a law establishing a racially neutral qualification for employment is nevertheless racially discriminatory and denies "any person . . . equal protection of the laws" simply because a greater proportion of Negroes fail to qualify than members of other racial or ethnic groups. Had respondents, along with all others who had failed Test 21, whether white or black, brought an action claiming that the test denied each of them equal protection of the laws as compared with those who had passed with high enough scores to qualify them as police recruits, it is most unlikely that their challenge would have been sustained. Test 21, which is administered generally to prospective Government employees, concededly seeks to ascertain whether those who take it have acquired a particular level of verbal skill; and it is untenable that

Page 246

the Constitution prevents the Government from seeking modestly to upgrade the communicative abilities of its employees rather than to be satisfied with some lower level of competence, particularly where the job requires special ability to communicate orally and in writing. Respondents, as Negroes, could no more successfully claim that the test denied them equal protection than could white applicants who also failed. The conclusion would not be different in the face of proof that more Negroes than whites had been disqualified by Test 21. That other Negroes also failed to score well would, alone, not demonstrate that respondents individually were being denied equal protection of the laws by the application of an otherwise valid qualifying test being administered to prospective police recruits.

          Nor on the facts of the case before us would the disproportionate impact of Test 21 warrant the conclusion that it is a purposeful device to discriminate against Negroes and hence an infringement of the constitutional rights of respondents as well as other black applicants. As we have said, the test is neutral on its face and rationally may be said to serve a purpose the Government is constitutionally empowered to pursue. Even agreeing with the District Court that the differential racial effect of Test 21 called for further inquiry, we think the District Court correctly held that the affirmative efforts of the Metropolitan Police Department to recruit black officers, the changing racial composition of the recruit classes and of the force in general, and the relationship of the test to the training program negated any inference that the Department discriminated on the basis of race or that "a police officer qualifies on the color of his skin rather than ability." 348 F.Supp., at 18.

          Under Title VII, Congress provided that when hiring

Page 247

and promotion practices disqualifying substantially disprortionate numbers of blacks are challenged, discriminatory purpose need not be proved, and that it is an insufficient response to demonstrate some rational basis for the challenged practices. It is necessary, in addition, that they be "validated" in terms of job performance in any one of several ways, perhaps by ascertaining the minimum skill, ability, or potential necessary for the position at issue and determining whether the qualifying tests are appropriate for the selection of qualified applicants for the job in question.¹³ However this process proceeds, it involves a more probing judicial review of, and less deference to, the seemingly reasonable acts of administrators and executives than is appropriate under the Constitution where special racial impact, without discriminatory purpose, is claimed. We are not disposed to adopt this more rigorous standard for the purposes

Page 248

of applying the Fifth and the Fourteenth Amendments in cases such as this

          A rule that a statute designed to serve neutral ends is nevertheless invalid, absent compelling justification, if in practice it benefits or burdens one race more than another would be far-reaching and would raise serious questions about, and perhaps invalidate, a whole range of tax, welfare, public service, regulatory, and licensing statutes that may be more burdensome to the poor and to the average black than to the more affluent white.¹⁴

            Given that rule, such consequences would perhaps be likely to follow. However, in our view, extension of the rule beyond those areas where it is already applicable by reason of statute, such as in the field of public employment, should await legislative prescription.

          As we have indicated, it was error to direct summary judgment for respondents based on the Fifth Amendment.

          We also hold that the Court of Appeals should have affirmed the judgment of the District Court granting the motions for summary judgment filed by petitioners and the federal parties. Respondents were entitled to relief on neither constitutional nor statutory grounds.

Page 249

          The submission of the defendants in the District Court was that Test 21 complied with all applicable statutory as well as constitutional requirements; and they appear not to have disputed that under the statutes and regulations governing their conduct standards similar to those obtaining under Title VII had to be satisfied.¹⁵ The District Court also assumed that Title VII standards were to control the case identified the determinative issue as whether Test 21 was sufficiently job related and proceeded to uphold use of the test because it was "directly related to a determination of whether the applicant possesses sufficient skills requisite to the demands of the curriculum a recruit must master at the police academy." 348 F.Supp., at 17. The Court of Appeals reversed because the relationship between Test 21 and training school success, if demonstrated at all, did not satisfy what it deemed to be the crucial requirement

Page 250

of a direct relationship between performance on Test 21 and performance on the policeman's job.

          We agree with petitioners and the federal parties that this was error. The advisability of the police recruit training course informing the recruit about his upcoming job, acquainting him with its demands, and attempting to impart a modicum of required skills seems conceded. It is also apparent to us, as it was to the District Judge, that some minimum verbal and communicative skill would be very useful, if not essential, to satisfactory progress in the training regimen. Based on the evidence before him, the District Judge concluded that Test 21 was directly related to the requirements of the police training program and that a positive relationship between the test and training-course performance was sufficient to validate the former, wholly aside from its possible relationship to actual performance as a police officer. This conclusion of the District Judge that training-program validation may itself be sufficient is supported by regulations of the Civil Service Commission, by the opinion evidence placed before the District Judge, and by the current views of the Civil Service Commissioners who were parties to the case.¹⁶ Nor is the

Page 251

conclusion closed by either Griggs or Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975); and it seems to us the much more sensible construction of the job-relatedness requirement.

          The District Court's accompanying conclusion that Test 21 was in fact directly related to the requirements of the police training program was supported by a validation study, as well as by other evidence of record; ¹⁷

Page 252

and we are not convinced that this conclusion was erroneous.

          The federal parties, whose views have somewhat changed since the decision of the Court of Appeals and who still insist that training-program validation is sufficient, now urge a remand to the District Court for the purpose of further inquiry into whether the training-program test scores, which were found to correlate with Test 21 scores, are themselves an appropriate measure of the trainee's mastership of the material taught in the course and whether the training program itself is sufficiently related to actual performance of the police officer's task. We think a remand is inappropriate. The District Court's judgment was warranted by the record before it, and we perceive no good reason to reopen it, particularly since we were informed at oral argument that although Test 21 is still being administered, the training program itself has undergone substantial modification in the course of this litigation. If there are now deficiencies in the recruiting practices under prevailing Title VII standards, those deficiencies are to be directly addressed in accordance with appropriate procedures mandated under that Title.

          The judgment of the Court of Appeals accordingly is reversed.

          So ordered.

          Mr. Justice STEWART joins Parts I and II of the Court's opinion.

           Mr. Justice STEVENS, concurring.

          While I agree with the Court's disposition of this case, I add these comments on the constitutional issue dis-

Page 253

cussed in Part II and the statutory issue discussed in Part III of the Court's opinion.

          The requirement of purposeful discrimination is a common thread running through the cases summarized in Part II. These cases include criminal convictions which were set aside because blacks were excluded from the grand jury, a reapportionment case in which political boundaries were obviously influenced to some extent by racial considerations, a school desegregation case, and a case involving the unequal administration of an ordinance purporting to prohibit the operation of laundries in frame buildings. Although it may be proper to use the same language to describe the constitutional claim in each of these contexts, the burden of proving a prima facie case may well involve differing evidentiary considerations. The extent of deference that one pays to the trial court's determination of the factual issue, and indeed, the extent to which one characterizes the intent issue as a question of fact or a question of law, will vary in different contexts.

          Frequently the most probative evidence of intent will be objective evidence of what actually happened rather than evidence describing the subjective state of mind of the actor. For normally the actor is presumed to have intended the natural consequences of his deeds. This is particularly true in the case of governmental action which is frequently the product of compromise, of collective decisionmaking, and of mixed motivation. It is unrealistic, on the one hand, to require the victim of alleged discrimination to uncover the actual subjective intent of the decisionmaker or, conversely, to invalidate otherwise legitimate action simply because an improper motive affected the deliberation of a participant in the decisional process. A law conscripting clerics should not be invalidated because an atheist voted for it.

Page 254

          My point in making this observation is to suggest that the line between discriminatory purpose and discriminatory impact is not nearly as bright, and perhaps not quite as critical, as the reader of the Court's opinion might assume. I agree, of course, that a constitutional issue does not arise every time some disproportionate impact is shown. On the other hand, when the disproportion is as dramatic as in Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 or Yick Wo v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064, 30 L.Ed. 220 (1886), it really does not matter whether the standard is phrased in terms of purpose or effect. Therefore, although I accept the statement of the general rule in the Court's opinion, I am not yet prepared to indicate how that standard should be applied in the many cases which have formulated the governing standard in different language.^*

          My agreement with the conclusion reached in Part II of the Court's opinion rests on a ground narrower than the Court describes. I do not rely at all on the evidence of good-faith efforts to recruit black police officers. In my judgment, neither those efforts nor the subjective good faith of the District administration, would save Test 21 if it were otherwise invalid.

          There are two reasons why I am convinced that the challenge to Test 21 is insufficient. First, the test serves the neutral and legitimate purpose of requiring all applicants to meet a uniform minimum standard of literacy. Reading ability is manifestly relevant to the police function, there is no evidence that the required passing grade was set at an arbitrarily high level, and there is sufficient disparity among high schools and high school graduates to justify the use of a separate uniform test. Second,

Page 255

the same test is used throughout the federal service. The applicants for employment in the District of Columbia Police Department represent such a small fraction of the total number of persons who have taken the test that their experience is of minimal probative value in assessing the neutrality of the test itself. That evidence, without more, is not sufficient to overcome the presumption that a test which is this widely used by the Federal Government is in fact neutral in its effect as well as its "purposes" that term is used in constitutional adjudication.

          My study of the statutory issue leads me to the same conclusion reached by the Court in Part III of its opinion. Since the Court of Appeals set aside the portion of the District Court's summary judgment granting the defendants' motion, I agree that we cannot ignore the statutory claims even though as the Court makes clear, Ante, at 238 n.10, there is not Title VII question in this case. The actual statutory holdings are limited to 42 U.S.C. § 1981 and § 1-320 of the District of Columbia Code, to which regulations of the Equal Employment Opportunity Commission have no direct application.

          The parties argued the case as though Title VII standards were applicable. In a general way those standards shed light on the issues, but there is sufficient individuality and complexity to that statute, and to the regulations promulgated under it, to make it inappropriate simply to transplant those standards in their entirety into a different statutory scheme having a different history. Moreover, the subject matter of this case the validity of qualifications for the law enforcement profession is one in which federal district judges have a greater expertise than in many others. I therefore do not regard this as a case in which the District Court was required to apply Title VII standards as strictly as would

Page 256

be necessary either in other contexts or in litigation actually arising under that statute.

          The Court's specific holding on the job-relatedness question contains, I believe, two components. First, as a matter of law, it is permissible for the police department to use a test for the purpose of predicting ability to master a training program even if the test does not otherwise predict ability to perform on the job. I regard this as a reasonable proposition and not inconsistent with the Court's prior holdings, although some of its prior language obviously did not contemplate this precise problem. Second, as a matter of fact, the District Court's finding that there was a correlation between success on the test and success in the training program has sufficient evidentiary support to withstand attack under the "clearly erroneous" standard mandated by Fed.Rule Civ.Proc. 52(a). Whether or not we would have made the same finding of fact, the opinion evidence identified in n. 17 of the Court's opinion and indeed the assumption made by the Court of Appeals quoted therein is surely adequate to support the finding under the proper standard of appellate review.

          On the understanding that nothing which I have said is inconsistent with the Court's reasoning, I join the opinion of the Court except to the extent that it expresses an opinion on the merits of the cases cited Ante, at 244-245, n. 12.

           Mr. Justice BRENNAN, with whom Mr. Justice MARSHALL joins, dissenting.

          The Court holds that the job qualification examination (Test 21) given by the District of Columbia Metropolitan Police Department does not unlawfully discriminate on the basis of race under either constitutional or statutory standards.

Page 257

          Initially, it seems to me that the Court should not pass on the statutory questions, because they are not presented by this case. The Court says that respondents' summary judgment motion "rested on purely constitutional grounds," Ante, at 236, and that "the Court of Appeals erroneously applied the legal standards applicable to Title VII cases in resolving the constitutional issue before it," Ante, at 238. There is a suggestion, however, that petitioners are entitled to prevail because they met the burden of proof imposed by 5 U.S.C. § 3304. Ante, at 249 n. 15. As I understand the opinion, the Court therefore holds that Test 21 is job-related under § 3304, but not necessarily under Title VII. But that provision, by the Court's own analysis, is no more in the case than Title VII; respondents' "complaint asserted no claim under § 3304." Ante, at 234 n. 2. Cf. Ante, at 238 n. 10. If it was "plain error" for the Court of Appeals to apply a statutory standard to this case, as the Court asserts, Ante, at 238-239, then it is unfortunate that the Court does not recognize that it is also plain error to address the statutory issues in Part III of its opinion.

          Nevertheless, although it appears unnecessary to reach the statutory questions, I will accept the Court's conclusion that respondents were entitled to summary judgment if they were correct in their statutory arguments, and I would affirm the Court of Appeals because petitioners have failed to prove that Test 21 satisfies the applicable statutory standards.¹ All parties' arguments and

Page 258

both lower court decisions were based on Title VII standards. In this context, I think it wrong to focus on § 3304 to the exclusion of the Title VII standards, particularly because the Civil Service Commission views the job-relatedness standards of Title VII and § 3304 as identical.² See also Infra, at 263.

          In applying a Title VII test,³ both the District Court and the Court of Appeals held that respondents had offered sufficient evidence of discriminatory impact to shift to petitioners the burden of proving job relatedness. 348 F.Supp. 15, 16; 168 U.S.App.D.C. 42, 45-47, 512 F.2d 956, 959-961. The Court does not question these rulings, and the only issue before us is what petitioners were required to show and whether they carried their burden. The Court agrees with the District Court's conclusion that Test 21 was validated by a positive relationship between Test 21 scores and performance in police training courses. This result is based upon the Court's reading of the record, its interpretation of in-

Page 259

structions governing testing practices issued by the Civil Service Commission (CSC), and "the current views of the Civil Service Commissioners who were parties to the case." We are also assured that today's result is not foreclosed by Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), and Albermarle Paper Co. v. Moody, 422 U.S. 405, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975). Finally, the Court asserts that its conclusion is "the much more sensible construction of the job relatedness requirement." Ante, at 251.

          But the CSC instructions cited by the Court do not support the District Court's conclusion. More importantly, the brief filed in this Court by the CSC takes the position that petitioners did not satisfy the burden of proof imposed by the CSC guidelines. It also appears that longstanding regulations of the Equal Employment Opportunity Commission (EEOC) previously endorsed by this Court require a result contrary to that reached by the Court. Furthermore, the Court's conclusion is inconsistent with my understanding of the interpretation of Title VII in Griggs and Albemarle. I do not find this conclusion "much more sensible" and with all respect I suggest that today's decision has the potential of significantly weakening statutory safeguards against discrimination in employment.

          On October 12, 1972, the CSC issued a supplement to the Federal Personnel Manual containing instructions for compliance with its general regulations concerning employment practices.⁴ The provision cited by the Court

Page 260

requires that Test 21 "have a demonstrable and rational relationship to important job-related performance objectives identified by management." "Success in training" is one example of a possible objective. The statistical correlation established by the Futransky validity study, Ante, at 251 n. 17, was between applicants' scores on Test 21 and recruits' average scores on final examinations given during the police training course.

          It is hornbook law that the Court accord deference to the construction of an administrative regulation when that construction is made by the administrative authority responsible for the regulation. E. g., Udall v. Tallman, 380 U.S. 1, 16, 85 S.Ct. 792, 801, 13 L.Ed.2d 616, 625 (1965). It is worthy of note, therefore, that the brief filed by the CSC in this case interprets the instructions in a manner directly contrary to the Court, despite the Court's claim that its result is supported by the Commissioners' "current views."

          "Under Civil Service Commission regulations and current professional standards governing criterion-related test validation procedures, the job-relatedness of an entrance examination may be demonstrated by proof that scores on the examination predict properly measured success in job-relevant training (regardless of whether they predict success on the job itself).

          "The documentary evidence submitted in the district court demonstrates that scores on Test 21 are predictive of Recruit School Final Averages. There

Page 261

          is little evidence, however, concerning the relationship between the Recruit School tests and the substance of the training program, and between the substance of the training program and the post-training job of a police officer. It cannot be determined, therefore, whether the Recruit School Final Averages are a proper measure of success in training and whether the training program is job-relevant." Brief for CSC 14-15 (emphasis added).

          The CSC maintains that a positive correlation between scores on entrance examinations and the criterion of success in training may establish the job relatedness of an entrance test thus relieving an employer from the burden of providing a relationship to job performance after training but only subject to certain limitations.

          "Proof that scores on an entrance examination predict scores on training school achievement tests, however, does not, by itself, satisfy the burden of demonstrating the job-relatedness of the entrance examination. There must also be evidence the nature of which will depend on the particular circumstances of the case showing that the achievement test scores are an appropriate measure of the trainee's mastery of the material taught in the training program and that the training program imparts to a new employee knowledge, skills, or abilities required for performance of the post-training job." Id., at 24-25.

          Applying its standards ⁵ the CSC concludes that none of

Page 262

the evidence presented in the District Court established "the appropriateness of using Recruit School Final Averages as the measure of training performance or the relationship of the Recruit School program to the job of a police officer." Id., at 30.⁶

          The CSC's standards thus recognize that Test 21 can be validated by a correlation between Test 21 scores and recruits' averages on training examinations only if (1) the training averages predict job performance or (2) the averages are proved to measure performance in job-related training. There is no proof that the recruits' average is correlated with job performance after completion of training. See n. 10, Infra. And although a positive relationship to the recruits' average might be sufficient to validate Test 21 if the average were proved to reflect mastery of material on the training curriculum that was in turn demonstrated to be relevant to job performance, the record is devoid of proof in this regard. First, there is no demonstration by petitioners that the training-course examinations measure comprehension of the training curriculum; indeed, these examinations do not even appear in the record. Furthermore, the Futransky study simply designated an average of 85 on the

Page 263

examination as a "good" performance and assumed that a recruit with such an average learned the material taught in the training course.⁷ Without any further proof of the significance of a score of 85, and there is none in the record, I cannot agree that Test 21 is predictive of "success in training."

          Today's decision is also at odds with EEOC regulations issued pursuant to explicit authorization in Title VII, 42 U.S.C. § 2000e-12(a). Although the dispute in this case is not within the EEOC's jurisdiction, as I noted above, the proper construction of Title VII nevertheless is relevant. Moreover, the 1972 extension of Title VII to public employees gave the same substantive protection to those employees as had previously been accorded in the private sector, Morton v. Mancari, 417 U.S. 535, 546-547, 94 S.Ct. 2474, 2480-2481, 41 L.Ed.2d 290, 298-299 (1974), and it is therefore improper to maintain different standards in the public and private sectors. Chandler v. Roudebush, 425 U.S. 840, 864, 96 S.Ct. 1949, 1961, 48 L.Ed.2d 416, 433 (1976). See n. 2, Supra.

          As with an agency's regulations, the construction of a statute by the agency charged with its administration is entitled to great deference. Trafficante v. Metropolitan Life Ins. Co., 409 U.S. 205, 210, 93 S.Ct. 364, 367, 34 L.Ed.2d 415, 419 (1972); Udall v. Tallman, 380 U.S., at 16, 85 S.Ct., at 801, 13 L.Ed.2d, at 625; Power Reactor Co. v. Electricians, 367 U.S. 396, 408, 81 S.Ct. 1529, 1535, 6 L.Ed.2d 924, 932 (1961). The defer-

Page 264

ence due the pertinent EEOC regulations is enhanced by the fact that they were neither altered nor disapproved when Congress extensively amended Title VII in 1972.⁸ Chemehuevi Tribe of Indians v. FPC, 420 U.S. 395, 410, 95 S.Ct. 1066, 1075, 43 L.Ed.2d 279, 290 (1975); Cammarano v. United States, 358 U.S. 498, 510, 79 S.Ct. 524, 531, 3 L.Ed.2d 462, 470 (1959); Allen v. Grand Central Aircraft Co., 347 U.S. 535, 547, 74 S.Ct. 745, 752, 98 L.Ed. 933, 943 (1954); Massachusetts Mut. Life Ins. Co. v. United States, 288 U.S. 269, 273, 53 S.Ct. 337, 339, 77 L.Ed. 739, 742 (1933). These principles were followed in Albemarle where the Court explicitly endorsed various regulations no fewer than eight times in its opinion, 422 U.S., at 431-436, 95 S.Ct., at 2378-2381, 45 L.Ed.2d, at 304-307 ⁹ and Griggs, 401 U.S., at 433-434, 91 S.Ct., at 854-855, 28 L.Ed.2d, at 165-166.

          The EEOC regulations require that the validity of a job qualification test be proved by "empirical data demonstrating that the test is predictive of or significantly correlated with important elements of work behavior which comprise or are relevant to the job or jobs for which candidates are being evaluated." 29 CFR § 1607.4(c) (1975). This construction of Title VII was approved in Albemarle, where we quoted this provision and remarked that "(t)he message of these Guidelines is the same as that of the Griggs case." 422 U.S., at 431, 95 S.Ct., at 2378, 45 L.Ed.2d, at 304. The regulations also set forth minimum standards for

Page 265

validation and delineate the criteria that may be used for this purpose.

          "The work behaviors or other criteria of employee adequacy which the test is intended to predict or identify must be fully described; and, additionally, in the case of rating techniques, the appraisal form(s) and instructions to the rater(s) must be included as a part of the validation evidence. Such criteria may include measures other than actual work proficiency, such as training time, supervisory ratings, regularity of attendance and tenure. Whatever criteria are used they must represent major or critical work behaviors as revealed by careful job analyses." 29 CFR § 1607.5(b)(3) (1975).

          This provision was also approved in Albemarle, 422 U.S., at 432, 95 S.Ct., at 2379, 45 L.Ed.2d, at 304, and n. 30.

          If we measure the validity of Test 21 by this standard, which I submit we are bound to do, petitioners' proof is deficient in a number of ways similar to those noted above. First, the criterion of final training examination averages does not appear to be "fully described." Although the record contains some general discussion of the training curriculum, the examinations are not in the record, and there is no other evidence completely elucidating the subject matter tested by the training examinations. Without this required description we cannot determine whether the correlation with training examination averages is sufficiently related to petitioners' need to ascertain "job-specific ability." See Albemarle, 422 U.S., at 433, 95 S.Ct., at 2379, 45 L.Ed.2d, at 305. Second, the EEOC regulations do not expressly permit validation by correlation to training performance, unlike the CSC instructions. Among the specified criteria the closest to training performance is "training time." All recruits to the Metropolitan Police Department, however, go through the

Page 266

same training course in the same amount of time, including those who experience some difficulty. See n. 7, supra. Third, the final requirement of § 1607.5(b)(3) has not been met. There has been no job analysis establishing the significance of scores on training examinations, nor is there any other type of evidence showing that these scores are of 'major or critical" importance.

          Accordingly, EEOC regulations that have previously been approved by the Court set forth a construction of Title VII that is distinctly opposed to today's statutory result.

          The Court also says that its conclusion is not foreclosed by Griggs and Albemarle, but today's result plainly conflicts with those cases. Griggs held that "(i)f an employment practice which operates to exclude Negroes cannot be shown to be Related to job performance, the practice is prohibited." 401 U.S., at 431, 91 S.Ct., at 853, 28 L.Ed.2d, at 164 (emphasis added). Once a discriminatory impact is shown, the employer carries the burden of proving that the challenged practice "bear(s) a Demonstrable relationship to successful performance of the jobs for which it was used." Ibid. (emphasis added). We observed further:

          "Nothing in the Act precludes the use of testing or measuring procedures; obviously they are useful. What Congress has forbidden is giving these devices and mechanisms controlling force unless they are demonstrably a reasonable measure of job performance. . . . What Congress has commanded is that any tests used must measure the person for the job and not the person in the abstract." Id., at 436, 91 S.Ct., at 856, 28 L.Ed.2d, at 167.

          Albemarle read Griggs to require that a discriminatory test be validated through proof "by professionally acceptable methods" that it is " 'predictive of or signifi-

Page 267

cantly correlated with important elements of work behavior which comprise or are relevant to the job or jobs for which candidates are being evaluated.' " 422 U.S., at 431, 95 S.Ct., at 2378, 45 L.Ed.2d, at 304 (emphasis added), quoting 29 CFR § 1607.4(c) (1975). Further, we rejected the employer's attempt to validate a written test by proving that it was related to supervisors' job performance ratings, because there was no demonstration that the ratings accurately reflected job performance. We were unable "to determine whether the criteria Actually considered were sufficiently related to the (employer's) legitimate interest in job-specific ability to justify a testing system with a racially discriminatory impact." 422 U.S., at 433, 95 S.Ct., at 2379, 45 L.Ed.2d, at 305 (emphasis in original). To me, therefore, these cases read Title VII as requiring proof of a significant relationship to job performance to establish the validity of a discriminatory test. See also McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668, 678 and n. 14 (1973). Petitioners do not maintain that there is a demonstrated correlation between Test 21 scores and job performance. Moreover, their validity study was unable to discern a significant positive relationship between training averages and job performance.¹⁰ Thus, there is no proof of a correlation either direct or indirect between Test 21 and performance of the job of being a police officer.

          It may well be that in some circumstances, proof of a relationship between a discriminatory qualification test and training performance is an acceptable substitute for establishing a relationship to job performance. But this question is not settled, and it should not be re-

Page 268

solved by the minimal analysis in the Court's opinion.¹¹ Moreover, it is particularly inappropriate to decide the question on this record. "Professionally acceptable methods" apparently recognize validation by proof of a correlation with training performance, rather than of performance, if (1) the training curriculum includes information proved to be important to job performance and (2) the standard used as a measure of training performance is shown to reflect the trainees' mastery of the material included in the training curriculum. See Brief for CSC 24-29; Brief for the Executive Committee of Division 14 of the American Psychological Assn. as Amicus Curiae 37-43. But no authority, whether professional, administrative, or judicial, has accepted the sufficiency of a correlation with training performance in the absence of such proof. For reasons that I have stated above, the record does not adequately establish either factor. As a result, the Court's conclusion cannot be squared with the focus on job performance in Griggs and Albemarle, even if this substitute showing is reconcilable with the holdings in those cases.

          Today's reduced emphasis on a relationship to job performance is also inconsistent with clearly expressed congressional intent. A section-by-section analysis of the 1972 amendments to Title VII states as follows:

          "In any area where the new law does not address itself, or in any areas where a specific contrary intention is not indicated, it was assumed that the present case law as developed by the courts would

Page 269

          continue to govern the applicability and construction of Title VII." 118 Cong.Rec. 7166 (1972).

          The pre-1972 judicial decisions dealing with standardized tests used as job qualification requirements uniformly follow the EEOC regulations discussed above and insist upon proof of a relationship to job performance to prove that a test is job related.¹² Furthermore, the Court ignores Congress' explicit hostility toward the use of written tests as job-qualification requirements; Congress disapproved the CSC's "use of general ability tests which are not aimed at any direct relationship to specific jobs." H.R.Rep. No. 92-238, p. 24 (1971). See S.Rep. No. 92-415, pp. 14-15 (1971). Petitioners concede that Test 21 was devised by the CSC for general use and was not designed to be used by police departments.

          Finally, it should be observed that every federal court, except the District Court in this case, presented with proof identical to that offered to validate Test 21 has reached a conclusion directly opposite to that of the

Page 270

Court today.¹³ Sound policy considerations support the view that, at a minimum, petitioners should have been required to prove that the police training examinations either measure job-related skills or predict job performance. Where employers try to validate written qualification tests by proving a correlation with written examinations in a training course, there is a substantial danger that people who have good verbal skills will achieve high scores on both tests due to verbal ability, rather than "job-specific ability." As a result, employers could validate any entrance examination that measures only verbal ability by giving another written test that measures verbal ability at the end of a training course. Any contention that the resulting correlation between examination scores would be evidence that the initial test is "job related" is plainly erroneous. It seems to me, however, that the Court's holding in this case can be read as endorsing this dubious proposition. Today's result will prove particularly unfortunate if it is extended to govern Title VII cases.

          Accordingly, accepting the Court's assertion that it is necessary to reach the statutory issue, I would hold that petitioners have not met their burden of proof and affirm the judgment of the Court of Appeals.

^{1. Under § 4-103 of the District of Columbia Code, appointments to the Metropolitan Police force were to be made by the Commissioner subject to the provisions of Title 5 of the United States Code relating to the classified civil service. The District of Columbia Council and the Office of Commissioner of the District of Columbia, established by Reorganization Plan No. 37 of 1967, were abolished as of January 2, 1975, and replaced by the Council of the District of Columbia and the Office of Mayor of the District of Columbia.}

^2. Title 42 U.S.C. § 1981 provides:

"All persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts, to sue, be parties, give evidence, and to the full and equal benefit of all laws and proceedings for the security of persons and property as is enjoyed by white citizens, and shall be subject to like punishment, pains, penalties, taxes, licenses, and exactions of every kind, and to no other."

Section 1-320 of the District of Columbia Code (1973) provides:

"In any program of recruitment or hiring of individuals to fill positions in the government of the District of Columbia, no officer or employee of the government of the District of Columbia shall exclude or give preference to the residents of the District of Columbia or any State of the United States on the basis of residence, religion, race, color, or national origin."

One of the provisions expressly made applicable to the Metropolitan Police force by § 4-103 is 5 U.S.C. § 3304(a), which provides:

"s 3304. Competitive service; examinations.

"(a) The President may prescribe rules which shall provide, as nearly as conditions of good administration warrant, for

"(1) open, competitive examinations for testing applicants for appointment in the competitive service which are practical in character and as far as possible relate to matters that fairly test the relative capacity and fitness of the applicants for the appointment sought; and

"(2) noncompetitive examinations when competent applicants do not compete after notice has been given of the existence of the vacancy."

The complaint asserted no claim under § 3304.

^3. Those proceedings included a hearing on respondents' motion for an order designating the case as a class action. A ruling on the motion was held in abeyance and was never granted insofar as the record before us reveals.

^4. In support of the motion, petitioners and the federal parties urged that they were in compliance with all applicable constitutional, statutory, and regulatory provisions, including the provisions of the Civil Service Act which since 1883 were said to have established a "job relatedness" standard for employment.

^5. When summary judgment was granted, the case with respect to discriminatory promotions was still pending. The District Court, however, made the determination and direction authorized by Fed.Rule Civ.Proc. 54(b). The promotion issue was subsequently decided adversely to the original plaintiffs. Davis v. Washington, 352 F.Supp. 187 (DC 1972).

^6. "Although appellants' complaint did not allege a violation of Title VII of the Civil Rights Act of 1964, which then was inapplicable to the Federal Government, decisions applying Title VII furnish additional instruction as to the legal standard governing the issues raised in this case. . . . The many decisions disposing of employment discrimination claims on constitutional grounds have made no distinction between the constitutional standard and the statutory standard under Title VII." 168 U.S.App.D.C. 42, 44 n. 2, 512 F.2d 956, 958 n. 2 (1975).

^7. The Civil Service Commissioners, defendants in the District Court, did not petition for writ of certiorari but have filed a brief as respondents. See our Rule 21(4). We shall at times refer to them as the "federal parties."

^8. Apparently not disputing the applicability of the Griggs and Title VII standards in resolving this case, petitioners presented issues going only to whether Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), had been misapplied by the Court of Appeals.

^9. See, E. g., Silber v. United States, 370 U.S. 717, 82 S.Ct. 1287, 8 L.Ed.2d 798 (1962); Carpenters v. United States, 330 U.S. 395, 412, 67 S.Ct. 775, 784, 91 L.Ed. 973, 987 (1947); Sibbach v. Wilson & Co., 312 U.S. 1, 16, 61 S.Ct. 422, 427, 85 L.Ed. 479, 486 (1941); Mahler v. Eby, 264 U.S. 32, 45, 44 S.Ct. 283, 288, 68 L.Ed. 549, 557 (1924); Weems v. United States, 217 U.S. 349, 362, 30 S.Ct. 544, 547, 54 L.Ed. 793, 796 (1910).

^10. Although Title VII standards have dominated this case, the statute was not applicable to federal employees when the complaint was filed; and although the 1972 amendments extending the Title to reach Government employees were adopted prior to the District Court's judgment, the complaint was not amended to state a claim under that Title, nor did the case thereafter proceed as a Title VII case. Respondents' motion for partial summary judgment, filed after the 1972 amendments, rested solely on constitutional grounds; and the Court of Appeals ruled that the motion should have been granted.

At the oral argument before this Court, when respondents' counsel was asked whether "this is just a purely Title VII case as it comes to us from the Court of Appeals without any constitutional overtones," counsel responded: "My trouble honestly with that proposition is the procedural requirements to get into court under Title VII, and this case has not met them." Tr. of Oral Arg. 66.

^11. To the extent that Palmer suggests a generally applicable proposition that legislative purpose is irrelevant in constitutional adjudication, our prior cases as indicated in the text are to the contrary; and very shortly after Palmer, all Members of the Court majority in that case joined the Court's opinion in Lemon v. Kurtzman, 403 U.S. 602, 91 S.Ct. 2105, 29 L.Ed.2d 745 (1971), which dealt with the issue of public financing for private schools and which announced, as the Court had several times before, that the validity of public aid to church-related schools includes close inquiry into the purpose of the challenged statute.

^12. Cases dealing with public employment include: Chance v. Board of Examiners, 458 F.2d 1167, 1176-1177 (CA2 1972); Castro v. Beecher, 459 F.2d 725, 732-733 (CA1 1972); Bridgeport Guardians v. Bridgeport Civil Service Comm'n, 482 F.2d 1333, 1337 (CA2 1973); Harper v. Mayor of Baltimore, 359 F.Supp. 1187, 1200 (D.Md.), aff'd in pertinent part Sub nom. Harper v. Kloster, 486 F.2d 1134 (CA4 1973); Douglas v. Hampton, 168 U.S.App.D.C. 62, 67, 512 F.2d 976, 981 (1975); but cf. Tyler v. Vickery, 517 F.2d 1089, 1096-1097 (CA5 1975), cert. pending, No. 75-1026. There are also District Court cases: Wade v. Mississippi Cooperative Extension Serv., 372 F.Supp. 126, 143 (ND Miss. 1974); Arnold v. Ballard, 390 F.Supp. 723, 736, 737 (N.D. Ohio 1975); United States v. City of Chicago, 385 F.Supp. 543, 553 (N.D. Ill. 1974); Fowler v. Schwarzwalder, 351 F.Supp. 721, 724 (D.Minn. 1972), rev'd on other grounds, 498 F.2d 143 (CA8 1974).

In other contexts there are Norwalk CORE v. Norwalk Redevelopment Agency, 395 F.2d 920 (CA2 1968) (urban renewal); Kennedy Park Homes Assn. v. City of Lackawanna, 436 F.2d 108, 114 (CA2 1970), cert. denied, 401 U.S. 1010, 91 S.Ct. 1256, 28 L.Ed.2d 546 (1971) (zoning); Southern Alameda Spanish Speaking Organization v. Union City, 424 F.2d 291 (CA9 1970) (dictum) (zoning); Metropolitan H. D. Corp. v. Village of Arlington Heights, 517 F.2d 409 (CA7), cert. granted, December 15, 1975, 423 U.S. 1030, 96 S.Ct. 560, 46 L.Ed.2d 404 (1975) (zoning); Gautreaux v. Romney, 448 F.2d 731, 738 (CA7 1971) (dictum) (public housing); Crow v. Brown, 332 F.Supp. 382, 391 (N.D. Ga. 1971), aff'd, 457 F.2d 788 (CA5 1972) (public housing); Hawkins v. Town of Shaw, 437 F.2d 1286 (CA5 1971), aff'd on rehearing en banc, 461 F.2d 1171 (1972) (municipal services).

^13. It appears beyond doubt by now that there is no single method for appropriately validating employment tests for their relationship to job performance. Professional standards developed by the American Psychological Association in its Standards for Educational and Psychological Tests and Manuals (1966), accept three basic methods of validation: "empirical" or "criterion" validity (demonstrated by identifying criteria that indicate successful job performance and then correlating test scores and the criteria so identified); "construct" validity (demonstrated by examinations structured to measure the degree to which job applicants have identifiable characteristics that have been determined to be important in successful job performance); and "content" validity (demonstrated by tests whose content closely approximates tasks to be performed on the job by the applicant). These standards have been relied upon by the Equal Employment Opportunity Commission in fashioning its Guidelines on Employee Selection Procedures, 29 CFR pt. 1607 (1975), and have been judicially noted in cases where validation of employment tests has been in issue. See, E.g., Albemarle Paper Co. v. Moody, 422 U.S. 405, 431, 95 S.Ct. 2362, 2378, 45 L.Ed.2d 280, 304 (1975); Douglas v. Hampton, 168 U.S.App.D.C., at 70, 512 F.2d, at 984; Vulcan Society v. Civil Service Comm'n, 490 F.2d 387, 394 (CA2 1973).

^14. Goodman, De Facto School Segregation: A Constitutional and Empirical Analysis, 60 Calif.L.Rev. 275, 300 (1972), suggests that disproportionate-impact analysis might invalidate "tests and qualifications for voting, draft deferment, public employment, jury service, and other government-conferred benefits and opportunities . . .; (s)ales taxes, bail schedules, utility rates, bridge tolls, license fees, and other state-imposed charges." It has also been argued that minimum wage and usury laws as well as professional licensing requirements would require major modifications in light of the unequal-impact rule. Silverman, Equal Protection, Economic Legislation, and Racial Discrimination, 25 Vand.L.Rev. 1183 (1972). See also Demsetz, Minorities in the Market Place, 43 N.C.L.Rev. 271 (1965).

^15. In their memorandum supporting their motion for summary judgment, the federal parties argued:

"In Griggs, supra, the Supreme Court set a job-relationship standard for the private sector employers which has been a standard for federal employment since the passage of the Civil Service Act in 1883. In that act Congress has mandated that the federal government must use '. . . examinations for testing applicants for appointment . . . which . . . as far as possible relate to matters that fairly test the relative capacity and fitness of the applicants for the appointments sought.' 5 U.S.C. § 3304(a)(1). Defendants contend that they have been following the job-related standards of Griggs, supra, for the past eighty-eight years by virtue of the enactment of the Civil Service Act which guaranteed open and fair competition for jobs."

They went on to argue that the Griggs standard had been satisfied. In granting the motions for summary judgment filed by petitioners and the federal parties, the District Court necessarily decided adversely to respondents the statutory issues expressly or tacitly tendered by the parties.

^16. See n. 17, Infra. Current instructions of the Civil Service Commission on "Examining, Testing, Standards, and Employment Practices" provide in pertinent part:

"S2-2 Use of applicant appraisal procedures

a. Policy. The Commission's staff develops and uses applicant appraisal procedures to assess the knowledges, skills, and abilities of persons for jobs and not persons in the abstract.

"(1) Appraisal procedures are designed to reflect real, reasonable, and necessary qualifications for effective job behavior.

"(2) An appraisal procedure must, among other requirements, have a demonstrable and rational relationship to important job-related performance objectives identified by management, such as:

"(a) Effective job performance;

"(b) Capability;

"(c) Success in training;

"(d) Reduced turnover; or

"(e) Job satisfaction." 37 Fed.Reg. 21557 (1972).

See also Equal Employment Opportunity Commission Guidelines on Employee Selection Procedures, 29 CFR § 1607.5(b)(3) (1975), discussed in Albemarle Paper Co. v. Moody, 422 U.S., at 430-435, 95 S.Ct. 2362, 2378-2380, 45 L.Ed.2d 280, 304-307.

^17. The record includes a validation study of Test 21's relationship to performance in the recruit training program. The study was made by D. L. Futransky of the Standards Division, Bureau of Policies and Standards, United States Civil Service Commission. App., at 99-109. Findings of the study included data "support(ing) the conclusion that T(est) 21 is effective in selecting trainees who can learn the material that is taught at the Recruit School." Id., at 103. Opinion evidence, submitted by qualified experts examining the Futransky study and/or conducting their own research, affirmed the correlation between scores on Test 21 and success in the training program. E. g., Affidavit of Dr. Donald J. Schwartz (personnel research psychologist, United States Civil Service Commission), App. 178, 183 ("It is my opinion . . . that Test 21 has a significant positive correlation with success in the MPD Recruit School for both Blacks and whites and is therefore shown to be job related . . ."); affidavit of Diane E. Wilson (personnel research psychologist, United States Civil Service Commission), App. 185, 186 ("It is my opinion that there is a direct and rational relationship between the content and difficulty of Test 21 and successful completion of recruit school training").

The Court of Appeals was "willing to assume for purposes of this appeal that appellees have shown that Test 21 is predictive of further progress in Recruit School." 168 U.S.App.D.C., at 48, 512 F.2d, at 962.

^* Specifically, I express no opinion on the merits of the cases listed in n. 12 of the Court's opinion.

^1. Although I do not intend to address the constitutional questions considered by the Court in Part II of its opinion, I feel constrained to comment upon the propriety of footnote 12, Ante, at 244-245. One of the cases "disapproved" therein is presently scheduled for plenary consideration by the Court in the 1976 Term, Metropolitan Housing Development Corp. v. Village of Arlington Heights, 517 F.2d 409 (CA7), cert. granted, 423 U.S. 1030, 96 S.Ct. 560, 46 L.Ed.2d 404 (1975). If the Court regarded this case only a few months ago as worthy of full briefing and argument, it ought not be effectively reversed merely by its inclusion in a laundry list of lower court decisions.

^2. The only administrative authority relied on by the Court in support of its result is a regulation of the Civil Service Commission construing the civil service employment standards in Title 5 of the United States Code. Ante, at 250-251 n. 16. I note, however, that 5 U.S.C. § 3304 was brought into this case by the CSC, not by respondents, and the CSC's only reason for referring to that provision was to establish that petitioners had been "following the job-related standards of Griggs (V. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971),) for the past eighty-eight years." Ante, at 249 n. 15.

^3. The provision in Title VII on which petitioners place principal reliance is 42 U.S.C. § 2000e-2(h). See Griggs v. Duke Power Co., supra, 401 U.S., at 433-436, 91 S.Ct., at 854-856, 28 L.Ed.2d, at 165-167.

^4. See 5 CFR § 300.101 Et seq. (1976). These instructions contain the "regulations" that the Court finds supportive of the District Court's conclusion, which was reached under Title VII, but neither the instructions nor the general regulations are an interpretation of Title VII. The instructions were issued "under authority of sections 3301 and 3302 of title 5, United States Code, and E.O. 10577, 3 CFR 1954-58 Comp., p. 218." 37 Fed.Reg. 21552 (1972). The pertinent regulations of the CSC in 5 CFR § 300.101 Et seq. were promulgated pursuant to the same authorities, as well as 5 U.S.C. §§ 7151, 7154 and Exec.Order No. 11478, 3 CFR (1966-1970 Comp.) 803.

^5. The CSC asserts that certain of its guidelines have some bearing on Test 21's job relatedness. Under the CSC instructions, " 'criterion-related' validity," see Douglas v. Hampton, 168 U.S.App.D.C. 62, 70 n. 60, 512 F.2d 976, 984 n. 60 (1975), can be established by demonstrating a correlation between entrance examination scores and "a criterion which is legitimately based on the needs of the Federal Government." P S3-2(a)(2), 37 Fed.Reg. 21558 (1972). Further, to prove validity, statistical studies must demonstrate that Test 21, "to a significant degree, measures performance or qualifications requirements which are relevant to the job or jobs for which candidates are being evaluated." P S3-3(a), 37 Fed.Reg. 21558 (1972). These provisions are ignored in the Court's opinion.

^6. On this basis, the CSC argues that the case ought to be remanded to enable petitioners to try to make such a demonstration, but this resolution seems to me inappropriate. Both lower courts recognized that petitioners had the burden of proof, and as this burden is yet unsatisfied, respondents are entitled to prevail.

^7. The finding in the Futransky study on which the Court relies, Ante, at 251 n. 17, was that Test 21 "is effective in selecting trainees who can learn the material that is taught at the Recruit School," because it predicts averages over 85. On its face, this would appear to be an important finding, but the fact is that Everyone learns the material included in the training course. The study noted that all recruits pass the training examinations; if a particular recruit has any difficulty, he is given assistance until he passes.

^8. Still another factor mandates deference to the EEOC regulations. The House and Senate committees considering the 1972 amendments to Title VII recognized that discrimination in employment, including the use of testing devices, is a "complex and pervasive phenomenon." S.Rep. No. 92-415, p. 5 (1971); H.R.Rep. No. 92-238, p. 8 (1971); U.S.Code Cong. & Admin.News 1972, p. 2137. As a result, both committees noted the need to obtain "expert assistance" in this area. S.Rep. No. 92-415, Supra, at 5; H.R.Rep. No. 92-238, Supra, at 8.

^9. Indeed, two Justices asserted that the Court relied too heavily on the EEOC guidelines. 422 U.S. 449, 95 S.Ct. 2389, 45 L.Ed.2d 316 (Blackmun, J., concurring in judgment); Id., at 451, 95 S.Ct., at 2387, 45 L.Ed.2d, at 317 (Burger, C. J., concurring in part and dissenting in part).

^10. Although the validity study found that Test 21 predicted job performance for white officers, but see Albemarle, 422 U.S., at 433, 95 S.Ct., at 2379, 45 L.Ed.2d, at 305, no similar relationship existed for black officers. The same finding was made as to the relationship between training examination averages and job performance. See Id., at 435, 95 S.Ct., at 2380, 45 L.Ed.2d, at 306.

^11. The Court of Appeals recognized that deciding whether 42 U.S.C. § 2000e-2(h) permitted such proof "is not a simple or insignificant endeavor." 168 U.S.App.D.C. 42, 50 n. 59, 512 F.2d 956, 964 n. 59. The court declined to express any view on this issue on the ground that petitioners had not satisfied this standard even if it were acceptable, which seems to me the proper treatment of the question.

^12. Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971); United States v. Jacksonville Terminal Co., 451 F.2d 418, 456-457 (CA5 1971), cert. denied, 406 U.S. 906, 92 S.Ct. 1607, 31 L.Ed.2d 815 (1972); Hicks v. Crown Zellerbach Corp., 319 F.Supp. 314, 319-321 (E.D.La.1970) (issuing preliminary injunction), 321 F.Supp. 1241, 1244 (1971) (issuing permanent injunction). See also Castro v. Beecher, 334 F.Supp. 930 (D.Mass.1971), aff'd in part and rev'd in part on other grounds, 459 F.2d 725 (CA1 1972); Western Addition Community Org. v. Alioto, 330 F.Supp. 536, 539-540 (N.D.Cal.1971), 340 F.Supp. 1351, 1354-1356 (1972) (issuing preliminary injunction), 360 F.Supp. 733 (1973) (issuing permanent injunction); Chance v. Board of Examiners, 330 F.Supp. 203 (S.D.N.Y.1971), aff'd, 458 F.2d 1167 (CA2 1972); Baker v. Columbus Mun. Sep. School Dist., 329 F.Supp. 706, 721-722 (N.D.Miss.1971), aff'd, 462 F.2d 1112 (CA5 1972); Arrington v. Massachusetts Bay Transp. Auth., 306 F.Supp. 1355 (D.Mass.1969).

^13. United States v. City of Chicago, 385 F.Supp. 543, 555-556 (N.D.Ill.1974) (police department); Officers for Justice v. CSC, 371 F.Supp. 1328, 1337 (N.D.Cal.1973) (police department); Smith v. City of East Cleveland, 363 F.Supp. 1131, 1148-1149 (N.D.Ohio 1973) (police department), aff'd in part and rev'd in part on other grounds, 520 F.2d 492 (CA6 1975); Harper v. Mayor of Baltimore, 359 F.Supp. 1187, 1202-1203 (D.Md.) (fire department), modified and aff'd, 486 F.2d 1134 (CA4 1973); Pennsylvania v. O'Neill, 348 F.Supp. 1084, 1090-1091 (E.D.Pa.1972) (police department), aff'd in pertinent part and vacated in part, 473 F.2d 1029 (CA3 1973).

          Syllabus

          Respondents brought this class action, alleging that the Detroit public school system is racially segregated as a result of the official policies and actions of petitioner state and city officials, and seeking implementation of a plan to eliminate the segregation and establish a unitary nonracial school system. The District Court, after concluding that various acts by the petitioner Detroit Board of Education had created and perpetuated school segregation in Detroit, and that the acts of the Board, as a subordinate entity of the State, were attributable to the State, ordered the Board to submit Detroit-only desegregation plans. The court also ordered the state officials to submit desegregation plans encompassing the three-county metropolitan area, despite the fact that the 85 outlying school districts in these three counties were not parties to the action and there was no claim that they had committed constitutional violations. Subsequently, outlying school districts were allowed to intervene, but were not permitted to assert any claim or defense on issues previously adjudicated or to reopen any issue previously decided, but were allowed merely to advise the court as to the propriety of a metropolitan plan and to submit any objections, modifications, or alternatives to any such plan. Thereafter, the District Court ruled that it was proper to consider metropolitan plans that Detroit-only plans submitted by the Board and respondents were inadequate to accomplish desegregation, and that therefore it would seek a solution beyond the limits of the Detroit School District, and concluded that '(s)chool district lines are simply matters of political convenience and may not be used to deny constitutional rights.' Without having evidence that the suburban school districts had committed acts of de jure segregation, the court appointed a panel to submit a plan for the

Page 718

Detroit schools that would encompass an entire designated desegregation area consisting of 53 of the 85 suburban school districts plus Detroit, and ordered the Detroit Board to acquire at least 295 school buses to provide transportation under an interim plan to be developed for the 1972—1973 school year. The Court of Appeals, affirming in part, held that the record supported the District Court's finding as to the constitutional violations committed by the Detroit Board and the state officials; that therefore the District Court was authorized and required to take effective measures to desegregate the Detroit school system; and that a metropolitan area plan embracing the 53 outlying districts was the only feasible solution and was within the District Court's equity powers. But the court remanded so that all suburban school districts that might be affected by a metropolitan remedy could be made parties and have an opportunity to be heard as to the scope and implementation of such a remedy, and vacated the order as to the bus acquisitions, subject to its reimposition at an appropriate time. Held: The relief ordered by the District Court and affirmed by the Court of Appeals was based upon erroneous standards and was unsupported by record evidence that acts of the outlying districts had any impact on the discrimination found to exist in the Detroit schools. A federal court may not impose a multidistrict, areawide remedy for single-district de jure school segregation violations, where there is no finding that the other included school districts have failed to operate unitary school systems or have committed acts that effected segregation within the other districts, there is no claim or finding that the school district boundary lines were established with the purpose of fostering racial segregation, and there is no meaningful opportunity for the included neighboring school districts to present evidence or be heard on the propriety of a multidistrict remedy or on the question of constitutional violations by those districts. Pp. 737—753.

          (a) The District Court erred in using as a standard the declared objective of development of a metropolitan area plan which, upon implementation, would leave 'no school, grade or classroom . . . substantially disproportionate to the overall pupil racial composition' of the metropolitan area as a whole. The clear import of Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554, is that desegregation, in the sense of dismantling a dual school system, does not require any particular racial balance. Pp. 739—741.

Page 719

          (b) While boundary lines may be bridged in circumstances where there has been a constitutional violation calling for inter-district relief, school district lines may not be casually ignored or treated as a mere administrative convenience; substantial local control of public education in this country is a deeply rooted tradition. Pp. 741—742.

          (c) The interdistrict remedy could extensively disrupt and alter the structure of public education in Michigan, since that remedy would require, in effect, consolidation of 54 independent school districts historically administered as separate governmental units into a vast new super school district, and, since—entirely apart from the logistical problems attending large-scale transportation of students—the consolidation would generate other problems in the administration, financing, and operation of this new school system. Pp. 742—743.

          (d) From the scope of the interdistrict plan itself, absent a complete restructuring of the Michigan school district laws, the District Court would become, first, a de facto 'legislative authority' to resolve the complex operational problems involved and thereafter a 'school superintendent' for the entire area, a task which few, if any, judges are qualified to perform and one which would deprive the people of local control of schools through elected school boards. Pp. 743—744.

          (e) Before the boundaries of separate and autonomous school districts may be set aside by consolidating the separate units for remedial purposes or by imposing a cross-district remedy, it must be first shown that there has been a constitutional violation within one district that produces a significant segregative effect in another district; i.e., specifically, it must be shown that racially discriminatory acts of the state or local school districts, or of a single school district have been a substantial cause of interdistrict segregation. Pp. 744—745.

          (f) With no showing of significant violation by the 53 outlying school districts and no evidence of any interdistrict violation or effect, the District Court transcended the original theory of the case as framed by the pleadings, and mandated a metropolitan area remedy, the approval of which would impose on the outlying districts, not shown to have committed any constitutional violation, a standard not previously hinted at in any holding of this Court. P. 745.

          (g) Assuming, arguendo, that the State was derivatively responsible for Detroit's segregated school conditions, it does not follow

Page 720

that an interdistrict remedy is constitutionally justified or required, since there has been virtually no showing that either the State or any of the 85 outlying districts engaged in any activity that had a cross-district effect. Pp. 748—749.

          (h) An isolated instance of a possible segregative effect as between two of the school districts involved would not justify the broad metropolitanwide remedy contemplated, particularly since that remedy embraced 52 districts having no responsibility for the arrangement and potentially involved 503,000 pupils in addition to Detroit's 276,000 pupils. Pp. 749—750.

          484 F.2d 215, reversed and remanded.

          Frank J. Kelley, Lansing, Mich., for petitioners William G. Milliken et al.

          William M. Saxton, Detroit, Mich., for petitioners Allen Park Public Schools and Grosse Pointe Public School System et al.

Page 721

          Solicitor Gen. Robert H. Bork for the United States, as amicus curiae, by special leave of Court.

          J. Harold Flannery, Cambridge, Mass., and Nathaniel R. Jones, New York City, for respondents.

           Mr. Chief Justice BURGER delivered the opinion of the Court.

          We granted certiorari in these consolidated cases to determine whether a federal court may impose a multidistrict, areawide remedy to a single-district de jure segregation problem absent any finding that the other included school districts have failed to operate unitary school systems within their districts, absent any claim or finding that the boundary lines of any affected school district were established with the purpose of fostering racial segregation in public schools, absent any finding that the included districts committed acts which effected segregation within the other districts, and absent a

Page 722

meaningful opportunity for the included neighboring school districts to present evidence or be heard on the propriety of a multidistrict remedy or on the question of constitutional violations by those neighboring districts.¹

          The action was commenced in August 1970 by the respondents, the Detroit Branch of the National Association for the Advancement of Colored People² and individual parents and students, on behalf of a class later defined by order of the United States District Court for the Eastern District of Michigan, dated February 16, 1971, to included 'all school children in the City of Detroit, Michigan, and all Detroit resident parents who have children of school age.' The named defendants in the District Court included the Governor of Michigan, the Attorney General, the State Board of Education, the State Superintendent of Public Instruction, the Board of Education of the city of Detroit, its members, the city's and its former superintendent of schools. The State of Michigan as such is not a party to this litigation and references to the State must be read as references to the public officials, state and local, through whom the State is alleged to have acted. In their complaint respondents attacked the constitutionality of a statute of the State of Michigan known as Act 48 of the 1970 Legislature on the ground that it put the State of Michigan in the position of unconstitutionally interfering with the execution and operation of a voluntary plan of partial high school desegregation, known as the April 7, 1970, Plan, which had been adopted by the Detroit Board of Education to be effective beginning

Page 723

with the fall 1970 semester. The complaint also alleged that the Detroit Public School System was and is segregated on the basis of race as a result of the official policies and actions of the defendants and their predecessors in office, and called for the implementation of a plan that would eliminate 'the racial identity of every school in the (Detroit) system and . . . maintain now and hereafter a unitary, nonracial school system.'

          Initially the matter was tried on respondents' motion for a preliminary injunction to restrain in enforcement of Act 48 so as to permit the April 7 Plan to be implemented. On that issue, the District Court ruled that respondents were not entitled to a preliminary injunction since at that stage there was no proof that Detroit had a dual segregated school system. On appeal, the Court of Appeals found that the 'implementation of the April 7 plan was (unconstitutionally) thwarted by State action in the form of the Act of the Legislature of Michigan,' 433 F.2d 897, 902 (CA6 1970), and that such action could not be interposed to delay, obstruct, or nullify steps lawfully taken for the purpose of protecting rights guaranteed by the Fourteenth Amendment. The case was remanded to the District Court for an expedited trial on the merits.

          On remand, the respondents moved for immediate implementation of the April 7 Plan in order to remedy the deprivation of the claimed constitutional rights. In response, the School Board suggested two other plans, along with the April 7 Plan, and urged that top priority be assigned to the so-called 'Magnet Plan' which was 'designed to attract children to a school because of its superior curriculum.' The District Court approved the Board's Magnet Plan, and respondents again appealed to the Court of Appeals, moving for summary reversal. The Court of Appeals refused to pass on the merits of the Magnet Plan and ruled that the District Court had

Page 724

not abused its discretion in refusing to adopt the April 7 Plan without an evidentiary hearing. The case was again remanded with instructions to proceed immediately to a trial on the merits of respondents' substantive allegations concerning the Detroit school system. 438 F.2d 945 (CA6 1971).

          The trial of the issue of segregation in the Detroit school system began on April 6, 1971, and continued through July 22, 1971, consuming some 41 trial days. On September 27, 1971, the District Court issued its findings and conclusions on the issue of segregation, finding that 'Governmental actions and inaction at all levels, federal, state and local, have combined, with those of private organizations, such as loaning institutions and real estate associations and brokerage firms, to establish and to maintain the pattern of residential segregation throughout the Detroit metropolitan area.' 338 F.Supp. 582, 587 (ED Mich.1971). While still addressing a Detroit-only violation, the District Court reasoned:

          'While it would be unfair to charge the present defendants with what other governmental officers or agencies have done, it can be said that the actions or the failure to act by the responsible school authorities, both city and state, were linked to that of these other governmental units. When we speak of governmental action we should not view the different agencies as a collection of unrelated units. Perhaps the most that can be said is that all of them, including the school authorities, are, in part, responsible for the segregated condition which exists. And we note that just as there is an interaction between residential patterns and the racial composition of the schools, so there is a corresponding effect on the residential pattern by the racial composition of the schools.' Ibid.

Page 725

          The District Court found that the Detroit Board of Education created and maintained optional attendance zones³ within Detroit neighborhoods undergoing racial transition and between high school attendance areas of opposite predominant racial compositions. These zones, the court found, had the 'natural, probable, foreseeable and actual effect' of allowing white pupils to escape identifiably Negro schools. Ibid. Similarly, the District Court found that Detroit school attendance zones had been drawn along north-south boundary lines despite the Detroit Board's awareness that drawing boundary lines in an east-west direction would result in significantly greater desegregation. Again, the District Court concluded, the natural and actual effect of these acts was the creation and perpetuation of school segregation within Detroit.

          The District Court found that in the operation of its school transportation program, which was designed to relieve overcrowding, the Detroit Board had admittedly bused Negro Detroit pupils to predominantly Negro schools which were beyond or away from closer white schools with available space.⁴ This practice was found to have continued in recent years despite the Detroit Board's avowed policy, adopted in 1967, of utilizing transportation to increase desegregation:

          'With one exception (necessitated by the burning of a white school), defendant Board has never bused

Page 726

          white children to predominantly black schools. The Board has not bused white pupils to black schools despite the enormous amount of space available in inner-city schools. There were 22,961 vacant seats in schools 90% or more black.' Id., at 588.

          With respect to the Detroit Board of Education's practices in school construction, the District Court found that Detroit school construction generally tended to have a segregative effect with the great majority of schools being built in either overwhelmingly all-Negro or all-white neighborhoods so that the new schools opened as predominantly one-race schools. Thus, of the 14 schools which opened for use in 1970—1971, 11 opened over 90% Negro and one opened less than 10% Negro.

          The District Court also found that the State of Michigan had committed several constitutional violations with respect to the exercise of its general responsibility for, and supervision of, public education.⁵ The State, for example, was found to have failed, until the 1971 Session of the Michigan Legislature, to provide authorization or

Page 727

funds for the transportation of pupils within Detroit regardless of their poverty or distance from the school to which they were assigned; during this same period the State provided many neighboring, mostly white, suburban districts the full range of state-supported transportation.

          The District Court found that the State, through Act 48, acted to 'impede, delay and minimize racial integration in Detroit schools.' The first sentence of § 12 of Act 48 was designed to delay the April 7, 1970, desegregation plan originally adopted by the Detroit Board. The remainder of § 12 sought to prescribe for each school in the eight districts criteria of 'free choice' and 'neighborhood schools,' which, the District Court found, 'had as their purpose and effect the maintenance of segregation.' 338 F.Supp., at 589.⁶

          The District Court also held that the acts of the Detroit Board of Education, as a subordinate entity of the State, were attributable to the State of Michigan, thus creating a vicarious liability on the part of the State. Under Michigan law, Mich.Comp.Laws § 388.851 (1970), for example, school building construction plans had to be approved by the State Board of Education, and, prior to 1962, the State Board had specific statutory authority to supervise school-site selection. The proofs concerning the effect of Detroit's school construction program were,

Page 728

therefore, found to be largely applicable to show state responsibility for the segregative results.⁷

          Turning to the question of an appropriate remedy for these several constitutional violations, the District Court deferred a pending motion⁸ by intervening parent de-

Page 729

fendants to join as additional parties defendant the 85 outlying school districts in the three-county Detroit metropolitan area on the ground that effective relief could not be achieved without their presence. ⁹ The District Court concluded that this motion to join was 'premature,' since it 'has to do with relief' and no reasonably specific desegregation plan was before the court. 338 F.Supp., at 595. Accordingly, the District Court proceeded to order the Detroit Board of Education to submit desegregation plans limited to the segregation problems found to be existing within the city of Detroit. At the same time, however, the state defendants were directed to submit desegregation plans encompassing the three-county metropolitan area ¹⁰ despite the fact that the 85 outlying school

Page 730

districts of these three counties were not parties to the action and despite the fact that there had been no claim that these outlying districts had committed constitutional violations.¹¹ An effort to appeal these orders to the Court of Appeals was dismissed on the ground that the orders were not appealable. 468 F.2d 902 (CA6), cert. denied, 409 U.S. 844, 93 S.Ct. 45, 34 L.Ed.2d 83 (1972). The sequence of the ensuing actions and orders of the District Court are significant factors and will therefore be catalogued in some detail.

          Following the District Court's abrupt announcement that it planned to consider the implementation of a multidistrict, metropolitan area remedy to the segregation problems identified within the city of Detroit, the District Court was again requested to grant the outlying school districts intervention as of right on the ground that the District Court's new request for multidistrict plans 'may, as a practical matter, impair or impede (the intervenors') ability to protect' the welfare of their students. The District Court took the motions to intervene under advisement pending submission of the requested desegregation plans by Detroit and the state officials. On March 7, 1972, the District Court notified all parties and the petitioner school districts seeking intervention, that March 14, 1972, was the deadline for submission of recommendations for conditions of intervention and the

Page 731

date of the commencement of hearings on Detroit-only desegregation plans. On the second day of the scheduled hearings, March 15, 1972, the District Court granted the motions of the intervenor school districts¹² subject, inter alia, to the following conditions:

          '1. No intervenor will be permitted to assert any claim or defense previously adjudicated by the court.

          '2. No intervenor shall reopen any question or issue which has previously been decided by the court.

          '7. New intervenors are granted intervention for two principal purposes: (a) To advise the court, by brief, of the legal propriety or impropriety of considering a metropolitan plan; (b) To review any plan or plans for the desegregation of the so-called larger Detroit Metropolitan area, and submitting objections, modifications or alternatives to it or them, and in accordance with the requirements of the United States Constitution and the prior orders of this court.' 1 Joint Appendix 206 (hereinafter App.).

          Upon granting the motion to intervene, on March 15, 1972, the District Court advised the petitioning intervenors that the court had previously set March 22, 1972, as the date for the filing of briefs on the legal propriety of a 'metropolitan' plan of desegregation and, accordingly, that the intervening school districts would have one week to muster their legal arguments on the issue.¹³

Page 732

Thereafter, and following the completion of hearings on the Detroit-only desegregation plans, the District Court issued the four rulings that were the principal issues in the Court of Appeals.

          (a) On March 24, 1972, two days after the intervenors' briefs were due, the District Court issued its ruling on the question of whether it could 'consider relief in the form of a metropolitan plan, encompassing not only the City of Detroit, but the larger Detroit metropolitan area.' It rejected the state defendants' arguments that no state action caused the segregation of the Detroit schools, and the intervening suburban districts' contention that interdistrict relief was inappropriate unless the suburban districts themselves had committed violations. The court concluded:

          '(I)t is proper for the court to consider metropolitan plans directed toward the desegregation of the Detroit public schools as an alternative to the present intra-city desegregation plans before it and, in the event that the court finds such intra-city plans inadequate to desegregate such schools, the court is of the opinion that it is required to consider a metropolitan remedy for desegregation.' Pet.App. 51a.

          (b) On March 28, 1972, the District Court issued its findings and conclusions on the three Detroit-only plans submitted by the city Board and the respondents. It found that the best of the three plans 'would make the Detroit school system more identifiably Black . . . thereby increasing the flight of Whites from the city and the system.' Id., at 55a. From this the court concluded that the plan 'would not accomplish desegregation . . . within the corporate geographical limits of the city.' Id., at 56a. Accordingly, the District Court held that it 'must look beyond the limits of the Detroit school

Page 733

district for a solution to the problem,' and that '(s)chool district lines are simply matters of political convenience and may not be used to deny constitutional rights.' Id., at 57a.

          (c) During the period from March 28 to April 14, 1972, the District Court conducted hearings on a metropolitan plan. Counsel for the petitioning intervenors was allowed to participate in these hearings, but he was ordered to confine his argument to 'the size and expanse of the metropolitan plan' without addressing the intervenors' opposition to such a remedy or the claim that a finding of a constitutional violation by the intervenor districts was an essential predicate to any remedy involving them. Thereafter, on June 14, 1972, the District Court issued its ruling on the 'desegregation area' and related findings and conclusions. The court acknowledged at the outset that it had 'taken no proofs with respect to the establishment of the boundaries of the 86 public school districts in the counties (in the Detroit area), nor on the issue of whether, with the exclusion of the city of Detroit school districts, such school districts have committed acts of de jure segregation.' Nevertheless, the court designated 53 of the 85 suburban school districts plus Detroit as the 'desegregation area' and appointed a panel to prepare and submit 'an effective desegregation plan' for the Detroit schools that would encompass the entire desegregation area.¹⁴ The plan was to be based on 15 clusters, each containing part of the Detroit system and two or more suburban districts,

Page 734

and was to 'achieve the greatest degree of actual desegregation to the end that, upon implementation, no school, grade or classroom (would be) substantially disproportionate to the overall pupil racial composition.' 345 F.Supp. 914, 918 (ED Mich.1972).

          (d) On July 11, 1972, and in accordance with a recommendation by the court-appointed desegregation panel, the District Court ordered the Detroit Board of Education to purchase or lease 'at least' 295 school buses for the purpose of providing transportation under an interim plan to be developed for the 1972 1973 school year. The costs of this acquisition were to be borne by the state defendants. Pet.App. 106a—107a.

          On June 12, 1973, a divided Court of Appeals, sitting en banc, affirmed in part, vacated in part, and remanded for further proceedings. 484 F.2d 215 (CA6).¹⁵ The Court of Appeals held, first, that the record supported the District Court's findings and conclusions on the constitutional violations committed by the Detroit Board, id., at 221—238, and by the state defendants, id., at 239—241.¹⁶ It stated that the acts of racial discrimina-

Page 735

tion shown in the record are 'causally related to the substantial amount of segregation found in the Detroit school system,' id., at 241, and that 'the District Court was therefore authorized and required to take effective measures to desegregate the Detroit Public School System.' Id., at 242.

          The Court of Appeals also agreed with the District Court that 'any less comprehensive a solution than a metropolitan area plan would result in an all black school system immediately surrounded by practically all white suburban school systems, with an overwhelmingly white majority population in the total metropolitan area.' Id., at 245. The court went on to state that it could '(not) see how such segregation can be any less harmful to the minority students than if the same result were accomplished within one school district.' Ibid.

          Accordingly, the Court of Appeals concluded that 'the only feasible desegregation plan involves the crossing of the boundary lines between the Detroit School District and adjacent or nearby school districts for the limited purpose of providing an effective desegregation plan.' Id., It reasoned that such a plan would be appropriate because of the State's violations, and could be implemented because of the State's authority to control local school districts. Without further elaboration, and without any discussion of the claims that no constitutional violation by the outlying districts had been

Page 736

shown and that no evidence on that point had been allowed, the Court of Appeals held:

          '(T)he State has committed de jure acts of segregation and . . . the State controls the instrumentalities whose action is necessary to remedy the harmful effects of the State acts.' Ibid.

          An interdistrict remedy was thus held to be 'within the equity powers of the District Court.' Id., at 250.¹⁷

          The Court of Appeals expressed no views on the propriety of the District Court's composition of the metropolitan 'desegregation area.' It held that all suburban school districts that might be affected by any metropolitanwide remedy should, under Fed.Rule Civ.Proc. 19, be made parties to the case on remand and be given an opportunity to be heard with respect to the scope and implementation of such a remedy. 484 F.2d, at 251—252. Under the terms of the remand, however, the District Court was not 'required' to receive further evidence on the issue of segregation in the Detroit schools or on the propriety of a Detroit-only remedy, or on the question of whether the affected districts had committed any violation of the constitutional rights of Detroit pupils or others. Id., at 252. Finally, the Court of Appeals vacated the District Court's order directing the acquisition of school buses, subject to the right of the District Court to consider reimposing the order 'at the appropriate time.' Ibid.

Page 737

          Ever since Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), judicial consideration of school desegregation cases has begun with the standard:

          '(I)n the field of public education the doctrine of 'separate but equal' has no place. Separate educational facilities are inherently unequal.' Id., at 495, 74 S.Ct., at 692.

          This has been reaffirmed time and again as the meaning of the Constitution and the controlling rule of law.

          The target of the Brown holding was clear and forthright: the elimination of state-mandated or deliberately maintained dual school systems with certain schools for Negro pupils and others for white pupils. This duality and racial segregation were held to violate the Constitution in the cases subsequent to 1954, including particularly Green v. County School Board of New Kent County, 391 U.S. 430, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968); Raney v. Board of Education, 391 U.S. 443, 88 S.Ct. 1697, 20 L.Ed.2d 727 (1968); Monroe v. Board of Comm'rs, 391 U.S. 450, 88 S.Ct. 1700, 20 L.Ed.2d 733 (1968); Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971); Wright v. Council of the City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972); United States v. Scotland Neck City Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75 (1972).

          The Swann case, of course, dealt

          'with the problem of defining in more precise terms than heretofore the scope of the duty of school authorities and district courts in implementing Brown I and the mandate to eliminate dual systems and establish unitary systems at once.' 402 U.S., at 6, 91 S.Ct., at 1271.

          In Brown v. Board of Education, 349 U.S. 294, 75 S.Ct. 753, 99 L.Ed. 1083 (1955) (Brown II), the Court's first encounter with the problem of remedies in school desegregation cases, the Court noted:

          'In fashioning and effectuating the decrees, the courts will be guided by equitable principles. Tra-

Page 738

          ditionally, equity has been characterized by a practical flexibility in shaping its remedies and by a facility for adjusting and reconciling public and private needs.' Id., at 300, 75 S.Ct., at 756 (footnote omitted).

          In further refining the remedial process, Swann held, the task is to correct, by a balancing of the individual and collective interests, 'the condition that offends the Constitution.' A federal remedial power may be exercised 'only on the basis of a constitutional violation' and, '(a)s with any equity case, the nature of the violation determines the scope of the remedy.' 402 U.S., at 16, 91 S.Ct., at 1276.

          Proceeding from these basic principles, we first note that in the District Court the complainants sought a remedy aimed at the condition alleged to offend the Constitution—the segregation within the Detroit City School District. ¹⁸ The court acted on this theory of the case and in its initial ruling on the 'Desegregation Area' stated:

          'The task before this court, therefore, is now, and . . . has always been, now to desegregate the Detroit public schools.' 345 F.Supp., at 921.

          Thereafter, however, the District Court abruptly rejected the proposed Detroit-only plans on the ground that 'while (they) would provide a racial mix more in keeping with the Black-White proportions of the student population (they) would accentuate the racial identifiability of the

Page 739

(Detroit) district as a Black school system, and would not accomplish desegregation.' Pet.App., 56a. '(T)he racial composition of the student body is such,' said the court, 'that the plan's implementation would clearly make the entire Detroit public school system racially identifiable' (Id., at 54a), 'leav(ing) many of its schools 75 to 90 per cent Black.' Id., at 55a. Consequently, the court reasoned, it was imperative to 'look beyond the limits of the Detroit school district for a solution to the problem of segregation in the Detroit public schools . . .' since '(s)chool district lines are simply matters of political convenience and may not be used to deny constitutional rights.' Id., at 57a. Accordingly, the District Court proceeded to redefine the relevant area to include areas of predominantly white pupil population in order to ensure that 'upon implementation, no school, grade or classroom (would be) substantially disproportionate to the overall pupil racial composition' of the entire metropolitan area.

          While specifically acknowledging that the District Court's findings of a condition of segregation were limited to Detroit, the Court of Appeals approved the use of a metropolitan remedy largely on the grounds that it is

          'impossible to declare 'clearly erroneous' the District Judge's conclusion that any Detroit only segregation plan will lead directly to a single segregated Detroit school district overwhelmingly black in all of its schools, surrounded by a ring of suburbs and suburban school districts overwhelmingly white in composition in a State in which the racial composition is 87 per cent white and 13 per cent black.' 484 F.2d, at 249.

          Viewing the record as a whole, it seems clear that the District Court and the Court of Appeals shifted the pri-

Page 740

mary focus from a Detroit remedy to the metropolitan area only because of their conclusion that total desegregation of Detroit would not produce the racial balance which they perceived as desirable. Both courts proceeded on an assumption that the Detroit schools could not be truly desegregated—in their view of what constituted desegregation—unless the racial composition of the student body of each school substantially reflected the racial composition of the population of the metropolitan area as a whole. The metropolitan area was then defined as Detroit plus 53 of the outlying school districts. That this was the approach the District Court expressly and frankly employed is shown by the order which expressed the court's view of the constitutional standard:

          'Within the limitations of reasonable travel time and distance factors, pupil reassignments shall be effected within the clusters described in Exhibit P.M. 12 so as to achieve the greatest degree of actual desegregation to the end that, upon implementation, no school, grade or classroom (will be) substantially disproportionate to the overall pupil racial composition.' 345 F.Supp., st 918 (emphasis added).

          In Swann, which arose in the context of a single independent school district, the Court held:

          'If we were to read the holding of the District Court to require, as a matter of substantive constitutional right, any particular degree of racial balance or mixing, that approach would be disapproved and we would be obliged to reverse.' 402 U.S., at 24, 91 S.Ct., at 1280.

          The clear import of this language from Swann is that desegregation, in the sense of dismantling a dual school system, does not require any particular racial balance in

Page 741

each 'school, grade or classroom.'¹⁹ See Spencer v. Kugler, 404 U.S. 1027, 92 S.Ct. 707, 30 L.Ed.2d 723 (1972).

          Here the District Court's approach to what constituted 'actual desegregation' raises the fundamental question, not presented in Swann, as to the circumstances in which a federal court may order desegregation relief that embraces more than a single school district. The court's analytical starting point was its conclusion that school district lines are no more than arbitrary lines on a map drawn 'for political convenience.' Boundary lines may be bridged where there has been a constitutional violation calling for interdistrict relief, but the nation that school district lines may be casually ignored or treated as a mere administrative convenience is contrary to the history of public education in our country. No single tradition in public education is more deeply rooted than local control over the operation of schools; local autonomy has long been thought essential both to the maintenance of community concern and support for public schools and to

Page 742

quality of the educational process. See Wright v. Council of the City of Emporia, 407 U.S., at 469, 92 S.Ct., at 2206. Thus, in San Antonio Independent School District v. Rodriguez, 411 U.S. 1, 50, 93 S.Ct. 1278, 1305, 36 L.Ed.2d 16 (1973), we observed that local control over the educational process affords citizens an opportunity to participate in decision-making, permits the structuring of school programs to fit local needs, and encourages 'experimentation, innovation, and a healthy competition for educational excellence.'

          The Michigan educational structure involved in this case, in common with most States, provides for a large measure of local control,²⁰ and a review of the scope and character of these local powers indicates the extent to which the interdistrict remedy approved by the two courts could disrupt and alter the structure of public edu-

Page 743

cation in Michigan. The metropolitan remedy would require, in effect, consolidation of 54 independent school districts historically administered as separate units into a vast new super school district. See n. 10, supra. Entirely apart from the logistical and other serious problems attending large-scale transportation of students, the consolidation would give rise to an array of other problems in financing and operating this new school system. Some of the more obvious questions would be: What would be the status and authority of the present popularly elected school boards? Would the children of Detroit be within the jurisdiction and operating control of a school board elected by the parents and residents of other districts? What board or boards would levy taxes for school operations in these 54 districts constituting the consolidated metropolitan area? What provisions could be made for assuring substantial equality in tax levies among the 54 districts, if this were deemed requisite? What provisions would be made for financing? Would the validity of long-term bonds be jeopardized unless approved by all of the component districts as well as the State? What body would determine that portion of the curricula now left to the discretion of local school boards? Who would establish attendance zones, purchase school equipment, locate and construct new schools, and indeed attend to all the myriad day-to-day decisions that are necessary to school operations affecting potentially more than three-quarters of a million pupils? See n. 10, supra.

          It may be suggested that all of these vital operational problems are yet to be resolved by the District Court, and that this is the purpose of the Court of Appeals' proposed remand. But it is obvious from the scope of the interdistrict remedy itself that absent a complete restructuring of the laws of Michigan relating to school districts the District Court will become first, a de facto

Page 744

'legislative authority' to resolve these complex questions, and then the 'school superintendent' for the entire area. This is a task which few, if any, judges are qualified to perform and one which would deprive the people of control of schools through their elected representatives.

          Of course, no state law is above the Constitution. School district lines and the present laws with respect to local control, are not sacrosanct and if they conflict with the Fourteenth Amendment federal courts have a duty to prescribe appropriate remedies. See, e.g., Wright v. Council of the City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972); United States v. Scotland Neck City Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75 (1972) (state or local officials prevented from carving out a new school district from an existing district that was in process of dismantling a dual school system); cf. Haney v. County Board of Education of Sevier County, 429 F.2d 364 (CA8 1970) (State contributed to separation of races by drawing of school district lines); United States v. Texas, 321 F.Supp. 1043 (ED Tex.1970), aff'd, 447 F.2d 441 (CA5 1971), cert. denied sub nom. Edgar v. United States, 404 U.S. 1016, 92 S.Ct. 675, 30 L.Ed.2d 663 (1972) (one or more school districts created and maintained for one race). But our prior holdings have been confined to violations and remedies within a single school district. We therefore turn to address, for the first time, the validity of a remedy mandating cross-district or interdistrict consolidation to remedy a condition of segregation found to exist in only one district.

          The controlling principle consistently expounded in our holdings is that the scope of the remedy is determined by the nature and extent of the constitutional violation. Swann, 402 U.S., at 16, 91 S.Ct., at 1276. Before the boundaries of separate and autonomous school districts may be set aside by consolidating the separate units for remedial purposes or by imposing a cross-district remedy, it must

Page 745

first be shown that there has been a constitutional violation within one district that produces a significant segregative effect in another district. Specifically, it must be shown that racially discriminatory acts of the state or local school districts, or of a single school district have been a substantial cause of interdistrict segregation. Thus an interdistrict remedy might be in order where the racially discriminatory acts of one or more school districts caused racial segregation in an adjacent district, or where district lines have been deliberately drawn on the basis of race. In such circumstances an interdistrict remedy would be appropriate to eliminate the interdistrict segregation directly caused by the constitutional violation. Conversely, without an interdistrict violation and interdistrict effect, there is no constitutional wrong calling for an interdistrict remedy.

          The record before us, voluminous as it is, contains evidence of de jure segregated conditions only in the Detroit schools; indeed, that was the theory on which the litigation was initially based and on which the District Court took evidence. See supra at 725—726. With no showing of significant violation by the 53 outlying school districts and no evidence of any interdistrict violation or effect, the court went beyond the original theory of the case as framed by the pleadings and mandated a metropolitan area remedy. To approve the remedy ordered by the court would impose on the outlying districts, not shown to have committed any constitutional violation, a wholly impermissible remedy based on a standard not hinted at in Brown I and II or any holding of this Court.

          In dissent, Mr. Justice WHITE and Mr. Justice MARSHALL undertake to demonstrate that agencies having statewide authority participated in maintaining the dual school system found to exist in Detroit. They are apparently of the view that once such participation is

Page 746

shown, the District Court should have a relatively free hand to reconstruct school districts outside of Detroit in fashioning relief. Our assumption, arguendo, see infra, p. 748, that state agencies did participate in the maintenance of the Detroit system, should make it clear that it is not on this point that we part company. ²¹ The difference between us arises instead from established doctrine laid down by our cases. Brown, supra; Green, supra; Swann, supra; Scotland Neck, supra; and Emporia, supra, each addressed the issue of constitutional wrong in terms of an established geographic and administrative school system populated by both Negro and white children. In such a context, terms such as 'unitary' and 'dual' systems, and 'racially identifiable schools,' have meaning, and the necessary federal authority to remedy the constitutional wrong is firmly established. But the remedy is necessarily designed, as all remedies are, to restore the victims of discriminatory conduct to the position they would have occupied in the absence of such conduct. Disparate treatment of white and Negro students occurred within the Detroit school system, and not elsewhere, and on this record the remedy must be limited to that system. Swann, supra, 402 U.S., at 16, 91 S.Ct., at 1276.

          The constitutional right of the Negro respondents residing in Detroit is to attend a unitary school system in that district. Unless petitioners drew the district lines in a discriminatory fashion. or arranged for white stu-

Page 747

dents residing in the Detroit district to attend schools in Oakland and Macomb Counties, they were under no constitutional duty to make provisions for Negro students to do so. The view of the dissenters, that the existence of a dual system in Detroit can be made the basis for a decree requiring cross-district transportation of pupils, cannot be supported on the grounds that it represents merely the devising of a suitably flexible remedy for the violation of rights already established by our prior decisions. It can be supported only by drastic expansion of the constitutional right itself, an expansion without any support in either constitutional principle or precedent.²²

Page 748

          We recognize that the six-volume record presently under consideration contains language and some specific incidental findings thought by the District Court to afford a basis for interdistrict relief. However, these comparatively isolated findings and brief comments concern only one possible interdistrict violation and are found in the context of a proceeding that, as the District Court conceded, included no proof of segregation practiced by any of the 85 suburban school districts surrounding Detroit. The Court of Appeals, for example, relied on five factors which, it held, amounted to unconstitutional state action with respect to the violations found in the Detroit system:

          (1) It held the State derivatively responsible for the Detroit Board's violations on the theory that actions of Detroit as a political subdivision of the State were attributable to the State. Accepting, arguendo, the correctness of this finding of state responsibility for the segregated conditions within the city of Detroit, it does not follow that an interdistrict remedy is constitutionally justified or required. With a single exception, discussed later, there has been no showing that either the State or any of the 85 outlying districts engaged in activity that had a cross-district effect. The boundaries of the Detroit School District, which are coterminous with the boundaries of the city of Detroit, were established over a century ago by neutral legislation when the city was incorporated; there is no evidence in the record, nor is there any suggestion by the respondents, that either the original boundaries of the Detroit School District, or any other school district in Michigan, were established for the purpose of creating, maintaining, or perpetuating segregation of races. There is no claim and there is no evidence hinting that petitioner outlying schools districts and their

Page 749

processors, or the 30-odd other school districts in the tricounty area—but outside the District Court's 'desegregation area'—have ever maintained or operated anything but unitary school systems. Unitary school systems have been required for more than a century by the Michigan Constitution as implemented by state law.²³ White the schools of only one district have been affected, there is no constitutional power in the courts to decree relief balancing the racial composition of that district's schools with those of the surrounding districts.

          (2) There was evidence introduced at trial that, during the late 1950's, Carver School District, a predominantly Negro suburban district, contracted to have Negro high school students sent to a predominantly Negro school in Detroit. At the time, Carver was an independent school district that had no high school because, according to the trial evidence, 'Carver District . . . did not have a place for adequate high school facilities.' 484 F.2d., at 231. Accordingly, arrangements were made with Northern High School in the abutting Detroit School District so that the Carver high school students could obtain a secondary school education. In 1960 the Oak Park School District, a predominantly white suburban district, annexed the predominantly Negro Carver School District, through the initiative of local officials.

Page 750

Ibid. There is, of course, no claim that the 1960 annexation had a segregative purpose or result or that Oak Park now maintains a dual system.

          According to the Court of Appeals, the arrangement during the late 1950's which allowed Carver students to be educated within the Detroit District was dependent upon the 'tacit or express' approval of the State Board of Education and was the result of the refusal of the white suburban districts to accept the Carver students. Although there is nothing in the record supporting the Court of Appeals' supposition that suburban white schools refused to accept the Carver students, it appears that this situation, whether with or without the State's consent, may have had a segregative effect on the school populations of the two districts involved. However, since 'the nature of the violation determines the scope of the remedy,' Swann, 402 U.S., at 16, 91 S.Ct., at 1276, this isolated instance effecting two of the school districts would not justify the broad metropolitanwide remedy contemplated by the District Court and approved by the Court of Appeals, particularly since it embraced potentially 52 districts having no responsibility for the arrangement and involved 503,000 pupils in addition to Detroit's 276,000 students.

          (3) The Court of Appeals cited the enactment of state legislation (Act 48) which had the effect of rescinding Detroit's voluntary desegregation plan (the April 7 Plan). That plan, however, affected only 12 of 21 Detroit high schools and had no causal connection with the distribution of pupils by race between Detroit and the other school districts within the tricounty area.

          (4) The court relied on the State's authority to supervise schoolsite selection and to approve building construction as a basis for holding the State responsible for the segregative results of the school construction program in Detroit. Specifically, the Court of Appeals asserted

Page 751

that during the period between 1949 and 1962 the State Board of Education exercised general authority as overseer of site acquisitions by local boards for new school construction, and suggested that this state-approved school construction 'fostered segregation throughout the Detroit Metropolitan area.' 484 F.2d, at 241. This brief comment, however, is not supported by the evidence taken at trial since that evidence was specifically limited to proof that schoolsite acquisition and school construction within the city of Detroit produced de jure segregation within the city itself. Id., at 235—238. Thus, there was no evidence suggesting that the State's activities with respect to either school construction or site acquisition within Detroit affected the racial composition of the school population outside Detroit or, conversely, that the State's school construction and site acquisition activities within the outlying districts affected the racial composition of the schools within Detroit.

          (5) The Court of Appeals also relied upon the District Court's finding:

          'This and other financial limitations, such as those on bonding and the working of the state aid formula whereby suburban districts were able to make far larger per pupil expenditures despite less tax effort, have created and perpetuated systematic educational inequalities.' Id., at 239.

          However, neither the Court of Appeals nor the District Court offered any indication in the record or in their opinions as to how, if at all, the availability of state-financed aid for some Michigan students outside Detroit, but not for those within Detroit, might have affected the racial character of any of the State's school districts. Furthermore, as the respondents recognize, the application of our recent ruling in San Antonio School District v. Rodriguez, 411 U.S. 1, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973), to this state education financing system is questionable, and this issue was not

Page 752

addressed by either the Court of Appeals or the District Court. This, again, underscores the crucial fact that the theory upon which the the case proceeded related solely to the establishment of Detroit city violations as a basis for desegregating Detroit schools and that, at the time of trial, neither the parties nor the trial judge was concerned with a foundation for interdistrict relief.²⁴

          Petitioners have urged that they were denied due process by the manner in which the District Court limited their participation after intervention was allowed, thus precluding adequate opportunity to present evidence that they had committed no acts having a segregative effect in Detroit. In light of our holding that, absent an interdistrict violation, there is no basis for an interdistrict remedy, we need not reach these claims. It is clear, however, that the District Court, with the approval of the Court of Appeals, has provided an interdistrict remedy in the face of a record which shows no constitutional violations that would call for equitable relief except within the city of Detroit. In these circumstances there was no occasion for the parties to address, or for the District Court to consider whether there were racially discriminatory acts for which any of the 53 outlying districts were responsible and which had direct and significant segregative effect on schools of more than one district.

          We conclude that the relief ordered by the District Court and affirmed by the Court of Appeals was based upon an erroneous standard and was unsupported by record evidence that acts of the outlying districts effected the discrimination found to exist in the schools of De-

Page 753

troit. Accordingly, the judgment of the Court of Appeals is reversed and the case is remanded for further proceedings consistent with this opinion leading to prompt formulation of a decree directed to eliminating the segregation found to exist in Detroit city schools, a remedy which has been delayed since 1970.

          Reversed and remanded.

           Mr. Justice STEWART, concurring.

          In joining the opinion of the Court, I think it appropriate, in view of some of the extravagant language of the dissenting opinions, to state briefly my understanding of what it is that the Court decides today.

          The respondents commenced this suit in 1970, claiming only that a constitutionally impermissible allocation of educational facilities along racial lines had occurred in public schools within a single school district whose lines were coterminous with those of the city of Detroit. In the course of the subsequent proceedings, the District Court found that public school officials had contributed to racial segregation within that district by means of improper use of zoning and attendance patterns, optional-attendance areas, and building and site selection. This finding of a violation of the Equal Protection Clause was upheld by the Court of Appeals, and is accepted by this Court today. See ante, at 738 n. 18. In the present posture of the case, therefore, the Court does not deal with questions of substantive constitutional law. The basic issue now before the Court concerns, rather, the appropriate exercise of federal equity jurisdiction.¹

Page 754

          No evidence was adduced and no findings were made in the District Court concerning the activities of school officials in districts outside the city of Detroit, and no school officials from the outside districts even participated in the suit until after the District Court had made the initial determination that is the focus of today's decision. In spite of the limited scope of the inquiry and the findings, the District Court concluded that the only effective remedy for the constitutional violations found to have existed within the city of Detroit was a desegregation plan calling for busing pupils to and from school districts outside the city. The District Court found that any desegregation plan operating wholly "within the corporate geographical limits of the city" would be deficient since it "would clearly make the entire Detroit public school system racially identifiable as Black." 484 F.2d 215, 244, 243. The Court of Appeals, in affirming the decision that an interdistrict remedy was necessary, noted that a plan limited to the city of Detroit 'would result in an all black school system immediately surrounded by practically all white suburban school systems, with an overwhelmingly white majority population in the total metropolitan area.' Id., at 245.

          The courts were in error for the simple reason that the remedy they thought necessary was not commensurate with the constitutional violation found. Within a single school district whose officials have been shown to have engaged in unconstitutional racial segregation, a remedial decree that affects every individual school may be dictated by 'common sense,' see Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 203, 93 S.Ct. 2686, 2695, 37 L.Ed.2d 548, and indeed may provide the only effective means to eliminate segregation 'root and branch,' Green v. County School Board of New Kent County, 391 U.S. 430, 438, 88 S.Ct. 1689, 1693, 20 L.Ed.2d 716, and to 'effectuate a transition to a racially nondiscriminatory school

Page 755

system.' Brown v. Board of Education, 349 U.S. 294, 301, 75 S.Ct. 753, 756, 99 L.Ed. 1083. See Keyes, supra, at 198—205, 93 S.Ct., at 2692—2696. But in this case the Court of Appeals approved the concept of a remedial decree that would go beyond the boundaries of the district where the constitutional violation was found, and include schools and schoolchildren in many other school districts that have presumptively been administered in complete accord with the Constitution.

          The opinion of the Court convincingly demonstrates, ante, at 742—743, that traditions of local control of schools, together with the difficulty of a judicially supervised restructuring of local administration of schools, render improper and inequitable such an interdistrict response to a constitutional violation found to have occurred only within a single school district.

          This is not to say, however, that an interdistrict remedy of the sort approved by the Court of Appeals would not be proper, or even necessary, in other factual situations. Were it to be shown, for example, that state officials had contributed to the separation of the races by drawing or redrawing school district lines, see Haney v. County Board of Education of Sevier County, 429 F.2d 364; cf. Wright v. Council of the City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51; United States v. Scotland Neck City Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75; by transfer of school units between districts, United States v. Texas, 321 F.Supp. 1043, aff'd, 447 F.2d 441; Turner v. Warren County Board of Education, 313 F.Supp. 380; or by purposeful racially discriminatory use of state housing or zoning laws, then a decree calling for transfer of pupils across district lines or for restructuring of district lines might well be appropriate.

          In this case, however, no such interdistrict violation was shown. Indeed, no evidence at all concerning the administration of schools outside the city of Detroit was presented other than the fact that these schools contained

Page 756

a higher proportion of white pupils than did the schools within the city. Since the mere fact of different racial compositions in contiguous districts does not itself imply or constitute a violation of the Equal Protection Clause in the absence of a showing that such disparity was imposed, fostered, or encouraged by the State or its political subdivisions, it follows that no interdistrict violation was shown in this case.² The formulation of an inter-distrit remedy was thus simply not responsive to the factual record before the District Court and was an abuse of that court's equitable powers.

Page 757

          In reversing the decision of the Court of Appeals this Court is in no way turning its back on the proscription of state-imposed segregation first voiced in Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873, or on the delineation of remedial powers and duties most recently expressed in Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554. In Swann the Court addressed itself to the range of equitable remedies available to the courts to effectuate the desegregation mandated by Brown and its progeny, noting that the task in choosing appropriate relief is 'to correct . . . the condition that offends the Constitution,' and that 'the nature of the violation determines the scope of the remedy . . .' Id., at 16, 91 S.Ct., at 1276.

          The disposition of this case thus falls squarely under these principles. The only 'condition that offends the Constitution' found by the District Court in this case is the existence of officially supported segregation in and among public schools in Detroit itself. There were no findings that the differing racial composition between schools in the city and in the outlying suburbs was caused by official activity of any sort. It follows that the decision to include in the desegregation plan pupils from school districts outside Detroit was not predicated upon any constitutional violation involving those school districts. By approving a remedy that would reach beyond the limits of the city of Detroit to correct a constitutional violation found to have occurred solely within that city the Court of Appeals thus went beyond the governing equitable principles established in this Court's decisions.

           Mr. Justice DOUGLAS, dissenting.

          The Court of Appeals has acted responsibly in these cases and we should affirm its judgment. This was the fourth time the case was before it over a span of less than three years. The Court of Appeals affirmed the District

Page 758

Court on the issue of segregation and on the 'Detroit-only' plans of desegregation. The Court of Appeals also approved in principle the use of a metropolitan area plan, vacating and remanding only to allow the other affected school districts to be brought in as parties, and in other minor respects.

          We have before us today no plan for integration. The only orders entered so far are interlocutory. No new principles of law are presented here. Metropolitan treatment of metropolitan problems is commonplace. If this were a sewage problem or a water problem, or an energy problem, there can be no doubt that Michigan would stay well within federal constitutional bounds if it sought a metropolitan remedy. In Bradley v. School Board of City of Richmond, 4 Cir., 462 F.2d 1058, aff'd by an equally divided Court, 412 U.S. 92, 93 S.Ct. 1952, 36 L.Ed.2d 771, we had a case involving the Virginia school system where local school boards had 'exclusive jurisdiction' of the problem, not 'the State Board of Education,' 462 F.2d, at 1067. Here the Michigan educational system is unitary, maintained and supported by the legislature and under the general supervision of the State Board of Education.¹ The State controls the boundaries of school districts.² The State supervises schoolsite selection.³ The construction is done through municipal bonds approved by several state agencies.⁴ Education in Michigan is a state project with very little completely local control,⁵ except that the schools are financed locally, not on a statewide basis. Indeed

Page 759

the proposal to put school funding in Michigan on a statewide basis was defeated at the polls in November 1972. ⁶ Yet the school districts by state law are agencies of the State.⁷ State action is indeed challenged as violating the Equal Protection Clause. Whatever the reach of that claim may be, it certainly is aimed at discrimination based on race.

          Therefore as the Court of Appeals held there can be no doubt that as a matter of Michigan law the State itself has the final say as to where and how school district lines should be drawn.⁸

          When we rule against the metropolitan area remedy we take a step that will likely put the problems of the blacks and our society back to the period that antedated the 'separate but equal' regime of Plessy v. Ferguson, 163 U.S. 537, 16 S.Ct. 1138, 41 L.Ed. 256. The reason is simple.

          The inner core of Detroit is now rather solidly black;⁹ and the blacks, we know, in many instances are likely to

Page 760

be poorer,¹⁰ just as were the Chicanos in San Antonio School District v. Rodriguez, 411 U.S. 1, 93 S.Ct. 1278, 36 L.Ed.2d 16. By that decision the poorer school districts¹¹ must pay their own way. It is therefore a foregone conclusion that we have now given the States a formula whereby the poor must pay their own way.¹²

Page 761

          Today's decision, given Rodriguez, means that there is no violation of the Equal Protection Clause though the schools are segregated by race and though the black schools are not only 'separate' but 'inferior.'

          So far as equal protection is concerned we are now in a dramatic retreat from the 7-to-1 decision in 1896 that blacks could be segregated in public facilities, provided they received equal treatment.

          As I indicated in Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 214—217, 93 S.Ct. 2686, 2700—2701, 37 L.Ed.2d 548, there is so far as the school cases go no constitutional difference between de facto and de jure segregation. Each school board performs state action for Fourteenth Amendment purposes when it draws the lines that confine it to a given area, when it builds schools at particular sites, or when it allocates students. The creation of the school districts in Metropolitan Detroit either maintained existing segregation or caused additional segregation. Restrictive covenants maintained by state action or inaction build black ghettos. It is state action when public funds are dispensed by housing agencies to build racial ghettos. Where a community is racially mixed and school authorities segregate schools, or assign black teachers to black schools or close schools in fringe areas and build new schools in black areas and in more distant white areas, the State creates and nurtures a segregated school system, just as surely as did those States involved in Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873, when they maintained dual school systems.

          All these conditions and more were found by the District Court to exist. The issue is not whether there should be racial balance but whether the State's use of

Page 762

various devices that end up with black schools and white schools brought the Equal Protection Clause into effect. Given the State's control over the educational system in Michigan, the fact that the black schools are in one district and the white schools are in another is not controlling—either constitutionally or equitably.¹³ No specific plan has yet been adopted. We are still at an interlocutory stage of a long drawn-out judicial effort at school desegregation. It is conceivable that ghettos develop on their own without any hint of state action. But since Michigan by one device or another has over the years created black school districts and white school districts, the task of equity is to provide a unitary system for the affected area where, as here, the State washes its hands of its own creations.

          Mr. Justice WHITE, with whom Mr. Justice DOUGLAS, Mr. Justice BRENNAN, and Mr. Justice MARSHALL join, dissenting.

          The District Court and the Court of Appeals found that over a long period of years those in charge of the Michigan public schools engaged in various practices calculated to effect the segregation of the Detroit school system. The Court does not question these findings, nor could it reasonably do so. Neither does it question the obligation of the federal courts to devise a feasible and effective remedy. But it promptly cripples the ability of the judiciary to perform this task, which is of fundamental importance to our constitutional system, by

Page 763

fashioning a strict rule that remedies in school cases must stop at the school district line unless certain other conditions are met. As applied here, the remedy for unquestioned violations of the protection rights of Detroit's Negroes by the Detroit School Board and the State of Michigan must be totally confined to the limits of the school district and may not reach into adjoining or surrounding districts unless and until it is proved there has been some sort of 'interdistrict violation'—unless unconstitutional actions of the Detroit School Board have had a segregative impact on other districts, or unless the segregated condition of the Detroit schools has itself been influenced by segregative practices in those surrounding districts into which it is proposed to extend the remedy.

          Regretfully, and for several reasons, I can join neither the Court's judgment nor its opinion. The core of my disagreement is that deliberate acts of segregation and their consequences will go unremedied, not because a remedy would be infeasible or unreasonable in terms of the usual criteria governing school desegregation cases, but because an effective remedy would cause what the Court considers to be undue administrative inconvenience to the State. The result is that the State of Michigan, the entity at which the Fourteenth Amendment is directed, has successfully insulated itself from its duty to provide effective desegregation remedies by vesting sufficient power over its public schools in its local school districts. If this is the case in Michigan, it will be the case in most States.

          There are undoubted practical as well as legal limits to the remedial powers of federal courts in school desegregation cases. The Court has made it clear that the achievement of any particular degree of racial balance in the school system is not required by the Constitution;

Page 764

nor may it be the primary focus of a court in devising an acceptable remedy for de jure segregation. A variety of procedures and techniques are available to a district court engrossed in fashioning remedies in a case such as this; but the courts must keep in mind that they are dealing with the process of educating the young, including the very young. The task is not to devise a system of pains and penalties to punish constituttional violations brought to light. Rather, it is to desegregate an educational system in which the races have been kept apart, without, at the same time, losing sight of the central educational function of the schools.

          Viewed in this light, remedies calling for school zoning, pairing, and pupil assignments, become more and more suspect as they require that schoolchildren spend more and more time in buses going to and from school and that more and more educational dollars be diverted to transportation systems. Manifestly, these considerations are of immediate and urgent concern when the issue is the desegregation of a city school system where residential patterns are predominantly segregated and the respective areas occupied by blacks and whites are heavily populated and geographically extensive. Thus, if one postulates a metropolitan school system covering a sufficiently large area, with the population evenly divided between whites and Negroes and with the races occupying identifiable residential areas, there will be very real practical limits on the extent to which racially identifiable schools can be eliminated within the school district. It is also apparent that the larger the proportion of Negroes in the area, the more difficult it would be to avoid having a substantial number of all-black or nearly all-black schools.

          The Detroit school district is both large and heavily populated. It covers 139.6 square miles, encircles two

Page 765

entirely separate cities and school districts, and surrounds a third city on three sides. Also, whites and Negroes live in identifiable areas in the city. The 1970 public school enrollment in the city school district totaled 289,763 and was 63.6% Negro and 34.8% white. ¹ If 'racial balance' were achieved in every school in the district, each school would be approximately 64% Negro. A remedy confined to the district could achieve no more desegregation. Furthermore, the proposed intracity remedies were beset with practical problems. None of the plans limited to the school district was satisfactory to the District Court. The most promising proposal, submitted by respondents, who were the plaintiffs in the District Court, would 'leave many of its schools 75 to 90 per cent Black.' 484 F.2d 215, 244 (CA6 1973).² Transportation on a 'vast scale' would be required; 900 buses would have to be purchased for the transportation of pupils who are not now bused. Id., at 243. The District Court also found that the plan 'would change a school system which is now Black and White to one that would be perceived as Black, thereby increasing the flight of Whites from the city and the system, thereby increasing the Black student population.' Id., at 244. For the District Court, '(t)he conclusion, under the evidence in this case, is inescapable that relief of segregation in the public schools of the

Page 766

City of Detroit cannot be accomplished within the corporate geographical limits of the city.' Ibid.

          The District Court therefore considered extending its remedy of the suburbs. After hearings, it concluded that a much more effective desegregation plan could be implemented if the suburban districts were included. In proceeding to design its plan on the basis that student bus rides to and from school should not exceed 40 minutes each way as a general matter, the court's express finding was that '(f)or all the reasons stated heretofore including time, distance, and transportation factors—desegregation within the area described is physically easier and more practicable and feasible, than desegregation efforts limited to the corporate geographic limits of the city of Detroit.' 345 F.Supp. 914, 930 (ED Mich.1972).

          The Court of Appeals agreed with the District Court that the remedy must extend beyond the city limits of Detroit. It concluded that '(i)n the instant case the only feasible desegregation plan involves the crossing of the boundary lines between the Detroit School District and adjacent or nearby school districts for the limited purpose of providing an effective desegregation plan.' 484 F.2d, at 249. (Emphasis added.) It also agreed that 'any Detroit only desegregation plan will lead directly to a single segregated Detroit school district overwhelmingly black in all of its schools, surrounded by a ring of suburbs and suburban school districts overwhelmingly white in composition in a State in which the racial composition is 87 per cent white and 13 per cent black.' Ibid. There was 'more than ample support for the District Judge's findings of unconstitutional segregation by race resulting in major part from action and inaction of public authorities, both local and State. . . . Under this record a remedial order of a court of equity which left the Detroit school system overwhelmingly black (for the fore-

Page 767

seeable future) surrounded by suburban school systems overwhelmingly white cannot correct the constitutional violations herein found.' Id., at 250. To conclude otherwise, the Court of Appeals announced, would call up 'haunting memories of the now long overruled and discredited 'separate but equal doctrine' of Plessy v. Ferguson, 163 U.S. 537 (16 S.Ct. 1138, 41 L.Ed. 256) . . . (1896),' and 'would be opening a way to nullify Brown v. Board of Education which overruled Plessy. . . .' 484 F.2d, at 249.

          This Court now reverses the Court of Appeals. It does not question the District Court's findings that any feasible Detroit-only plan would leave many schools 75 to 90 percent black and that the district would become progressively more black as whites left the city. Neither does the Court suggest that including the suburbs in a desegregation plan would be impractical or infeasible because of educational considerations, because of the number of children requiring transportation, or because of the length of their rides. Indeed, the Court leaves unchallenged the District Court's conclusion that a plan including the suburbs would be physically easier and more practical and feasible than a Detroit-only plan. Whereas the most promising Detroit-only plan, for example, would have entailed the purchase of 900 buses, the metropolitan plan would involve the acquisition of no more than 350 new vehicles.

          Despite the fact that a metropolitan remedy, if the findings of the District Court accepted by the Court of Appeals are to be credited, would more effectively desegregate the Detroit schools, would prevent resegregation,³ and would be easier and more feasible from many

Page 768

standpoints, the Court fashions out of whole cloth an arbitrary rule that remedies for constitutional violations occurring in a single Michigan school district must stop at the school district line. Apparently, no matter how much less burdensome or more effective and efficient in many respects, such as transportation, the metropolitan plan might be, the school district line may not be crossed. Otherwise, it seems, there would be too much disruption of the Michigan scheme for managing its educational system, too much confusion, and too much administrative burden.

          The District Court, on the scene and familiar with local conditions, had a wholly different view. The Court of Appeals also addressed itself at length to matters of local law and to the problems that interdistrict remedies might present to the State of Michigan. Its conclusion, flatly contrary to that of this Court, was that 'the constitutional right to equality before the law (is not) hemmed in by the boundaries of a school district' and that an interdistrict remedy

          'is supported by the status of school districts under Michigan law and by the historical control exercised over local school districts by the legislature of Michigan and by State agencies and officials . . .. (I)t is well established under the Constitution and laws of Michigan that the public school system is a State function and that local school districts are instrumentalities of the State created for administrative convenience.'⁴ 484 F.2d, at 245—246.

Page 769

          I am surprised that the Court, sitting at this distance from the State of Michigan, claims better insight than the Court of Appeals and the District Court as to whether an interdistrict remedy for equal protection violations practiced by the State of Michigan would involve undue difficulties for the State in the management of its public schools. In the area of what constitutes an acceptable desegregation plan, 'we must of necessity rely to a large extent, as this Court has for more than 16 years, on the informed judgment of the district courts in the first instance and on courts of appeals.' Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 28, 91 S.Ct. 1267, 1282, 28 L.Ed.2d 554 (1971). Obviously, whatever difficulties there might be, they are surmountable; for the Court itself concedes that, had there been sufficient evidence of an interdistrict violation, the District Court could have fashioned a single remedy for the districts implicated rather than a different remedy for each district

Page 770

in which the violation had occurred or had an impact.

          I am even more mystified as to how the Court can ignore the legal reality that the constitutional violations, even if occurring locally, were committed by governmental entities for which the State is responsible and that it is the State that must respond to the command of the Fourteenth Amendment. An interdistrict remedy for the infringements that occurred in this case is well within the confines and powers of the State, which is the governmental entity ultimately responsible for desegregating its schools. The Michigan Supreme Court has observed that '(t)he school district is a State agency,' Attorney General ex rel. Kies v. Lowrey, 131 Mich. 639, 644, 92 N.W. 289, 290 (1902), and that "(e)ducation in Michigan belongs to the State. It is no part of the local self-government inherent in the township or municipality, except so far as the legislature may choose to make it such. The Constitution has turned the whole subject over to the legislature. . . ." Attorney General ex rel. Lacharias v. Detroit Board of Education, 154 Mich. 584, 590, 118 N.W. 606, 609 (1908).

          It is unnecessary to catalogue at length the various public misdeeds found by the District Court and the Court of Appeals to have contributed to the present segregation of the Detroit public schools. The legislature contributed directly by enacting a statute overriding a partial high school desegregation plan voluntarily adopted by the Detroit Board of Education. Indirectly, the trial court found the State was accountable for the thinly disguised, pervasive acts of segregation committed by the Detroit Board,⁵ for Detroit's school construction

Page 771

plans that would promote segregation, and for the Detroit school district's not having funds for pupil transportation within the district. The State was also chargeable with responsibility for the transportation of Negro high school students in the late 1950's from the suburban Ferndale School District, past closer suburban and Detroit high schools with predominantly white student bodies, to a predominantly Negro high school within Detroit. Swann v. Charlotte-Mecklenburg Board of Education, supra, 402 U.S., at 20—21, 91 S.Ct. at 1278, and Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973), make abundantly clear that the tactics employed by the Detroit Board of Education, a local instrumentality of the State, violated the constitutional rights of the Negro students in Detroit's public schools and required equitable relief sufficient to accomplish the maximum, practical desegregation within the power of the political body against which the Fourteenth Amendment directs its proscriptions. No 'State' may deny any individual the equal protection of the laws; and if the Constitution and the Supremacy Clause are to have any substance at all, the courts must be free to devise workable remedies against the political entity with the effective power to determine local choice. It is also the case here that the State's legislative interdiction of Detroit's voluntary effort to desegregate its school system was unconstitutional. See North Carolina State Board of Education v. Swann, 402 U.S. 43, 91 S.Ct. 1284, 28 L.Ed.2d 586 (1971).

          The Court draws the remedial line at the Detroit school district boundary, even though the Fourteenth Amendment is addressed to the State and even though

Page 772

the State denies equal protection of the laws when its public agencies, acting in its behalf, invidiously discriminate. The State's default is 'the condition that offends the Constitution,' Swann v. Charlotte-Mecklenburg Board of Education, supra, 402 U.S., at 16, 91 S.Ct. at 1277, and state officials may therefore be ordered to take the necessary measures to completely eliminate from the Detroit public schools 'all vestiges of state-imposed segregation.' Id., at 15, 91 S.Ct. at 1275. I cannot understand, nor does the majority satisfactorily explain, why a federal court may not order an appropriate interdistrict remedy, if this is necessary or more effective to accomplish this constitutionally mandated task. As the Court unanimously observed in Swann: 'Once a right and a violation have been shown, the scope of a district court's equitable powers to remedy past wrongs is broad, for breadth and flexibility are inherent in equitable remedies.' Ibid. In this case, both the right and the State's Fourteenth Amendment violation have concededly been fully established, and there is no acceptable reason for permitting the party responsible for the constitutional violation to contain the remedial powers of the federal court within administrative boundaries over which the transgressor itself has plenary power.

          The unwavering decisions of this Court over the past 20 years support the assumption of the Court of Appeals that the District Court's remedial power does not cease at the school district line. The Court's first formulation of the remedial principles to be followed in disestablishing racially discriminatory school systems recognized the variety of problems arising from different local school conditions and the necessity for that 'practical flexibility' traditionally associated with courts of equity. Brown v. Board of Education, 349 U.S. 294, 299—301, 75 S.Ct. 753, 755 756, 99 L.Ed. 1083, (1955) (Brown II). Indeed, the district courts to which

Page 773

the Brown cases were remanded for the formulation of remedial decrees were specifically instructed that they might consider, inter alia, 'revision of school districts and attendance areas into compact units to achieve a system of determining admission to the public schools on a nonracial basis . . ..' Id., at 300—301, 75 S.Ct. at 756. The malady addressed in Brown II was the statewide policy of requiring or permitting school segregation on the basis of race, while the record here concerns segregated schools only in the city of Detroit. The obligation to rectify the unlawful condition nevertheless rests on the State. The permissible revision of school districts contemplated in Brown II rested on the State's responsibility for desegregating its unlawfully segregated schools, not on any segregative effect which the condition of segregation in one school district might have had on the schools of a neighboring district. The same situation obtains here and the same remedial power is available to the District Court.

          Later cases reinforced the clearly essential rules that state officials are fully answerable for unlawfully caused conditions of school segregation which can effectively be controlled only by steps beyond the authority of local school districts to take, and that the equity power of the district courts includes the ability to order such measures implemented. When the highest officials of the State of Arkansas impeded a federal court order to desegregate the public schools under the immediate jurisdiction of the Little Rock School Board, this Court refused to accept the local board's assertion of its good faith as a legal excuse for delay in implementing the desegregation order. The Court emphasized that 'from the point of view of the Fourteenth Amendment, they (the local school board members) stand in this litigation as the agents of the State.' Cooper v. Aaron, 358 U.S. 1, 16, 78 S.Ct. 1401, 1408, 3 L.Ed.2d 5 (1958). Per-

Page 774

haps more importantly for present purposes, the Court went on to state:

          'The record before us clearly establishes that the growth of the Board's difficulties to a magnitude beyond its unaided power to control is the product of state action. Those difficulties . . . can also be brought under control by state action.' Ibid.

          See also Griffin v. School Board, 377 U.S. 218, 228, 233—234, 84 S.Ct. 1226, 1231, 1234—1235, 12 L.Ed.2d 256 (1964).

          In the context of dual school systems, the Court subsequently made clear the 'affirmative duty to take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch' and to come forward with a desegregation plan that 'promises realistically to work now.' Green v. County School Board of New Kent County, 391 U.S. 430, 437—438, 439, 88 S.Ct. 1689, 1694, 20 L.Ed.2d 716 (1968). 'Freedom of choice' plans were rejected as acceptable desegregation measures where 'reasonably available other ways . . . promising speedier and more effective conversion to a unitary, nonracial school system . . .' exist. Id., at 441, 88 S.Ct., at 1696. Imperative insistence on immediate full desegregation of dual school systems 'to operate now and hereafter only unitary schools' was reiterated in Alexander v. Holmes County Board of Education, 396 U.S. 19, 20, 90 S.Ct. 29, 24 L.Ed.2d 19 (1969), and Carter v. West Feliciana Parish School Board, 396 U.S. 290, 90 S.Ct. 608, 24 L.Ed.2d 477 (1970).

          The breadth of the equitable authority of the district courts to accomplish these comprehensive tasks was reaffirmed in much greater detail in Swann v. Charlotte-Mecklenburg Board of Education, supra, and the companion case of Davis v. School Comm'rs of Mobile County, 402 U.S. 33, 91 S.Ct. 1289, 28 L.Ed.2d 577 (1971), where there was unanimous assent to the following propositions:

          'Having once found a violation, the district judge or school authorities should make every effort to

Page 775

          achieve the greatest possible degree of actual desegregation, taking into account the practicalities of the situation. A district court may and should consider the use of all available techniques including restructuring of attendance zones and both contiguous and noncontiguous attendance zones. . . . The measure of any desegregation plan is its effectiveness.' Id., at 37, 91 S.Ct. at 1292.

          No suggestion was made that interdistrict relief was not an available technique. In Swann v. Charlotte-Mecklenburg Board of Education itself, the Court, without dissent, recognized that the District Judge, in fulfilling his obligation to 'make every effort to achieve the greatest possible degree of actual desegregation(,) will thus necessarily be concerned with the elimination of one-race schools.' 402 U.S., at 26, 91 S.Ct., at 1281. Nor was there any dispute that to break up the dual school system, it was within the District Court's 'broad remedial powers' to employ a 'frank—and sometimes drastic—gerrymandering of school districts and attendance zones,' as well as 'pairing, 'clustering,' or 'grouping' of schools,' to desegregate the 'formerly all-Negro schools,' despite the fact that these zones might not be compact or contiguous and might be 'on opposite ends of the city.' Id., at 27, 91 S.Ct. at 1282. The school board in that case had jurisdiction over a 550-square-mile area encompassing the city of Charlotte and surrounding Mecklenburg County, North Carolina. The Mobile County, Alabama, board in Davis embraced a 1,248-squaremile area, including the city of Mobile. Yet the Court approved the District Court's authority to award countywide relief in each case in order to accomplish desegregation of the dual school system.

          Even more recently, the Court specifically rejected the claim that a new school district, which admittedly would operate a unitary school system within its borders, was beyond the reach of a court-ordered desegregation plan

Page 776

for other school districts, where the effectiveness of the plan as to the other districts depended upon the availability of the facilities and student population of the new district. In Wright v. Council of City of Emporia, 407 U.S. 451, 470, 92 S.Ct. 2196, 2207, 33 L.Ed.2d 51 (1972), we held 'that a new school district may not be created where its effect would be to impede the process of dismantling a dual system.' Mr. Justice Stewart's opinion for the Court made clear that if a proposal to erect new district boundary lines 'would impede the dismantling of the (pre-existing) dual system, then a district court, in the exercise of its remedial discretion, may enjoin it from being carried out.' Id., at 460, 92 S.Ct. at 2203. In United States v. Scotland Neck Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75 (1972), this same standard was applied to forbid North Carolina from creating a new city school district within a larger district which was in the process of dismantling a dual school system. The Court noted that if establishment of the new district were permitted, the 'traditional racial identities of the schools in the area would be maintained,' id., at 490, 92 S.Ct., at 2717.

          Until today, the permissible contours of the equitable authority of the district courts to remedy the unlawful establishment of a dual school system have been extensive, adaptable, and fully responsive to the ultimate goal of achieving 'the greatest possible degree of actual desegregation.' There are indeed limitations on the equity powers of the federal judiciary, but until now the Court had not accepted the proposition that effective enforcement of the Fourteenth Amendment could be limited by political or administrative boundary lines demarcated by the very State responsible for the constitutional violation and for the disestablishment of the dual system. Until now the Court has instead looked to practical considerations in effectuating a desegregation

Page 777

decree, such as excessive distance, transportation time, and hazards to the safety of the schoolchildren involved in a proposed plan. That these broad principles have developed in the context of dual school systems compelled or authorized by state statute at the time of Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1945) (Brown I), does not lessen their current applicability to dual systems found to exist in other contexts, like that in Detroit, where intentional school segregation does not stem from the compulsion of state law, but from deliberate individual actions of local and state school authorities directed at a particular school system. The majority properly does not suggest that the duty to eradicate completely the resulting dual system in the latter context is any less than in the former. But its reason for incapacitating the remedial authority of the federal judiciary in the presence of school district perimeters in the latter context is not readily apparent.

          The result reached by the Court certainly cannot be supported by the theory that the configuration of local governmental units is immune from alteration when necessary to redress constitutional violations. In addition to the well-established principles already noted, the Court has elsewhere required the public bodies of a State to restructure the State's political subdivisions to remedy infringements of the constitutional rights of certain members of its populace, notably in the reapportionment cases. In Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964), for example, which held that equal protection of the laws demands that the seats in both houses of a bicameral state legislature be apportioned on a population basis, thus necessitating wholesale revision of Alabama's voting districts, the Court remarked:

          'Political subdivisions of States—counties, cities, or whatever—never were and never have been con-

Page 778

          sidered as sovereign entities. Rather, they have been traditionally regarded as subordinate governmental instrumentalities created by the State to assist in the carrying out of state governmental functions.' Id., at 575, 84 S.Ct., at 1389.

          And even more pointedly, the Court declared in Gomillion v. Lightfoot, 364 U.S. 339, 334—345, 81 S.Ct. 125, 129, 5 L.Ed.2d 110 (1960), that '(l) egislative control of municipalities, no less than other state power, lies within the scope of relevant limitations imposed by the United States Constitution.

          Nor does the Court's conclusion follow from the talismanic invocation of the desirability of local control over education. Local autonomy over school affairs, in the sense of the community's participation in the decisions affecting the education of its children, is, of course, an important interest. But presently constituted school district lines do not delimit fixed and unchangeable areas of a local educational community. If restructuring is required to meet constitutional requirements, local authority may simply be redefined in terms of whatever configuration is adopted, with the parents of the children attending schools in the newly demarcated district or attendance zone continuing their participation in the policy management of the schools with which they are concerned most directly. The majority's suggestion that judges should not attempt to grapple with the administrative problems attendant on a reorganization of school attendance patterns is wholly without foundation. It is precisely this sort of task which the district courts have been properly exercising to vindicate the constitutional rights of Negro students since Brown I and which the Court has never suggested they lack the capacity to perform. Intradistrict revisions of attendance zones, and pairing and grouping of schools, are techniques unanimously approved in Swann v. Charlotte-Mecklenburg

Page 779

Board of Education which entail the same sensitivity to the interet of parents in the education their children receive as would an interditrict plan which is likely to employ the very same methods. There is no reason to suppose that the District Court, which has not yet adopted a final plan of desegregation, would not be as capable of giving or as likely to give sufficient weight to the interest in community participation in schools in an interdistrict setting, consistent with the dictates of the Fourteenth Amendment. The majority's assumption that the District Court would act otherwise is a radical departure from the practical flexibility previously left to the equity powers of the federal judiciary.

          Finally, I remain wholly unpersuaded by the Court's assertion that 'the remedy is necessarily designed, as all remedies are, to restore the victims of discriminatory conduct to the position they would have occupied in the absence of such conduct.' Ante, p. 746. In the first place, under this premise the Court's judgment is itself infirm; for had the Detroit school system not followed an official policy of segregation throughout the 1950's and 1960's, Negroes and whites would have been going to school together. There would have been no, or at least not as many, recognizable Negro schools and no, or at least not as many, white schools, but 'just schools,' and neither Negroes nor whites would have suffered from the effects of segregated education, will all its shortcomings. Surely the Court's remedy will not restore to the Negro community, stigmatized as it was by the dual school system, what it would have enjoyed over all or most of this period if the remedy is confined to present-day Detroit; for the maximum remedy available within that area will leave many of the schools almost totally black, and the system itself will be predominantly black and will become increasingly so. Moreover, when a State has engaged in acts of official segregation over a lengthy

Page 780

period of time, as in the case before us, it is unrealistic to suppose that the children who were victims of the State's unconstitutional conduct could now be provided the benefits of which they were wrongfully deprived. Nor can the benefits which accrue to school systems in which schoolchildren have not been officially segregated, and to the communities supporting such school systems, be fully and immediately restored after a substantial period of unlawful segregation. The education of children of different races in a desegregated environment has unhappily been lost, along with the social, economic, and political advantages which accompany a desegregated school system as compared with an unconstitutionally segregated system. It is for these reasons that the Court has consistently followed the course of requiring the effects of past official segregation to be eliminated 'root and branch' by imposing, in the present, the duty to provide a remedy which will achieve 'the greatest possible degree of actual desegregation, taking into account the practicalities of the situation.' It is also for these reasons that once a constitutional violation has been found, the district judge obligated to provide such a remedy 'will thus necessarily be concerned with the elimination of one-race schools.' These concerns were properly taken into account by the District Judge in this case. Confining the remedy to the boundaries of the Detroit district is quite unrelated either to the goal of achieving maximum desegregation or to those intensely practical considerations, such as the extent and expense of transportation, that have imposed limits on remedies in cases such as this. The Court's remedy, in the end, is essentially arbitrary and will leave serious violations of the Constitution substantially unremedied.

          I agree with my Brother DOUGLAS that the Court of Appeals has acted responsibly in these cases. Regre-

Page 781

tably, the majority's arbitrary limitation on the equitable power of federal district courts, based on the invisible borders of local school districts, is unrelated to the State's responsibility for remedying the constitutional wrongs visited upon the Negro schoolchildren of Detroit. It is oblivious to the potential benefits of metropolitan relief, to the noneducational communities of interest among neighborhoods located in and sometimes bridging different school districts, and to the considerable interdistrict cooperation already existing in various educational areas. Ultimately, it is unresponsive to the goal of attaining the utmost actual desegregation consistent with restraints of practicability and thus augurs the frequent frustration of the remedial powers of the federal courts.

          Here the District Court will be forced to impose an intracity desegregation plan more expensive to the district, more burdensome for many of Detroit's Negro students, and surely more conductive to white flight than a metropolitan plan would be—all of this merely to avoid what the Detroit School Board, the District Court, and the en banc Court of Appeals considered to be the very manageable and quite surmountable difficulties that would be involved in extending the desegregation remedy to the suburban school districts.

          I am therefore constrained to record my disagreement and dissent.

           Mr. Justice MARSHALL, with whom Mr. Justice DOUGLAS, Mr. Justice BRENNAN, and Mr. Justice WHITE join, dissenting.

          In Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), this Court held that segregation of children in public schools on the basis of race deprives minority group childen of equal educational opportunities and therefore denies them the equal protection of the laws under the

Page 782

Fourteenth Amendment. This Court recognized then that remedying decades of segregation in public education would not be an easy task. Subsequent events, unfortunately, have seen that prediction bear bitter fruit. But however imbedded old ways, however ingrained old prejudices, this Court has not been diverted from its appointed task of making 'a living truth' of our constitutional ideal of equal justice under law. Cooper v. Aaron, 358 U.S. 1, 20, 78 S.Ct. 1401, 1410, 3 L.Ed.2d 5 (1958).

          After 20 years of small, often difficult steps toward that great end, the Court today takes a giant step backwards. Notwithstanding a record showing widespread and pervasive racial segregation in the educational system provided by the State of Michigan for children in Detroit, this Court holds that the District Court was powerless to require the State to remedy its constitutional violation in any meaningful fashion. Ironically purporting to base its result on the principle that the scope of the remedy in a desegregation case should be determined by the nature and the extent of the constitutional violation, the Court's answer is to provide no remedy at all for the violation proved in this case, thereby guaranteeing that Negro children in Detroit will receive the same separate and inherently unequal education in the future as they have been unconstitutionally afforded in the past.

          I cannot subscribe to this emasculation of our constitutional guarantee of equal protection of the laws and must respectfully dissent. Our precedents, in my view, firmly establish that where, as here, state-imposed segregation has been demonstrated, it becomes the duty of the State to eliminate root and branch all vestiges of racial discrimination and to achieve the greatest possible degree of actual desegregation. I agree with both the District Court and the Court of Appeals that, under the facts of this case, this duty cannot be fulfilled unless the State

Page 783

of Michigan involves outlying metropolitan area school districts in its desegregation remedy. Furthermore, I perceive no basis either in law or in the practicalities of the situation justifying the State's interposition of school district boundaries as absolute barriers to the implementation of an effective desegregation remedy. Under established and frequently used Michigan procedures, school district lines are both flexible and permeable for a wide variety of purposes, and there is no reason why they must now stand in the way of meaningful desegregation relief.

          The rights at issue in this case are too fundamental to be abridged on grounds as superficial as those relied on by the majority today. We deal here with the right of all of our children, whatever their race, to an equal start in life and to an equal opportunity to reach their full potential as citizens. Those children who have been denied that right in the past deserve better than to see fences thrown up to deny them that right in the future. Our Nation, I fear, will be ill served by the Court's refusal to remedy separate and unequal education, for unless our children begin to learn together, there is little hope that our people will ever learn to live together.

          The great irony of the Court's opinion and, in my view, its most serious analytical flaw may be gleaned from its concluding sentence, in which the Court remands for 'prompt formulation of a decree directed to eliminating the segregation found to exist in Detroit city schools, a remedy which has been delayed since 1970.' Ante, at 753. The majority, however, seems to have forgotten the District Court's explicit finding that a Detroit-only decree, the only remedy permitted under today's decision, 'would not accomplish desegregation.'

Page 784

          Nowhere in the Court's opinion does the majority confront, let alone respond to, the District Court's conclusion that a remedy limited to the city of Detroit would not effectively desegregate the Detroit city schools. I, for one, find the District Court's conclusion well supported by the record and its analysis compelled by our prior cases. Before turning to these questions, however, it is best to begin by laying to rest some mischaracterizations in the Court's opinion with respect to the basis for the District Court's decision to impose a metropolitan remedy.

          The Court maintains that while the initial focus of this lawsuit was the condition of segregation within the Detroit city schools, the District Court abruptly shifted focus in mid-course and altered its theory of the case. This new theory, in the majority's words, was 'equating racial imbalance with a constitutional violation calling for a remedy.' Ante, at 741, n. 19. As the following review of the District Court's handling of the case demonstrates, however, the majority's characterization is totally inaccurate. Nowhere did the District Court indicate that racial imbalance between school districts in the Detroit metropolitan area or within the Detroit School District constituted constitutional violation calling for interdistrict relief. The focus of this case was from the beginning, and has remained, the segregated system of education in the Detroit city schools and the steps necessary to cure that condition which offends the Fourteenth Amendment.

          The District Court's consideration of this case began with its finding, which the majority accepts, that the State of Michigan, through its instrumentality, the Detroit Board of Education, engaged in widespread purposeful acts of racial segregation in the Detroit School District. Without belaboring the details, it is sufficient to

Page 785

note that the various techniques used in Detroit were typical of methods employed to segregate students by race in areas where no statutory dual system of education has existed. See, e.g., Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973). Exacerbating the effects of extensive residential segregation between Negroes and whites, the school board consciously drew attendance zones along lines which maximized the segregation of the races in schools as well. Optional attendance zones were created for neighborhoods undergoing racial transition so as to allow whites in these areas to escape integration. Negro students in areas with overcrowded schools were transported past or away from closer white schools with available space to more distant Negro schools. Grade structures and feeder-school patterns were created and maintained in a manner which had the foreseeable and actual effect of keeping Negro and white pupils in separate schools. Schools were also constructed in locations and in sizes which ensured that they would open with predominantly one-race student bodies. In sum, the evidence adduced below showed that Negro children had been intentionally confined to an expanding core of virtually all-Negro schools immediately surrounded by a receding band of all-white schools.

          Contrary to the suggestions in the Court's opinion, the basis for affording a desegregation remedy in this case was not some perceived racial imbalance either between schools within a single school district or between independent school districts. What we confront here is 'a systematic program of segregation affecting a substantial portion of the students, schools . . . and facilities within the school system . . ..' Id., 413 U.S., at 201, 93 S.Ct., at 2694. The constitutional violation found here was not some de facto racial imbalance, but rather the purposeful, intentional, massive, de jure segregation of the Detroit city schools,

Page 786

which under our decision in Keyes, forms 'a predicate for a finding of the existence of a dual school system,' ibid., 93 S.Ct., at 2694, and justifies 'all-out desegregation.' Id., at 214, 93 S.Ct., at 2700.

          Having found a de jure segregated public school system in operation in the city of Detroit, the District Court turned next to consider which officials and agencies should be assigned the affirmative obligation to cure the constitutional violation. The court concluded that responsibility for the segregation in the Detroit city schools rested not only with the Detroit Board of Education, but belonged to the State of Michigan itself and the state defendants in this case that is, the Governor of Michigan, the Attorney General, the State Board of Education, and the State Superintendent of Public Instruction. While the validity of this conclusion will merit more extensive analysis below, suffice it for now to say that it was based on three considerations. First, the evidence at trial showed that the State itself had taken actions contributing to the segregation within the Detroit schools. Second, since the Detroit Board of Education was an agency of the State of Michigan, its acts of racial discrimination were acts of the State for purposes of the Fourteenth Amendment. Finally, the District Court found that under Michigan law and practice, the system of education was in fact a state school system, characterized by relatively little local control and a large degree of centralized state regulation, with respect to both educational policy and the structure and operation of school districts.

          Having concluded, then, that the school system in the city of Detroit was a de jure segregated system and that the State of Michigan had the affirmative duty to remedy that condition of segregation, the District Court then turned to the difficult task of devising an effective remedy. It bears repeating that the District Court's focus at this stage of the litigation remained what it had

Page 787

been at the beginning—the condition of segregation within the Detroit city schools. As the District Court stated: 'From the initial ruling (on segregation) to this day, the basis of the proceedings has been and remains the violation: de jure school segregation. . . . The task before this court, therefore, is now, and . . . has always been, how to desegregate the Detroit public schools.'

          The District Court first considered three desegregation plans limited to the geographical boundaries of the city of Detroit. All were rejected as ineffective to desegregate the Detroit city schools. Specifically, the District Court determined that the racial composition of the Detroit student body is such that implementation of any Detroit-only plan 'would clearly make the entire Detroit public school system racially identifiable as Black' and would 'leave many of its schools 75 to 90 per cent Black.' The District Court also found that a Detroit-only plan 'would change a school system which is now Black and White to one that would be perceived as Black, thereby increasing the flight of Whites from the city and the system, thereby increasing the Black student population.' Based on these findings, the District Court reasoned that 'relief of segregation in the public schools of the City of Detroit cannot be accomplished within the corporate geographical limits of the city' because a Detroit-only decree 'would accentuate the racial identifiability of the district as a Black school system, and would not accomplish desegregation.' The District Court therefore concluded that it 'must look beyond the limits of the Detroit school district for a solution to the problem of segregation in the Detroit public schools . . ..'

          In seeking to define the appropriate scope of that expanded desegregation area, however, the District Court continued to maintain as its sole focus the condition shown to violate the Constitution in this case—the segregation of the Detroit school system. As it stated, the

Page 788

primary question 'remains the determination of the area necessary and practicable effectively to eliminate 'root and branch' the effects of state-imposed and supported segregation and to desegregate the Detroit public schools.'

          There is simply no foundation in the record, then, for the majority's accusation that the only basis for the District Court's order was some desire to achieve a racial balance in the Detroit metropolitan area.¹ In fact, just the contrary is the case. In considering proposed desegregation areas, the District Court had occasion to criticize one of the State's proposals specifically because it had no basis other than its 'particular racial ratio' and did not focus on 'relevant factors, like eliminating racially identifiable schools (and) accomplishing maximum actual desegregation of the Detroit public schools.' Similarly, in rejecting the Detroit School Board's proposed desegregation area, even though it included more all-white districts and therefore achieved a higher white-Negro ratio, the District Court commented:

          'There is nothing in the record which suggests that these districts need be included in the desegregation area in order to disestablish the racial

Page 789

          identifiability of the Detroit public schools. From the evidence, the primary reason for the Detroit School Board's interest in the inclusion of these school districts is not racial desegregation but to increase the average socio-economic balance of all the schools in the abutting regions and clusters.'

          The Court also misstates the basis for the District Court's order by suggesting that since the only segregation proved at trial was within the Detroit school system, any relief which extended beyond the jurisdiction of the Detroit Board of Education would be inappropriate because it would impose a remedy on outlying districts 'not shown to have committed any constitutional violation.' Ante, at 745.² The essential foundation of interdistrict relief in this case was not to correct conditions within outlying districts which themselves engaged in purposeful segregation. Instead, interdistrict relief was seen as a necessary part of any meaningful effort by the State of Michigan to remedy the state-caused segregation within the city of Detroit.

          Rather than consider the propriety of interdistrict relief on this basis, however, the Court has conjured up a largely fictional account of what the District Court was attempting to accomplish. With all due respect, the Court, in my view, does a great disservice to the District Judge who labored long and hard with this complex litigation by accusing him of changing horses in midstream and shifting the focus of this case from the pursuit of a remedy for the condition of segregation

Page 790

within the Detroit school system to some unprincipled attempt to impose his own philosophy of racial balance on the entire Detroit metropolitan area. See ante, at 738—739. The focus of this case has always been the segregated system of education in the city of Detroit. The District Court determined that interdistrict relief was necessary and appropriate only because it found that the condition of segregation within the Detroit school system could not be cured with a Detroit-only remedy. It is on this theory that the interdistrict relief must stand or fall. Unlike the Court, I perceive my task to be to review the District Court's order for what it is, rather than to criticize it for what it manifestly is not.

          As the foregoing demonstrates, the District Court's decision to expand its desegregation decree beyond the geographical limits of the city of Detroit rested in large part on its conclusions (A) that the State of Michigan was ultimately responsible for curing the condition of segregation within the Detroit city schools, and (B) that a Detroit-only remedy would not accomplish this task. In my view, both of these conclusions are well supported by the facts of this case and by this Court's precedents.

          To begin with, the record amply supports the District Court's findings that the State of Michigan, through state officers and state agencies, had engaged in purposeful acts which created or aggravated segregation in the Detroit schools. The State Board of Education, for example, prior to 1962, exercised its authority to supervise local schoolsite selection in a manner which contributed to segregation. 484 F.2d 215, 238 (CA6 1973). Furthermore, the State's continuing authority, after 1962,

Page 791

to approve school building construction plans³ had intertwined the State with site-selection decisions of the Detroit Board of Education which had the purpose and effect of maintaining segregation.

          The State had also stood in the way of past efforts to desegregate the Detroit city schools. In 1970, for example, the Detroit School Board had begun implementation of its own desegregation plan for its high schools, despite considerable public and official resistance. The State Legislature intervened by enacting Act 48 of the Public Acts of 1970, specifically prohibiting implementation of the desegregation plan and thereby continuing the growing segregation of the Detroit school system. Adequate desegregation of the Detroit system was also hampered by discriminatory restrictions placed by the State on the use of transportation within Detroit. While state aid for transportation was provided by statute for suburban districts, many of which were highly urbanized, aid for intracity transportation was excepted. One of the effects of this restriction was to encourage the construction of small walk-in neighborhood schools in Detroit, thereby lending aid to the intentional policy of creating a school system which reflected, to the greatest extent feasible, extensive residential segregation. Indeed, that one of the purposes of the transportation restriction was to impede desegregation was evidenced when the Michigan Legislature amended the State Transportation Aid Act to cover intracity transportation but expressly prohibited the allocation of funds for cross-busing of students within a school district to achieve racial balance. ⁴ Cf. North Carolina State Board of Education v. Swann, 402 U.S. 43, 91 S.Ct. 1284, 28 L.Ed.2d 586 (1971).

Page 792

          Also significant was the State's involvement during the 1950's in the transportation of Negro high school students from the Carver School-District past a closer white high school in the Oak Park District to a more distant Negro high school in the Detroit system. Certainly the District Court's finding that the State Board of Education had knowledge of this action and had given its tacit or express approval was not clearly erroneous. Given the comprehensive statutory powers of the State Board of Education over contractual arrangements between school districts in the enrollment of students on a nonresident tuition basis, including certification of the number of pupils involved in the transfer and the amount of tuition charged, over the review of transportation routes and distances, and over the disbursement of transportation funds,⁵ the State Board inevitably knew and understood the significance of this discriminatory act.

          Aside from the acts of purposeful segregation committed by the State Legislature and the State Board of Education, the District Court also concluded that the State was responsible for the many intentional acts of segregation committed by the Detroit Board of Education, an agency of the State. The majority is only willing to accept this finding arguendo. See ante, at 748. I have no doubt, however, as to its validity under the Fourteenth Amendment.

          'The command of the Fourteenth Amendment,' it should be recalled, 'is that no 'State' shall deny to any person within its jurisdiction the equal protection of the laws.' Cooper v. Aaron, 358 U.S. 1, 16, 78 S.Ct. 1401, 1409, 3 L.Ed.2d 5 (1958). While a State can act only through 'the officers or agents by whom its powers are exerted,' Ex parte Virginia, 100 U.S. 339, 347, 25 L.Ed. 676 (1880), actions by an agent or officer of

Page 793

the State are encompassed by the Fourteenth Amendment for, 'as he acts in the name and for the State, and is clothed with the State's power, his act is that of the State.' Ibid. See also Cooper v. Aaron, supra; Virginia v. Rives, 100 U.S. 313, 318, 25 L.Ed. 667 (1880); Shelley v. Kraemer, 334 U.S. 1, 14, 68 S.Ct. 836, 842, 92 L.Ed. 1161 (1948). Under Michigan law a 'school district is an agency of the City of State government.' School District of Lansing v. State Board of Education, 367 Mich. 591, 600, 116 N.W.2d 866, 870 (1962). It is 'a legal division of territory, created by the State for educational purposes, to which the State has granted such powers as are deemed necessary to permit the district to function as a State agency.' Detroit Board of Education v. Superintendent of Public Instruction, 319 Mich. 436, 450, 29 N.W.2d 902, 908 (1947). Racial discrimination by the school district, an agency of the State, is therefore racial discrimination by the State itself, forbidden by the Fourteenth Amendment. See, e.g., Pennsylvania v. Board of Trusts, 353 U.S. 230, 77 S.Ct. 806, 1 L.Ed.2d 792 (1957).

          We recognized only last Term in Keyes that it was the State itself which was ultimately responsible for de jure acts of segregation committed by a local school board. A deliberate policy of segregation by the local board, we held, amounted to 'state-imposed segregation.' 413 U.S., at 200, 93 S.Ct., at 2693. Wherever a dual school system exists, whether compelled by state statute or created by a local board's systematic program of segregation, 'the State automatically assumes an affirmative duty 'to effectuate a transition to a racially nondiscriminatory school system' (and) to eliminate from the public schools within their school system 'all vestiges of state-imposed segregation." Ibid. (emphasis added).

          Vesting responsibility with the State of Michigan for Detroit's segregated schools is particularly appropriate as

Page 794

Michigan, unlike some other States, operates a single statewide system of education rather than several separate and independent local school systems. The majority's emphasis on local governmental control and local autonomy of school districts in Michigan will come as a surprise to those with any familiarity with that State's system of education. School districts are not separate and distinct sovereign entities under Michigan law, but rather are "auxiliaries of the State," subject to its 'absolute power.' Attorney General of Michigan ex rel. Kies v. Lowrey, 199 U.S. 233, 240, 26 S.Ct. 27, 29, 50 L.Ed. 167 (1905). The courts of the State have repeatedly emphasized that education in Michigan is not a local governmental concern, but a state function.

          'Unlike the delegation of other powers by the legislature to local governments, education is not inherently a part of the local self-government of a municipality . . .. Control of our public school system is a State matter delegated and lodged in the State legislature by the Constitution. The policy of the State has been to retain control of its school system, to be administered throughout the State under State laws by local State agencies organized with plenary powers to carry out the delegated functions given (them) by the legislature.' School District of the City of Lansing v. State Board of Education, supra, at 595, 116 N.W.2d, at 868.

          The Supreme Court of Michigan has noted the deep roots of this policy:

          'It has been settled by the Ordinance of 1787, the several Constitutions adopted in this state, by its uniform course of legislation, and by the decisions of this court, that education in Michigan is a matter of state concern, that it is no part of the local self-government of a particular township or munic-

Page 795

          ipality . . .. The legislature has always dictated the educational policy of the state.' In re School District No. 6, 284 Mich. 132, 145—146, 278 N.W. 792, 797 (1938).

          The State's control over education is reflected in the fact that, contrary to the Court's implication, there is little or no relationship between school districts and local political units. To take the 85 outlying local school districts in the Detroit metropolitan area as examples, 17 districts lie in two counties, two in three counties. One district serves five municipalities; other suburban municipalities are fragmented into as many as six school districts. Nor is there any apparent state policy with regard to the size of school districts, as they now range from 2,000 to 285,000 students.

          Centralized state control manifests itself in practice as well as in theory. The State controls the financing of education in several ways. The legislature contributes a substantial portion of most school districts' operating budgets with funds appropriated from the State's General Fund revenues raised through statewide taxation.⁶ The State's power over the purse can be and is in fact used to enforce the State's powers over local districts.⁷ In addition, although local districts obtain funds through local property taxation, the State has assumed the responsibility to ensure equalized property valuations throughout the State.⁸ The State also establishes

Page 796

standards for teacher certification and teacher tenure;⁹ determines part of the required curriculum;¹⁰ sets the minimum school term;¹¹ approves bus routes, equipment, and drivers;¹² approves textbooks;¹³ and establishes procedures for student discipline.¹⁴ The State Superintendent of Public Instruction and the State Board of Education have the power to remove local school board members from office for neglect of their duties.¹⁵

          Most significantly for present purposes, the State has wide-ranging powers to consolidate and merge school districts, even without the consent of the districts themselves or of the local citizenry.¹⁶ See, e.g., Attorney General ex rel. Kies, v. Lowrey, 131 Mich. 639, 92 N.W. 289 (1902), aff'd, 199 U.S. 233, 26 S.Ct. 27, 50 L.Ed. 167 (1905). Indeed, recent years have witnessed an accelerated program of school district consolidations, mergers, and annexations, many of which were state imposed. Whereas the State had 7,362 local districts in 1912, the number had been reduced to 1,438 in 1964 and to 738 in 1968.¹⁷ By June 1972, only 608 school districts remained. Furthermore, the State has broad powers to transfer property from one district to another, again without the consent of the local school districts affected by the transfer.¹⁸ See, e.g., School Dis-

Page 797

trict of the City of Lansing v. State Board of Education, supra; Imlay Township District v. State Board of Education, 359 Mich. 478, 102 N.W.2d 720 (1960).

          Whatever may be the history of public education in other parts of our Nation, it simply files in the face of reality to say, as does the majority, that in Michigan, '(n)o single tradition in public education is more deeply rooted than local control over the operation of schools . . ..' Ante, as 741. As the State's Supreme Court has said: 'We have repeatedly held that education in this state is not a matter of local concern, but belongs to the state at large.' Collins v. City of Detroit, 195 Mich. 330, 335—336, 161 N.W. 905, 907 (1917). See also Sturgis v. County of Allegan, 343 Mich. 209, 215, 72 N.W.2d 56, 59 (1955); Van Fleet v. Oltman, 244 Mich. 241, 244, 221 N.W. 299, 300 (1928); Child Welfare Society of Flint v. Kennedy School District, 220 Mich 290, 296, 189 N.W. 1002, 1004 (1922). Indeed, a study prepared for the 1961 Michigan Constitutional Convention noted that the Michigan Constitution's articles on education had resulted in 'the establishment of a state system of education in contrast to a series of local school systems.' Elemen-Michigan Constitution, Michigan Constitutional Convention Studies 1 (1961).

          In sum, several factors in this case coalesce to support the District Court's ruling that it was the State of Michigan itself, not simply the Detroit Board of Education, which bore the obligation of curing the condition of segregation within the Detroit city schools. The actions of the State itself directly contributed to Detroit's segregation. Under the Fourteenth Amendment, the State is ultimately responsible for the actions of its local agencies. And, finally, given the structure of Michigan's educational system, Detroit's segregation cannot be

Page 798

viewed as the problem of an independent and separate entity. Michigan operates a single statewide system of education, a substantial part of which was shown to be segregated in this case.

          What action, then, could the District Court require the State to take in order to cure Detroit's condition of segregation? Our prior cases have not minced words as to what steps responsible officials and agencies must take in order to remedy segregation in the public schools. Not only must distinctions on the basis of race be terminated for the future, but school officials are also 'clearly charged with the affirmative duty to take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch.' Green v. County School Board of New Kent County, 391 U.S. 430, 437—438, 88 S.Ct. 1689, 1694, 20 L.Ed.2d 716 (1968). See also Lee v. Macon County Board of Education, 267 F.Supp. 458 (MD Ala.), aff'd sub nom. Wallace v. United States, 389 U.S. 215, 88 S.Ct. 415, 19 L.Ed.2d 422 (1967). Negro students are not only entitled to neutral nondiscriminatory treatment in the future. They must receive 'what Brown II promised them: a school system in which all vestiges of enforced racial segregation have been eliminated.' Wright v. Council of the City of Emporia, 407 U.S. 451, 463, 92 S.Ct. 2196, 2203, 33 L.Ed.2d 51 (1972). See also Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 15, 91 S.Ct. 1267, 1275, 28 L.Ed.2d 554 (1971). These remedial standards are fully applicable not only to school districts where a dual system was compelled by statute, but also where, as here, a dual system was the product of purposeful and intentional state action. See Keyes, 413 U.S., at 200—201, 93 S.Ct., at 2693—2694.

          After examining three plans limited to the city of Detroit, the District Court correctly concluded that none would eliminate root and branch the vestiges of

Page 799

unconstitutional segregation. The plans' effectiveness, of course, had to be evaluated in the context of the District Court's findings as to the extent of segregation in the Detroit city schools. As indicated earlier, the most essential finding was that Negro children in Detroit had been confined by intentional acts of segregation to a growing core of Negro schools surrounded by a receding ring of white schools.¹⁹ Thus, in 1960, of Detroit's 251

Page 800

regular attendance schools, 100 were 90% or more white and 71 were 90% or more Negro. In 1970, of Detroit's 282 regular attendance schools, 69 were 90% or more white and 133 were 90% or more Negro. While in 1960, 68% of all schools were 90% or more one race, by 1970, 71.6% of the schools fell into that category. The growing core of all-Negro schools was further evidenced in total school district population figures. In 1960 the Detroit system had 46% Negro students and 54% white students, but by 1970, 64% of the students were Negro and only 36% were white. This increase in the proportion of Negro students was the highest of any major Northern city.

          It was with these figures in the background that the District Court evaluated the adequacy of the three Detroit-only plans submitted by the parties. Plan A, proposed by the Detroit Board of Education, desegregated the high schools and about a fifth of the middle-level schools. It was deemed inadequate, however, because it did not desegregate elementary schools and left the middle-level schools not included in the plan more segregated than ever. Plan C, also proposed by the Detroit Board, was deemed inadequate because it too covered only some grade levels and would leave elementary schools segregated. Plan B, the plaintiffs' plan, though requiring the transportation of 82,000 pupils and the acquisition of 900 school buses, would make little

Page 801

headway in rooting out the vestiges of segregation. To begin with, because of practical limitations, the District Court found that the plan would leave many of the Detroit city schools 75 to 90% Negro. More significantly, the District Court recognized that in the context of a community which historically had a school system marked by rigid de jure segregation, the likely effect of a Detroit-only plan would be to 'change a school system which is now Black and White to one that would be perceived as Black . . ..' The result of this changed perception, the District Court found, would be to increase the flight of whites from the city to the outlying suburbs, compounding the effects of the present rate of increase in the proportion of Negro students in the Detroit system. Thus, even if a plan were adopted which, at its outset, provided in every school a 65% Negro-35% white racial mix in keeping with the Negro-white proportions of the total student population, such a system would, in short order, devolve into an all-Negro system. The net result would be a continuation of the all-Negro schools which were the hallmarks of Detroit's former dual system of one-race schools.

          Under our decisions, it was clearly proper for the District Court to take into account the so-called 'white flight' from the city schools which would be forthcoming from any Detroit-only decree. The court's prediction of white flight was well supported by expert testimony based on past experience in other cities undergoing desegregation relief. We ourselves took the possibility of white flight into account in evaluating the effectiveness of a desegregation plan in Wright, supra, where we relied on the District Court's finding that if the city of Emporia were allowed to withdraw from the existing system, leaving a system with a higher proportion of Negroes, it 'may be anticipated that the pro-

Page 802

portion of whites in county schools may drop as those who can register in private academies' . . ..' 407 U.S., at 464, 92 S.Ct., at 2204. One cannot ignore the white-flight problem, for where legally imposed segregation has been established, the District Court has the responsibility to see to it not only that the dual system is terminated at once but also that future events do not serve to perpetuate or re-establish segregation. See Swann, 402 U.S. at 21, 91 S.Ct., at 1278. See also Green, 391 U.S., at 438 n. 4, 88 S.Ct., at 1694; Monroe v. Board of Comm'rs, 391 U.S. 450, 459, 88 S.Ct. 1700, 1705, 20 L.Ed.2d 733 (1968).

          We held in Swann, supra, that where de jure segregation is shown, school authorities must make 'every effort to achieve the greatest possible degree of actual desegregation.' 402 U.S., at 26, 91 S.Ct., at 1281. This is the operative standard re-emphasized in Davis v. School Comm'rs of Mobile County, 402 U.S. 33, 37, 91 S.Ct. 1289, 1292, 28 L.Ed.2d 577 (1971). If these words have any meaning at all, surely it is that school authorities must, to the extent possible, take all practicable steps to ensure that Negro and white children in fact go to school together. This is, in the final analysis, what desegregation of the public schools is all about.

          Because of the already high and rapidly increasing percentage of Negro students in the Detroit system, as well as the prospect of white flight, a Detroit-only plan simply has no hope of achieving actual desegregation. Under such a plan white and Negro students will not go to school together. Instead, Negro children will continue to attend all-Negro schools. The very evil that Brown I was aimed at will not be cured, but will be perpetuated for the future.

          Racially identifiable schools are one of the primary vestiges of state-imposed segregation which an effective desegregation decree must attempt to eliminate. In Swann, supra, for example, we held that '(t)he district judge or school authorities . . . will thus necessarily be concerned with the elimination of one-race schools.' 402

Page 803

U.S., at 26, 91 S.Ct., at 1281. There is 'a presumption,' we stated, 'against schools that are substantially disproportionate in their racial composition.' Ibid. And in evaluating the effectiveness of desegregation plans in prior cases, we ourselves have considered the extent to which they discontinued racially identifiable schools. See, e.g., Green v. County School Board of New Kent County, supra; Wright v. Council of the City of Emporia, supra. For a principal end of any desegregation remedy is to ensure that it is no longer 'possible to identify a 'white school' or a 'Negro school." Swann, supra, 402 U.S., 18, 91 S.Ct., at 1277. The evil to be remedied in the dismantling of a dual system is the '(r) acial identification of the system's schools.' Green, supra, 391 U.S., at 435, 88 S.Ct., at 1693. The goal is a system without white schools or Negro schools—a system with 'just schools.' Id., at 442, 88 S.Ct., at 1696. A school authority's remedial plan or a district court's remedial decree is to be judged by its effectiveness in achieving this end. See Swann, 402 U.S., at 25, 91 S.Ct., at 1280; Davis, supra, 402 U.S., at 37, 91 S.Ct., at 1292; Green, supra, 391 U.S., at 439, 88 S.Ct., at 1694.

          We cautioned in Swann, of course, that the dismantling of a segregated school system does not mandate any particular racial balance. 402 U.S., at 24, 91 S.Ct., at 1280. We also concluded that a remedy under which there would remain a small number of racially identifiable schools was only presumptively inadequate and might be justified. Id., at 26, 91 S.Ct., at 1281. But this is a totally different case. The flaw of a Detroit-only decree is not that it does not reach some ideal degree of racial balance or mixing. It simply does not promise to achieve actual desegregation at all. It is one thing to have a system where a small number of students remain in racially identifiable schools. It is something else entirely to have a system where all students continue to attend such schools.

          The continued racial identifiability of the Detroit schools under a Detroit-only remedy is not simply a reflection of their high percentage of Negro students.

Page 804

What is or is not a racially identifiable vestige of de jure segregation must necessarily depend on several factors. Cf. Keyes, 413 U.S., at 196, 93 S.Ct., at 2691. Foremost among these should be the relationship between the schools in question and the neighboring community. For these purposes the city of Detroit and its surrounding suburbs must be viewed as a single community. Detroit is closely connected to its suburbs in many ways, and the metropolitan area is viewed as a single cohesive unit by its residents. About 40% of the residents of the two suburban counties included in the desegregation plan work in Wayne County, in which Detroit is situated. Many residents of the city work in the suburbs. The three counties participate in a wide variety of cooperative governmental ventures on a metropolitan-wide basis, including a metropolitan transit system, park authority, water and sewer system, and council of governments. The Federal Government has classified the tri-county area as a Standard Metropolitan Statistical Area, indicating that it is an area of 'economic and social integration.' United States v. Connecticut National Bank, 418 U.S. 656, 670, 94 S.Ct. 2788, 2797, 41 L.Ed.2d 1016 (1974).

          Under a Detroit-only decree, Detroit's schools will clearly remain racially identifiable in comparison with neighboring schools in the metropolitan community. Schools with 65% and more Negro students will stand in sharp and obvious contrast to schools in neighboring districts with less than 2% Negro enrollment. Negro students will continue to perceive their schools as segregated educational facilities and this perception will only be increased when whites react to a Detroit-only decree by fleeing to the suburbs to avoid integration. School district lines, however innocently drawn, will surely be perceived as fences to separate the races when, under a Detroit-only decree, white parents withdraw their chil-

Page 805

dren from the Detroit city schools and move to the suburbs in order to continue them in all-white schools. The message of this action will not escape the Negro children in the city of Detroit. See Wright, 407 U.S., at 466, 92 S.Ct., at 2205. It will be of scant significance to Negro children who have for years been confined by de jure acts of segregation to a growing core of all-Negro schools surrounded by a ring of all-white schools that the new dividing line between the races is the school district boundary.

          Nor can it be said that the State is free from any responsibility for the disparity between the racial makeup of Detroit and its surrounding suburbs. The State's creation, through de jure acts of segregation, of a growing core of all-Negro schools inevitably acted as a magnet to attract Negroes to the areas served by such schools and to deter them from settling either in other areas of the city or in the suburbs. By the same token, the growing core of all-Negro schools inevitably helped drive whites to other areas of the city or to the suburbs. As we recognized in Swann:

          'People gravitate toward school facilities, just as schools are located in response to the needs of people. The location of schools may thus influence the patterns of residential development of a metropolitan area and have important impact on composition of inner-city neighborhoods. . . . (Action taken) to maintain the separation of the races with a minimum departure from the formal principles of 'neighborhood zoning' . . . does more than simply influence the short-run composition of the student body . . .. It may well promote segregated residential patterns which, when combined with 'neighborhood zoning,' further lock the school system into the mold of separation of the races. Upon a proper

Page 806

          showing a district court may consider this in fashioning a remedy.' 402 U.S., at 20—21, 91 S.Ct., at 1278.

          See also Keyes, 413 U.S., at 202, 93 S.Ct., at 2694. The rippling effects on residential patterns caused by purposeful acts of segregation do not automatically subside at the school district border. With rare exceptions, these effects naturally spread through all the residential neighborhoods within a metropolitan area. See id., at 202—203, 93 S.Ct., at 2694—2695.

          The State must also bear part of the blame for the white flight to the suburbs which would be forthcoming from a Detroit-only decree and would render such a remedy ineffective. Having created a system where whites and Negroes were intentionally kept apart so that they could not become accustomed to learning together, the State is responsible for the fact that many whites will react to the dismantling of that segregated system by attempting to flee to the suburbs. Indeed, by limiting the District Court to a Detroit-only remedy and allowing that flight to the suburbs to succeed, the Court today allows the State to profit from its own wrong and to perpetuate for years to come the separation of the races it achieved in the past by purposeful state action.

          The majority asserts, however, that involvement of outlying districts would do violence to the accepted principle that 'the nature of the violation determines the scope of the remedy.' Swann, supra, 402 U.S., at 16, 91 S.Ct., at 1276. See ante, at 744 745. Not only is the majority's attempt to find in this single phrase the answer to the complex and difficult questions presented in this case hopelessly simplistic, but more important, the Court reads these words in a manner which perverts their obvious meaning. The nature of a violation determines the scope of the remedy simply because the function of any remedy is to cure the violation to which it is addressed. In school segregation

Page 807

cases, as in other equitable causes, a remedy which effectively cures the violation is what is required. See Green, 391 U.S., at 439, 88 S.Ct., at 1694; Davis, 402 U.S., at 37, 91 S.Ct., at 1292. No more is necessary, but we can tolerate no less. To read this principle as barring a district court from imposing the only effective remedy for past segregation and remitting the court to a patently ineffective alternative is, in my view, to turn a simple commonsense rule into a cruel and meaningless paradox. Ironically, by ruling out an interdistrict remedy, the only relief which promises to cure segregation in the Detroit public schools, the majority flouts the very principle on which it purports to rely.

          Nor should it be of any significance that the suburban school districts were not shown to have themselves taken any direct action to promote segregation of the races. Given the State's broad powers over local school districts, it was well within the State's powers to require those districts surrounding the Detroit school district to participate in a metropolitan remedy. The State's duty should be no different here than in cases where it is shown that certain of a State's voting districts are malapportioned in violation of the Fourteenth Amendment. See Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Overrepresented electoral districts are required to participate in reapportionment although their only 'participation' in the violation was to do nothing about it. Similarly, electoral districts which themselves meet representation standards must frequently be redrawn as part of a remedy for other over-and under-inclusive districts. No finding of fault on the part of each electoral district and no finding of a discriminatory effect on each district is a prerequisite to its involvement in the constitutionally required remedy. By the same logic, no finding of fault on the part of the suburban school districts in this case

Page 808

and no finding of a discriminatory effect on each district should be a prerequisite to their involvement in the constitutionally required remedy.

          It is the State, after all, which bears the responsibility under Brown of affording a nondiscriminatory system of education. The State, of course, is ordinarily free to choose any decentralized framework for education it wishes, so long as it fulfills that Fourteenth Amendment obligation. But the State should no more be allowed to hide behind its delegation and compartmentalization of school districts to avoid its constitutional obligations to its children than it could hide behind its political subdivisions to avoid its obligations to its voters. Reynolds v. Sims, at 575, 84 S.Ct., at 1388. See also Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960).

          It is a hollow remedy indeed where 'after supposed 'desegregation' the schools remained segregated in fact.' Hobson v. Hansen, 269 F.Supp. 401, 495 (D.D.C. 1967). We must do better than "substitute . . . one segregated school system for another segregated school system." Wright, 407 U.S., at 456, 92 S.Ct., at 2200. To suggest, as does the majority, that a Detroitonly plan somehow remedies the effects of de jure segregation of the races is, in my view, to make a solemn mockery of Brown I's holding that separate educational facilities are inherently unequal and of Swann's unequivocal mandate that the answer to de jure segregation is the greatest possible degree of actual desegregation.

          One final set of problems remains to be considered. We recognized in Brown II, and have re-emphasized ever since, that in fashioning relief in desegregation cases, 'the courts will be guided by equitable principles. Traditionally, equity has been characterized by a practical flexibility in shaping its remedies and by a facility for

Page 809

adjusting and reconciling public and private needs.' Brown II, 349 U.S., at 300, 75 S.Ct., at 756. See also Swann, supra.

          Though not resting its holding on this point, the majority suggests that various equitable considerations militate against interdistrict relief. The Court, for example, refers to financing and administrative problems, the logistical problems attending large-scale transportation of students, and the prospect of the District Court's becoming a 'de facto 'legislative authority" and "school superintendent' for the entire area.' Ante, at 743—744. The entangling web of problems woven by the Court, however, appears on further consideration to be constructed of the flimsiest of threads.

          I deal first with the last of the problems posed by the Court the specter of the District Court qua 'school superintendent' and 'legislative authority'—for analysis of this problem helps put the other issues in proper perspective. Our cases, of course, make clear that the initial responsibility for devising an adequate desegregation plan belongs with school authorities, not with the District Court. The court's primary role is to review the adequacy of the school authorities' efforts and to substitute its own plan only if and to the extent they default. See Swann, 402 U.S., at 16, 91 S.Ct., at 1276; Green, 391 U.S., at 439, 88 S.Ct., at 1294. Contrary to the majority's suggestions, the District Judge in this case consistently adhered to these procedures and there is every indication that he would have continued to do so. After finding de jure segregation the court ordered the parties to submit proposed Detroit-only plans. The state defendants were also ordered to submit a proposed metropolitan plan extending beyond Detroit's boundaries. As the District Court stated, 'the State defendants . . . bear the initial burden of coming forward with a proposal that promises to work.' The state defendants defaulted in this obligation, however.

Page 810

Rather than submit a complete plan, the State Board of Education submitted six proposals, none of which was in fact a desegregation plan. It was only upon this default that the District Court began to take steps to develop its own plan. Even then the District Court maximized school authority participation by appointing a panel representing both plaintiffs and defendants to develop a plan. Pet.App. 99a—100a. Furthermore, the District Court still left the state defendants the initial responsibility for developing both interim and final financial and administrative arrangements to implement interdistrict relief. Id., at 104A—105a. The Court of Appeals further protected the interests of local school authorities by ensuring that the outlying suburban districts could fully participate in the proceedings to develop a metropolitan remedy.

          These processes have not been allowed to run their course. No final desegregation plan has been proposed by the panel of experts, let alone approved by the District Court. We do not know in any detail how many students will be transported to effect a metropolitan remedy, and we do not know how long or how far they will have to travel. No recommendations have yet been submitted by the state defendants on financial and administrative arrangements. In sum, the practicality of a final metropolitan plan is simply not before us at the present time. Since the State and the panel of experts have not yet had an opportunity a come up with a workable remedy, there is no foundation for the majority's suggestion of the impracticality of interdistrict relief. Furthermore, there is no basis whatever for assuming that the District Court will inevitably be forced to assume the role of legislature or school superintendent.²⁰

Page 811

Were we to hold that it was its constitutional duty to do so, there is every indication that the State of Michigan would fulfill its obligation and develop a plan which is workable, administrable, financially sound, and, most important, in the best interest of quality education for all of the children in the Detroit metropolitan area.

          Since the Court chooses, however, to speculate on the feasibility of a metropolitan plan, I feel constrained to comment on the problem areas it has targeted. To begin with, the majority's question concerning the practicality of consolidation of school districts need not give us pause. The State clearly has the power, under existing law, to effect a consolidation if it is ultimately determined that this offers the best prospect for a workable and stable desegregation plan. See supra, at 796—797. And given the 1,000 or so consolidations of school districts which have taken place in the past, it is hard to believe that the State has not already devised means of solving most, if not all, of the practical problems which the Court suggests consolidation would entail.

          Furthermore, the majority ignores long-established Michigan procedures under which school districts may enter into contractual agreements to educate their pupils in other districts using state or local funds to finance nonresident education.²¹ Such agreements could form an

Page 812

easily administrable framework for interdistrict relief short of outright consolidation of the school districts. The District Court found that interdistrict procedures like these were frequently used to provide special educational services for handicapped children, and extensive statutory provision is also made for their use in vocational education.²² Surely if school districts are willing to engage in interdistrict programs to help those unfortunate children crippled by physical or mental handicaps, school districts can be required to participate in an inter-district program to help those children in the city of Detroit whose educations and very futures have been crippled by purposeful state segregation.

          Although the majority gives this last matter only fleeting reference, it is plain that one of the basic emotional and legal issues underlying these cases concerns the propriety of transportation of students to achieve desegregation. While others may have retreated from its standards, see, e.g., Keyes, 413 U.S., at 217, 93 S.Ct., at 2701 (Powell, J., concurring in part and dissenting in part), I continue to adhere to the guidelines set forth in Swann on this issue. See 402 U.S., at 29—31, 91 S.Ct., at 1282—1283. And though no final desegregation plan is presently before us, to the extent the outline of such a plan is now visible, it is clear that the transportation it would entail will be fully consistent with these guidelines.

          First of all, the metropolitan plan would not involve the busing of substantially more students than already ridebuses. The District Court found that, statewide, 35%—40% of all students already arrive at school on a bus. In those school districts in the tri-county Detroit metropolitan area eligible for state reimbursement of transportation costs, 42%—52% of all students rode buses to school. In the tri-county areas as a whole, ap-

Page 813

proximately 300,000 pupils arrived at school on some type of bus, with about 60,000 of these apparently using regular public transit. In comparison, the desegregation plan, according to its present rough outline, would involve the transportation of 310,000 students, about 40% of the population within the desegregation area.

          With respect to distance and amount of time traveled, 17 of the outlying school districts involved in the plan are contiguous to the Detroit district. The rest are all within 8 miles of the Detroit city limits. The trial court, in defining the desegregation area, placed a ceiling of 40 minutes one way on the amount of travel time, and many students will obviously travel for far shorter periods. As to distance, the average statewide bus trip is 8 1/2 miles one way, and in some parts of the tri-county area, students already travel for one and a quarter hours or more each way. In sum, with regard to both the number of students transported and the time and distances involved, the outlined desegregation plan 'compares favorably with the transportation plan previously operated . . ..' Swann, 402 U.S., at 30, 91 S.Ct., at 1283.

          As far as economics are concerned, a metropolitan remedy would actually be more sensible than a Detroit-only remedy. Because of prior transportation aid restrictions, see at 791, Detroit largely relied on public transport, at student expense, for those students who lived too far away to walk to school. Since no inventory of school buses existed, a Detroit-only plan was estimated to require the purchase of 900 buses to effectuate the necessary transportation. The tri-county area, in contrast, already has an inventory of 1,800 buses, many of which are now under-utilized. Since increased utilization of the existing inventory can take up much of the increase in transportation involved in the interdistrict remedy, the District Court found that only 350 additional buses would

Page 814

probably be needed, almost two-thirds fewer than a Detroit-only remedy. Other features of an interdistrict remedy bespeak its practicality, such as the possibility of pairing up Negro schools near Detroit's boundary with nearby white schools on the other side of the present school district line.

          Some disruption, of course, is the inevitable product of any desegregation decree, whether it operates within one district or on an interdistrict basis. As we said in Swann, however:

          'Absent a constitutional violation there would be no basis for judicially ordering assignment of students on a racial basis. All things being equal, with no history of discrimination, it might well be desirable to assign pupils to schools nearest their homes. But all things are not equal in a system that has been deliberately constructed and maintained to enforce racial segregation. The remedy for such segregation may be administratively awkward, inconvenient, and even bizarre in some situations and may impose burdens on some; but all awkwardness and inconvenience cannot be avoided . . ..' 402 U.S., at 28, 91 S.Ct., at 1282.

          Desegregation is not and was never expected to be an easy task. Racial attitudes ingrained in our Nation's childhood and adolescence are not quickly thrown aside in its middle years. But just as the inconvenience of some cannot be allowed to stand in the way of the rights of others, so public opposition, no matter how strident, cannot be permitted to divert this Court from the enforcement of the constitutional principles at issue in this case. Today's holding, I fear, is more a reflection of a perceived public mood that we have gone far enough in enforcing the Constitution's guarantee of equal justice than it is the product of neutral principles of law. In

Page 815

the short run, it may seem to be the easier course to allow our great metropolitan areas to be divided up each into two cities—one white, the other black—but it is a course, I predict, our people will ultimately regret. I dissent.

^{1. 484 F.2d 215 (CA6), cert. granted, 414 U.S. 1038, 94 S.Ct. 538, 38 L.Ed.2d 329 (1973).}

^2. The standing of the NAACP as a proper party plaintiff was not contested in the trial court and is not an issue in this case.

^3. Optional zones, sometimes referred to as dual zones or dual overlapping zones, provide pupils living within certain areas a choice of attendance at one of two high schools.

^4. The Court of Appeals found record evidence that in at least one instance during the period 1957—1958, Detroit served a suburban school district by contracting with it to educate its Negro high school students by transporting them away from nearby suburban white high schools, and past Detroit high schools which were predominantly white, to all-Negro or predominantly Negro Detroit schools. 484 F.2d, at 231.

^5. School districts in the State of Michigan are instrumentalities of the State and subordinate to its State Board of Education and legislature. The Constitution of the State of Michigan, Art. 8, § 2, provides in relevant part:

'The legislature shall maintain and support a system of free public elementary and secondary schools as defined by law.'

Similarly, the Michigan Supreme Court has stated: 'The school district is a State agency. Moreover, it is of legislative creation. . . .' Attorney General ex rel. Kies v. Lowrey, 131 Mich. 639, 644, 92 N.W. 289, 290 (1902): "Education in Michigan belongs to the State. It is no part of the local self-government inherent in the township or municipality, except so far as the Legislature may choose to make it such. The Constitution has turned the whole subject over to the Legislature. . . ." Attorney General ex rel. Zacharias v. Detroit Board of Education, 154 Mich. 584, 590, 118 N.W. 606, 609 (1908).

^6. 'Sec. 12. The implementation of any attendance provisions for the 1970—71 school year determined by any first class school district board shall be delayed pending the date of commencement of functions by the first class school district boards established under the provisions of this amendatory act but such provision shall not impair the right of any such board to determine and implement prior to such date such changes in attendance provisions as are mandated by practical necessity. . . .' Act No. 48, § 12, Mich.Pub. Acts of 1970; Mich.Comp.Laws § 388.182 (1970).

^7. The District Court briefly alluded to the possibility that the State, along with private persons, had caused, in part, the housing patterns of the Detroit metropolitan area which, in turn, produced the predominantly white and predominantly Negro neighborhoods that characterize Detroit:

'It is no answer to say that restricted practices grew gradually (as the black population in the area increased between 1920 and 1970), or that since 1948 racial restrictions on the ownership of real property have been removed. The policies pursued by both government and private persons and agencies have a continuing and present effect upon the complexion of the community as we know, the choice of a residence is a relatively infrequent affair. For many years FHA and VA openly advised and advocated the maintenance of 'harmonious' neighborhoods, i.e., racially and economically harmonious. The conditions created continue.' 338 F.Supp. 582, 587 (ED Mich.1971).

Thus, the District Court concluded:

'The affirmative obligation of the defendant Board has been and is to adopt and implement pupil assignment practices and policies that compensate for and avoid incorporation into the school system the effects of residential racial segregation.' Id., at 593.

The Court of Appeals, however, expressly noted that:

'In affirming the District Judge's findings of constitutional violations by the Detroit Board of Education and by the State defendants resulting in segregated schools in Detroit, we have not relied at all upon testimony pertaining to segregated housing except as school construction programs helped cause or maintain such segregation.' 484 F.2d., at 242.

Accordingly, in its present posture, the case does not present any question concerning possible state housing violations.

^8. On March 22, 1971, a group of Detroit residents, who were parents of children enrolled in the Detroit public schools, were permitted to intervene as parties defendant. On June 24, 1971, the District Judge alluded to the 'possibility' of a metropolitan school system stating: '(A)s I have said to several witnesses in this case: 'How do you desegrate a black city, or a black school system." Petitioners' Appendix 243a (hereinafter Pat.App.). Subsequently, on July 16, 1971, various parents filed a motion to require joinder of all of the 85 outlying independent school districts within the tri-county area.

^9. The respondents, as plaintiffs below, opposed the motion to join the additional school districts, arguing that the presence of the state defendants was sufficient and all that was required, even if, in shaping a remedy, the affairs of these other districts was to be affected. 338 F.Supp. at 595.

^10. At the time of the 1970 census, the population of Michigan was 8,875,083, almost half of which, 4,199,931, resided in the tri-county area of Wayne, Oakland, and Macomb. Oakland and Macomb Counties abut Wayne County to the north, and Oakland County abuts Macomb County to the west. These counties cover 1,952 square miles, Michigan Statistical Abstract (9th ed. 1972), and the area is approximately the size of the State of Delaware (2,057 square miles), more than half again the size of the State of Rhode Island (1,214 square miles) and almost 30 times the size of the District of Columbia (67 square miles). Statistical Abstract of the United States (93d ed. 1972). The populations of Wayne, Oakland, and Macomb Counties were 2,666,751; 907,871; and 625,309, respectively, in 1970. Detroit, the State's largest city, is located in Wayne County.

In the 1970—1971 school year, there were 2,157,449 children enrolled in school districts in Michigan. There are 86 independent, legally distinct school districts within the tri-county area, having a total enrollment of approximately 1,000,000 children. In 1970, the Detroit Board of Education operated 319 schools with approximately 276,000 students.

^11. In its formal opinion, subsequently announced, the District Court candidly recognized:

'It should be noted that the court has taken no proofs with respect to the establishment of the boundaries of the 86 public school districts in the counties of Wayne, Oakland and Macomb, nor on the issue of whether, with the exclusion of the city of Detroit school district, such school districts have committed acts of de jure segregation.' 345 F.Supp. 914, 920 (ED Mich.1972).

^12. According to the District Court, intervention was permitted under Fed.Rule Civ.Proc. 24(a), 'Intervention of Right,' and also under Rule 24(b), 'Permissive Intervention.'

^13. This rather abbreviated briefing schedule was maintained despite the fact that the District Court had deferred consideration of a motion made eight months earlier, to bring the suburban districts into the case. See text accompanying n. 8 supra.

^14. As of 1970, the 53 school districts outside the city of Detroit that were included in the court's 'desegration area' had a combined student population of approximately 503,000 students compared to Detroit's approximately 276,000 students. Nevertheless, the District Court directed that the intervening districts should be represented by only one member on the desegregation panel while the Detroit Board of Education was granted three panel members. 345 F.Supp., at 917.

^15. The District Court had certified most of the foregoing rulings for interlocutory review pursuant to 28 U.S.C. § 1292(b) (1 App. 265—266) and the case was initially decided on the merits by a panel of three judges. However, the panel's opinion and judgment were vacated when it was determined to rehear the case en banc, 484 F.2d, at 218.

^16. With respect to the State's violations, the Court of Appeals held: (1) that, since the city Board is an instrumentality of the State and subordinate to the State Board, the segregative actions of the Detroit Board 'are the actions of an agency of the State,' id., at 238; (2) that the state legislation rescinding Detroit's voluntary desegregation plan contributed to increasing segregation in the Detroit schools, ibid.; (3) that under state law prior to 1962 the State Board had authority over school construction plans and therefore had to be held responsible 'for the segregative results,' ibid.; (4) that the 'State statutory scheme of support of transportation for school children directly discriminated against Detroit;' id., at 240, by not providing transportation funds to Detroit on the same basis as funds were provided to surburban districts, id., at 238; and (5) that the transportation of Negro students from one suburban district to a Negro school in Detroit must have had the 'approval, tacit or express, of the State Board of Education,' ibid.

^17. The court sought to distinguish Bradley v. School Board of the City of Richmond, 462 F.2d 1058 (CA4 1972), aff'd by an equally divided Court, 412 U.S. 92, 93 S.Ct. 1952, 36 L.Ed.2d 771 (1973), on the grounds that the District Court in that case had ordered an actual consolidation of three school districts and that Virginia's Constitution and statutes, unlike Michigan's gave the local boards exclusive power to operate the public schools. 484 F.2d, at 251.

^18. Although the list of issues presented for review in petitioners' briefs and petitions for writs of certiorari do not include arguments on the findings of segregative violations on the part of the Detroit defendants, two of the petitioners argue in brief that these findings constitute error. This Court's Rules 23(1)(c) and 40(1)(d)(2), at a minimum limit our review to the Detroit violation findings to 'plain error,' and, under our decision last Term in Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 1973), the findings appear to be correct.

^19. Disparity in the racial composition of pupils within a single district may well constitute a 'signal' to a district court at the outset, leading to inquiry into the causes accounting for a pronounced racial identifiability of schools within one school system. In Swann, for example, we were dealing with a large but single independent school system, . . . and a unanimous Court noted: 'Where the school authority's proposed plan for conversion from a dual to a unitary system contemplates the continued existence of some schools that are all or predominantly of one race (the school authority has) the burden of showing that such school assignments are genuinely nondiscriminatory.' 402 U.S., at 26, 91 S.Ct., at 1281. See also Keyes, supra, at 208, 93 S.Ct., at 2697. However, the use of significant racial imbalance in schools within an autonomous school district as a signal which operates simply to shift the burden of proof, is a very different matter from equating racial imbalance with a constitutional violation calling for a remedy. Keyes, supra, also involved a remedial order within a single autonomous school district.

^20. Under the Michigan School Code of 1955, the local school district is an autonomous political body corporate, operating through a Board of Education popularly elected. Mich.Comp.Laws §§ 340.27, 340.55, 340.107, 340.148, 340.149, 340.188. As such, the day-to-day affairs of the school district are determined at the local level in accordance with the plenary power to acquire real and personal property, §§ 340.26, 340.77, 340.113, 340.165, 340.192, 340.352; to hire and contract with personnel, §§ 340.569, 340.574; to levy taxes for operations, § 340.563; to borrow against receipts, § 340.567; to determine the length of school terms, § 340.575; to control the admission of nonresident students, § 340.582; to determine courses of study, § 340.583; to provide a kindergarten program, § 340.584; to establish and operate vocational schools, § 340.585; to offer adult education programs, § 340.586; to establish attendance areas, § 340.589; to arrange for transportation of nonresident students, § 340.591; to acquire transportation equipment, § 340.594; to receive gifts and bequests for educational purposes, § 340.605; to employ an attorney, § 340.609; to suspend or expel students, § 340.613; to make rules and regulations for the operation of schools, § 340.614; to cause to be levied authorized millage, § 340.643a; to acquire property by eminent domain, § 340.711 et seq.; and to approve and select textbooks, § 340.882.

^21. Since the Court has held that a resident of a school district has a fundamental right protected by the Federal Constitution to vote in a district election, it would seem incongruous to disparage the importance of the school district in a different context. Kramer v. Union Free School District No. 15, 395 U.S. 621, 626, 89 S.Ct. 1886, 1889, 23 L.Ed.2d 583 (1969). White the district there involved was located in New York, none of the facts in our possession suggest that the relation of school districts to the State is significantly different in New York from that in Michigan.

^22. The suggestion in the dissent of Mr. Justice MARSHALL that schools which have a majority of Negro students are not 'desegregated,' whatever the racial makeup of the school district's population and however neutrally the district lines have been drawn and administered, finds no support in our prior cases. In Green v. County School Board of New Kent County, 391 U.S. 430, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968), for example, this Court approved a desegregation plan which would have resulted in each of the schools within the district having a racial composition of 57% Negro and 43% White. In Wright v. Council of the City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972), the optimal desegregation plan would have resulted in the schools' being 66% Negro and 34% white, substantially the same percentages as could be obtained under one of the plans involved in this case. And in United States v. Scotland Neck City Board of Education, 407 U.S. 484, 491 n. 5, 92 S.Ct. 2214, 2218, 33 L.Ed.2d 75 (1972), a desegregation plan was implicitly approved for a school district which had a racial composition of 77% Negro and 22% white. In none of these cases was it even intimated that 'actual desegregation' could not be accomplished as long as the number of Negro students was greater than the number of white students.

The dissents also seem to attach importance to the metropolitan character of Detroit and neighboring school districts. But the constitutional principles applicable in school desegregation cases cannot vary in accordance with the size or population dispersal of the particular city, county, or school district as compared with neighboring areas.

^23. People ex rel. Workman v. Board of Education of Detroit, 18 Mich. 400 (1869); Act 34, § 28, Mich.Pub.Acts of 1867. The Michigan Constitution and laws provide that 'every school district shall provide for the education of its pupils without discrimination as to religion, creed, race, color or national origin,' Mich.Const.1963, Art. 8, § 2; that 'no separate school or department shall be kept for any person or persons on account of race or color,' Mich.Comp.Laws § 340.355; and that '(a)ll persons, residents of a school district . . . shall have an equal right to attend school therein,' id., § 340.356. See also Act 319, Part II, c. 2, § 9, Mich.Pub.Acts of 1927.

^24. Apparently, when the District Court sua sponte, abruptly altered the theory of the case to include the possibility of multidistrict relief, neither the plaintiffs nor the trial judge considered amending the complaint to embrace the new theory.

^1. As this Court stated in Brown v. Board of Education, 349 U.S. 294, 300, 75 S.Ct. 753, 756, 99 L.Ed. 1083: '(E)quity has been characterized by a practical flexibility in shaping its remedies and by a facility for adjusting and reconciling public and private needs. These (school desegregation) cases call for the exercise of these traditional attributes of equity power.'

^2. My Brother MARSHALL seems to ignore this fundamental fact when he states, post at 799, that 'the most essential finding (made by the District Court) was that Negro children in Detroit had been confined by intentional acts of segregation to a growing core of Negro schools surrounded by a receding ring of white schools.' This conclusion is simply not substantiated by the record presented in this case. The record here does support the claim made by the respondents that white and Negro students within Detroit who otherwise would have attended school together were separated by acts of the State or its subdivision. However, segregative acts within the city alone cannot be presumed to have produced—and no factual showing was made that they did produce—an increase in the number of Negro students in the city as a whole. It is this essential fact of a predominantly Negro school population in Detroit—caused by unknown and perhaps unknowable factors such as in-migration, birth rates, economic changes, or cumulative acts of private racial fears—that accounts for the 'growing core of Negro schools,' a 'core' that has grown to include virtually the entire city. The Constitution simply does not allow federal courts to attempt to change that situation unless and until it is shown that the State, or its political subdivisions, have contributed to cause the situation to exist. No record has been made in this case showing that the racial composition of the Detroit school population or that residential patterns within Detroit and in the surrounding areas were in any significant measure caused by governmental activity, and it follows that the situation over which my dissenting Brothers express concern cannot serve as the predicate for the remedy adopted by the District Court and approved by the Court of Appeals.

^1. Mich.Const., Art. 8, §§ 2, 3.

^2. See 484 F.2d 215, 247—248; Mich.Comp.Laws §§ 340.402, 340.431, 340.447, 388.681 (1970).

^3. Mich.Comp.Laws § 388.851 (1948), as amended by Act 231, Mich.Pub.Acts of 1949, and Act 175, Mich.Pub.Acts 1962.

^4. See Mich.Comp.Laws §§ 132.1 and 132.2 (1970); 3 App. 157.

^5. See 484 F.2d at 248—249.

^6. See Detroit Free Press, Nov. 8, 1972, p. 1A, col. 3. Michigan has recently passed legislation which could eliminate some, but not all, of the inequities in school financing. See Act 101, Mich.Pub.Acts of 1973.

^7. See 484 F.2d, at 246—247; Mich.Const. Art. 8, §§ 2, 3.

^8. See n. 2, supra.

^9. A tremendous change has occurred in the distribution of this country's black population since World War I. See Hauser, Demographic Factors in the Integration of the Negro, Daedalus 847 877 (fall 1965). In 1910, 73% of all blacks lived on farms and in rural areas; by 1960, 73% lived in urban areas, mainly in the largest metropolitan areas. Moreover, due to the fact that the black population is younger than the white population, the concentration of blacks in the cities is even more pronounced for the schoolage population. The pattern of change which has existed since World War I is continuing, and hence the proportion of blacks in the urban North and West will continue to increase. Dept. of Health, Education, and Welfare, J. Coleman et al., Equality of Educational Opportunity 39—40 (1966).

^10. There are some definite and systematic directions of difference between the schools attended by minorities and those attended by the majority. It appears to be in the most academically related areas that the schools of minority pupils show the most consistent deficiencies.' Dept. of Health, Education, and Welfare, Coleman et al., supra n. 9, at 120.

^11. That some school districts are markedly poorer than others is beyond question. The California Supreme Court has noted that per-pupil expenditures in two different districts—both located in the same county—were $2,223 and $616. Serrano v. Priest, 5 Cal.3d 584, 600 n. 15 (1971). In New York the Fleischmann Commission reported that the two Long Island districts of Great Neck and Levittown spent $2,078 and $1,189 respectively per pupil. 1 New York State Commission on the Quality, Cost, and Financing of Elementary and Secondary Education, Fleischmann Report 58 (1973). 'A further glaring inequity resulting from the current systems of school finance is that variations in per pupil expenditures among school districts tend to be inversely related to educational need. City students, with greater than average education deficiencies, consistently have less money spent on their education and have higher pupil/teacher ratios than do their high-income counterparts in the favored schools of suburbia.' Glickstein & Want, Inequality in School Financing: The Role of the Law, 25 Stan.L.Rev. 335, 338 (1973).

^12. Cities face an especially difficult problem in paying the cost of education, since they have the 'municipal overburden' which results from greater costs for health, public safety, sanitation, public works, transportation, public welfare, public housing, and recreation. Because of municipal overburden, cities on the average devote only about 30% of their budgets to their schools. This compares with the over 50% which is spent on schools by the suburbs. J. Berke & J. Callahan, Inequities in School Finance (1971), reprinted in Senate Select Committee on Equal Educational Opportunity, 92d Cong., 2d Sess., Report on Issues in School Finance 129, 142 (Comm.Print 1972); see Glickstein & Want, supra, n. 11, at 387.

^13. Mr. Justice STEWART indicates that equitable factors weigh in favor of local school control and the avoidance of administrative difficulty given the lack of an 'interdistrict' violation. Ante, at 755. It would seem to me that the equities are stronger in favor of the children of Detroit who have been deprived of their constitutional right to equal treatment by the State of Michigan.

^1. The percentage of Negro pupils in the Detroit student population rose to 64.9% in 1971, to 67.3% in 1972, and to 69.8% in 1973, amid a metropolitan school population whose racial composition in 1970 was 81% white and 19% Negro. 5 App. 16; Racial-Ethnic Distribution of Students and Employees in the Detroit Public Schools, October 1972, and October 1973; 484 F.2d 215, 250.

^2. The District Court's ruling on the Detroit-only desegregation plans is set out in full by the Court of Appeals, id., at 242—245, and is not otherwise officially reported.

^3. The Court has previously disapproved the implementation of proposed desegregation plans which operate to permit resegregation. Monroe v. Board of Comm'rs, 391 U.S. 450, 459—460, 88 S.Ct. 1700, 1705, 20 L.Ed.2d 733 (1968), ('free transfer' plan).

^4. The Court of Appeals also noted several specific instances of school district mergers ordered by the State Board of Education for financial reasons. 484 F.2d, at 247. Limitations on the authority of local school districts were also outlined by the Court of Appeals:

'Local school districts, unless they have the approval of the State Board of Education or the Superintendent of Public Instruction, cannot consolidate with another school district, annex territory, divide or attach parts of other districts, borrow monies in anticipation of State aid, or construct, reconstruct or remodel school buildings or additions to them.' Id., at 249. (Footnotes and supporting statutory citations omitted.)

And the Court of Appeals properly considered the State's statutory attempt to undo the adoption of a voluntary high school desegregation plan by the Detroit Board of Education as evidencing state control over local school district affairs. Ibid. Finally, it is also relevant to note that the District Court found that the school district boundaries in that segment of the metropolitan area preliminarily designated as the desegregation area 'in general bear no relationship to other municipal, county, or special district governments, needs or services,' that some educational services are already provided to students on an interdistrict basis requiring their travel from one district to another, and that local communities in the metropolitan area share noneducational interests in common, which do not adhere to school district lines, and have applied metropolitan solutions to other governmental needs. 345 F.Supp. 914, 934—935 (E.D.Mich.1972).

^5. These included the creation and alteration of attendance zones and feeder patterns from the elementary to the secondary schools in a manner naturally and predictably perpetuating racial segregation of students, the transportation of Negro students beyond predominantly white schools with available space to predominantly Negro schools, the use of optional attendance areas in neighborhoods in which Negro families had recently begun to settle to permit white students to transfer to predominantly white schools nearer the city limits, and the construction of schools in the heart of residentially segregated areas, thereby maximizing school segregation.

^1. Contrary to the Court's characterization, the use of racial ratios in this case in no way differed from that in Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971). Here, as there, mathematical ratios were used simply as 'a starting point in the process of shaping a remedy, rather than an inflexible requirement.' Id., at 25, 91 S.Ct., at 1280. It may be expected that a final desegregation plan in this case would deviate from a pure mathematical approach. Indeed, the District Court's most recent order appointing a panel of experts to draft an interdistrict plan requires only that the plan be designed 'to achieve the greatest degree of actual desegregation . . . (w)ithin the limitations of reasonable travel time and distance factors.' 345 F.Supp. 914, 918 (ED Mich.1972). Cf. 402 U.S., at 23, 91 S.Ct., at 1279.

^2. It does not appear that even the majority places any real weight on this consideration since it recognizes that interdistrict relief would be proper where a constitutional violation within one district produces a significant segregative effect in another district, see ante, at 744—745, thus allowing interdistrict relief to touch districts which have not themselves violated the Constitution.

^3. See Mich.Comp.Laws § 388.851 (1970).

^4. See § 388.1179.

^5. See §§ 388.629 and 340.600.

^6. See § 388.611. The State contributed an average of 34% of the operating budgets of the 54 school districts included in the original proposed desegregation area. In 11 of these districts, state contributions exceeded 50% of the operating budgets.

^7. See, e.g., id., § 340.575. See also 1949—1950 Report of the Attorney General 104 (Roth); Vol. 1, 1955 Report of the Attorney General 561 (Kavanagh); 1961—1962 Report of the Attorney General 533 (Kelley).

^8. See Mich.Comp.Laws §§ 211.34 and 340.681.

^9. § 340.569.

^10. §§ 257.811(c), 340.361, 340.781, 340.782, 388.371.

^11. § 340.575.

^12. § 388.1171.

^13. § 340.887(1).

^14. Op.Atty.Gen. No. 4705 (July 7, 1970), 1969—1970 Report of the Attorney General 156 (Kelley).

^15. See Mich.Comp.Laws § 340.253.

^16. See generally, §§ 340.401—340.415 (consolidations), 340.431—340.449 (annexations).

^17. See 1 Michigan Senate Journal, 1968, p. 423.

^18. See generally Mich.Comp.Laws §§ 340.461—340.468.

^19. Despite Mr. Justice STEWART's claim to the contrary, ante, at 756, n. 2, of his concurring opinion, the record fully supports my statement that Negro students were intentionally confined to a core of Negro schools within the city of Detroit. See, e.g., supra, at 784—785, 790—792. Indeed, Mr. Justice STEWART acknowledges that intentional acts of segregation by the State have separated white and Negro students within the city, and that the resulting core of all-Negro schools has grown to encompass most of the city. In suggesting that my approval of an interdistrict remedy rests on a further conclusion that the State or its political subdivisions have been responsible for the increasing percentage of Negro students in Detroit, my Brother STEWART misconceives the thrust of this dissent. In light of the high concentration of Negro students in Detroit, the District Judge's finding that a Detroit-only remedy cannot effectively cure the constitutional violation within the city should be enough to support the choice of an interdistrict remedy. Whether state action is responsible for the growth of the core of all-Negro schools in Detroit is, in my view, quite irrelevant.

The difficulty with Mr. Justice STEWART's position is that he, like the Court, confuses the inquiry required to determine whether there has been a substantive constitutional violation with that necessary to formulate an appropriate remedy once a constitutional violation has been shown. While a finding of state action is of course a prerequisite to finding a violation, we have never held that after unconstitutional state action has been shown, the District Court at the remedial stage must engage in a second inquiry to determine whether additional state action exists to justify a particular remedy. Rather, once a constitutional violation has been shown, the District Court is duty-bound to formulate an effective remedy and, in so doing, the court is entitled—indeed, it is required—to consider all the factual circumstances relevant to the framing of an effective decree. Thus, in Swann v. Charlotte-Mecklenburg Board of Education we held that the District Court must take into account the existence of extensive residential segregation in determining whether a racially neutral 'neighborhood school' attendance plan was an adequate desegregation remedy, regardless of whether this residential segregation was caused by state action. So here, the District Court was required to consider the facts that the Detroit school system was already predominantly Negro and would likely become all-Negro upon issuance of a Detroit-only decree in framing an effective desegregation remedy, regardless of state responsibility for this situation.

^20. In fact, the District Court remarked 'that this court's task is to enforce constitutional rights not to act as a schoolmaster; the court's task is to protect the constitutional rights here found violated with as little intrusion into the education process as possible. The court's objective is to establish the minimum constitutional framework within which the system of public schools may operate now and hereafter in a racially unified, non-discriminatory fashion. Within that framework the body politic, educators, parents, and most particularly the children must be given the maximum opportunity to experiment and secure a high quality, and equal, educational opportunity.' Pet.App. 82a.

^21. See, e.g., Mich.Comp.Laws §§ 340.69, 340.121(d), 340.359, 340.582, 340.582a, 340.590.

^22. See id., §§ 340.330—340.330u.

          See 411 U.S. 959, 93 S.Ct. 1919.

                    Syllabus

          The financing of public elementary and secondary schools in Texas is a product of state and local participation. Almost half of the revenues are derived from a largely state-funded program designed to provide a basic minimum educational offering in every school. Each district supplements state aid through an ad valorem tax on property within its jurisdiction. Appellees brought this class action on behalf of schoolchildren said to be members of poor families who reside in school districts having a low property tax base, making the claim that the Texas system's reliance on local property taxation favors the more affluent and violates equal protection requirements because of substantial interdistrict disparities in per-pupil expenditures resulting primarily from differences in the value of assessable property among the districts. The District Court, finding that wealth is a 'suspect' classification and that education is a 'fundamental' right, concluded that the system could be upheld only upon a showing, which appellants failed to make, that there was a compelling state interest for the system. The court also concluded that appellants failed even to

Page 2

demonstrate a reasonable or rational basis for the State's system. Held:

          1. This is not a proper case in which to examine a State's laws under standards of strict judicial scrutiny, since that test is reserved for cases involving laws that operate to the disadvantage of suspect classes or interfere with the exercise of fundamental rights and liberties explicitly or implicitly protected by the Constitution. Pp. 18—44.

          (a) The Texas system does not disadvantage any suspect class. It has not been shown to discriminate against any definable class of 'poor' people or to occasion discriminations depending on the relative wealth of the families in any district. And, insofar as the financing system disadvantages those who, disregarding their individual income characteristics, reside in comparatively poor school districts, the resulting class cannot be said to be suspect. Pp. 18—28.

          (b) Nor does the Texas school-financing system impermissibly interfere with the exercise of a 'fundamental' right or liberty. Though education is one of the most important services performed by the State, it is not within the limited category of rights recognized by this Court as guaranteed by the Constitution. Even if some identifiable quantum of education is arguably entitled to constitutional protection to make meaningful the exercise of other constitutional rights, here there is no showing that the Texas system fails to provide the basic minimal skills necessary for that purpose. Pp. 29—39.

          (c) Moreover, this is an inappropriate case in which to invoke strict scrutiny since it involves the most delicate and difficult questions of local taxation, fiscal planning, educational policy, and federalism, considerations counseling a more restrained form of review. Pp. 40—44.

          2. The Texas system does not violate the Equal Protection Clause of the Fourteenth Amendment. Though concededly imperfect, the system bears a rational relationship to a legitimate state purpose. While assuring a basic education for every child in the State, it permits and encourages participation in and significant control of each district's schools at the local level. Pp. 44—53.

          D.C., 337 F.Supp. 280, reversed.

Page 3

          Charles Alan Wright, Austin, Tex., for appellants.

          Arthur Gochman, San Antonio, Tex., for appellees.

  [Amicus Curiae Information from pages 3-5 intentionally omitted]

Page 4

           Mr. Justice POWELL delivered the opinion of the Court.

          This suit attacking the Texas system of financing public education was initiated by Mexican-American parents whose children attend the elementary and sec-

Page 5

ondary schools in the Edgewood Independent School District, an urban school district in San Antonio, Texas.¹ They brought a class action on behalf of schoolchildren throughout the State who are members of minority groups or who are poor and reside in school districts having a low property tax base. Named as defendants² were the State Board of Education, the Commissioner of Education, the State Attorney General, and the Bexar County (San Antonio) Board of Trustees. The com-

Page 6

plaint was filed in the summer of 1968 and a three-judge court was impaneled in January 1969.³ In December 1971⁴ the panel rendered its judgment in a per curiam opinion holding the Texas school finance system unconstitutional under the Equal Protection Clause of the Fourteenth Amendment.⁵ The State appealed, and we noted probable jurisdiction to consider the far-reaching constitutional questions presented. 406 U.S. 966, 92 S.Ct. 2413, 32 L.Ed.2d 665 (1972). For the reasons stated in this opinion, we reverse the decision of the District Court.

          The first Texas State Constitution, promulgated upon Texas' entry into the Union in 1845, provided for the establishment of a system of free schools. ⁶ Early in its history, Texas adopted a dual approach to the financing of its schools, relying on mutual participation by the local school districts and the State. As early as 1883, the state

Page 7

constitution was amended to provide for the creation of local school districts empowered to levy ad valorem taxes with the consent of local taxpayers for the 'erection . . . of school buildings' and for the 'further maintenance of public free schools.'⁷ Such local funds as were raised were supplemented by funds distributed to each district from the State's Permanent and Available School Funds.⁸ The Permanent School Fund, its predecessor established in 1854 with $2,000,000 realized from an annexation settlement,⁹ was thereafter endowed with millions of acres of public land set aside to assure a continued source of income for school support.¹⁰ The Available School Fund, which received income from the Permanent School Fund as well as from a state ad valorem property tax and other designated taxes,¹¹ served as the disbursing arm for most state educational funds throughout the late 1800's and first half of this century. Additionally, in 1918 an increase in state property taxes was used to finance a program providing free textbooks throughout the State.¹²

          Until recent times, Texas was a predominantly rural State and its population and property wealth were spread

Page 8

relatively evenly across the State.¹³ Sizable differences in the value of assessable property between local school districts became increasingly evident as the State became more industrialized and as rural-to-urban population shifts became more pronounced.¹⁴ The location of commercial and industrial property began to play a significant role in determining the amount of tax resources available to each school district. These growing disparities in population and taxable property between districts were responsible in part for increasingly notable differences in levels of local expenditure for education.¹⁵

          In due time it became apparent to those concerned with financing public education that contributions from the Available School Fund were not sufficient to ameliorate these disparities.¹⁶ Prior to 1939, the Available School Fund contributed money to every school district at a rate of $17.50 per school-age child.¹⁷ Although the amount was increased several times in the early 1940's,¹⁸

Page 9

the Fund was providing only $46 per student by 1945.¹⁹

          Recognizing the need for increased state funding to help offset disparities in local spending and to meet Texas' changing educational requirements, the state legislature in the late 1940's undertook a thorough evaluation of public education with an eye toward major reform. In 1947, an 18-member committee, composed of educators and legislators, was appointed to explore alternative systems in other States and to propose a funding scheme that would guarantee a minimum or basic educational offering to each child and that would help overcome interdistrict disparities in taxable resources. The Committee's efforts led to the passage of the Gilmer-Aikin bills, named for the Committee's co-chairmen, establishing the Texas Minimum Foundation School Program²⁰. Today, this Program accounts for approximately half of the total educational expenditures in Texas.²¹

          The Program calls for state and local contributions to a fund earmarked specifically for teacher salaries, operating expenses, and transportation costs. The State, supplying funds from its general revenues, finances approximately 80% of the Program, and the school districts are responsible—as a unit—for providing the remaining 20%. The districts' share, known as the Local Fund Assignment, is apportioned among the school districts

Page 10

under a formula designed to reflect each district's relative taxpaying ability. The Assignment is first divided among Texas' 254 counties pursuant to a complicated economic index that takes into account the relative value of each county's contribution to the State's total income from manufacturing, mining, and agricultural activities. It also considers each county's relative share of all payrolls paid within the State and, to a lesser extent, considers each county's share of all property in the State.²² Each county's assignment is then divided among its school districts on the basis of each district's share of assessable property within the county.²³ The district, in turn, finances its share of the Assignment out of revenues from local property taxation.

          The design of this complex system was twofold. First, it was an attempt to assure that the Foundation Program would have an equalizing influence on expenditure levels between school districts by placing the heaviest burden on the school districts most capable of paying. Second, the Program's architects sought to establish a Local Fund Assignment that would force every school district to contribute to the education of its children²⁴ but that would not by itself exhaust any district's resources.²⁵ Today every school district does impose a property tax from which it derives locally expendable

Page 11

funds in excess of the amount necessary to satisfy its Local Fund Assignment under the Foundation Program.

          In the years since this program went into operation in 1949, expenditures for education—from state as well as local sources have increased steadily. Between 1949 and 1967, expenditures increased approximately 500%.²⁶ In the last decade alone the total public school budget rose from $750 million to.$2.1 billion²⁷ and these increases have been reflected in consistently rising perpupil expenditures throughout the State.²⁸ Teacher salaries, by far the largest item in any school's budget, have increased dramatically—the state-supported minimum salary for teachers possessing college degrees has risen from $2,400 to $6,000 over the last 20 years.²⁹

          The school district in which appellees reside, the Edgewood Independent School District, has been compared throughout this litigation with the Alamo Heights Independent School District. This comparison between the least and most affluent districts in the San Antonio area serves to illustrate the manner in which the dual system of finance operates and to indicate the extent to which substantial disparities exist despite the State's impressive progress in recent years. Edgewood is one of seven public school districts in the metropolitan are enrolled in its 25 elementary

Page 12

and secondary schools. The district is are enrolled in its 25 elementary situated in the core-city sector of San Antonio in a residential neighborhood that has little commercial or industrial property. The residents are predominantly of Mexican-American descent: approximately 90% of the student population is Mexican-American and over 6% is Negro. The average assessed property value per pupil is $5,960—the lowest in the metropolitan area—and the median family income ($4,686) is also the lowest.³⁰ At an equalized tax rate of $1.05 per $100 of assessed property the highest in the metropolitan area—the district contributed $26 to the education of each child for the 1967—1968 school year above its Local Fund Assignment for the Minimum Foundation Program. The Foundation Program contributed $222 per pupil for a state-local total of $248.³¹ Federal funds added another $108 for a total of $356 per pupil.³²

          Alamo Heights is the most affluent school district in San Antonio. Its six schools, housing approximately 5,000 students, are situated in a residential community quite unlike the Edgewood District. The school population is predominantly 'Anglo,' having only 18% Mexican-Amer-

Page 13

icans and less than 1% Negroes. The assessed property value per pupil exceeds $49,000,³³ and the median family income is $8,001. In 1967—1968 the local tax rate of $.85 per $100 of valuation yielded $333 per pupil over and above its contribution to the Foundation Program. Coupled with the $225 provided from that Program, the district was able to supply $558 per student. Supplemented by a $36 per-pupil grant from federal sources, Alamo Heights spent $594 per pupil.

          Although the 1967—1968 school year figures provide the only complete statistical breakdown for each category of aid,³⁴ more recent partial statistics indicate that the previously noted trend of increasing state aid has been significant. For the 1970—1971 school year, the Foundation School Program allotment for Edgewood was $356 per pupil, a 62% increase over the 1967—68 school year. Indeed, state aid alone in 1970—1971 equaled Edgewood's entire 1967—1968 school budget from local, state, and federal sources. Alamo Heights enjoyed a similar increase under the Foundation Program, netting $491 per pupil in 1970—1971.³⁵ These recent figures

Page 14

also reveal the extent to which these two districts' allotments were funded from their own required contributions to the Local Fund Assignment. Alamo Heights, because of its relative wealth, was required to contribute out of its local property tax collections approximately $100 per pupil, or about 20% of its Foundation grant. Edgewood, on the other hand, paid only $8.46 per pupil, which is about 2.4% of its grant.³⁶ It appears then that, at least as to these two districts, the Local Fund Assignment does reflect a rough approximation of the relative taxpaying potential of each.³⁷

Page 15

                    Despite these recent increases, substantial interdistrict disparities in school expenditures found by the District Court to prevail in San Antonio and in varying degrees throughout the State³⁸ still exist. And it was

Page 16

these disparities, largely attributable to differences in the amounts of money collected through local property taxation, that led the District Court to conclude that Texas' dual system of public school financing violated the Equal Protection Clause. The District Court held that the Texas system discriminates on the basis of wealth in the manner in which education is provided for its people. 337 F.Supp., at 282. Finding that wealth is a 'suspect' classification and that education is a 'fundamental' interest, the District Court held that the Texas system could be sustained only if the State could show that it was premised upon some compelling state interest. Id., at 282—284. On this issue the court concluded that '(n)ot only are defendants unable to demonstrate compelling state interests . . . they fail even to establish a reasonable basis for these classifications.' Id., at 284.

          Texas virtually concedes that its historically rooted dual system of financing education could not withstanding the strict judicial scrutiny that this Court has found appropriate in reviewing legislative judgments that interfere with fundamental constitutional rights³⁹ or that involve suspect classifications.⁴⁰ If, as previous decisions have indicated, strict scrutiny means that the State's system is not entitled to the usual presumption of validity, that the State rather than the complainants must carry a 'heavy burden of justification,' that the State must

Page 17

demonstrate that its educational system has been structured with 'precision,' and is 'tailored' narrowly to serve legitimate objectives and that it has selected the 'less drastic means' for effectuating its objectives,⁴¹ the Texas financing system and its counterpart in virtually every other State will not pass muster. The State candidly admits that '(n)o one familiar with the Texas system would contend that it has yet achieved perfection.'⁴² Apart from its concession that educational financing in Texas has 'defects'⁴³ and 'imperfections,'⁴⁴ the State defends the system's rationality with vigor and disputes the District Court's finding that it lacks a 'reasonable basis.'

          This, then, establishes the framework for our analysis. We must decide, first, whether the Texas system of financing public education operates to the disadvantage of some suspect class or impinges upon a fundamental right explicitly or implicitly protected by the Constitution, thereby requiring strict judicial scrutiny. If so, the judgment of the District Court should be affirmed. If not, the Texas scheme must still be examined to determine whether it rationally furthers some legitimate, articulated state purpose and therefore does not constitute an invidious discrimination in violation of the Equal Protection Clause of the Fourteenth Amendment.

          The District Court's opinion does not reflect the novelty and complexity of the constitutional questions posed by appellees' challenge to Texas' system of school financing. In concluding that strict judicial scrutiny was required,

Page 18

that court relied on decisions dealing with the rights of indigents to equal treatment in the criminal trial and appellate processes,⁴⁵ and on cases disapproving wealth restrictions on the right to vote.⁴⁶ Those cases, the District Court concluded, established wealth as a suspect classification. Finding that the local property tax system discriminated on the basis of wealth, it regarded those precedents as controlling. It then reasoned, based on decisions of this Court affirming the undeniable importance of education, ⁴⁷ that there is a fundamental right to education and that, absent some compelling state justification, the Texas system could not stand.

          We are unable to agree that this case, which in significant aspects is sui generis, may be so neatly fitted into the conventional mosaic of constitutional analysis under the Equal Protection Clause. Indeed, for the several reasons that follow, we find neither the suspect-classification not the fundamental-interest analysis persuasive.

            The wealth discrimination discovered by the District Court in this case, and by several other courts that have recently struck down school-financing laws in other States,⁴⁸ is quite unlike any of the forms of wealth dis-

Page 19

crimination heretofore reviewed by this Court. Rather than focusing on the unique features of the alleged discrimination, the courts in these cases have virtually assumed their findings of a suspect classification through a simplistic process of analysis: since, under the traditional systems of financing public schools, some poorer people receive less expensive educations than other more affluent people, these systems discriminate on the basis of wealth. This approach largely ignores the hard threshold questions, including whether it makes a difference for purposes of consideration under the Constitution that the class of disadvantaged 'poor' cannot be identified or defined in customary equal protection terms, and whether the relative—rather than absolute—nature of the asserted deprivation is of significant consequence. Before a State's laws and the justifications for the classifications they create are subjected to strict judicial scrutiny, we think these threshold considerations must be analyzed more closely than they were in the court below.

          The case comes to us with no definitive description of the classifying facts or delineation of the disfavored class. Examination of the District Court's opinion and of appellees' complaint, briefs, and contentions at oral argument suggests, however, at least three ways in which the discrimination claimed here might be described. The Texas system of school financing might be regarded as discriminating (1) against 'poor' persons whose incomes fall below some identifiable level of poverty or who might be characterized as functionally 'indigent,'⁴⁹ or

Page 20

(2) against those who are relatively poorer than others, ⁵⁰ or (3) against all those who, irrespective of their personal incomes, happen to reside in relatively poorer school districts.⁵¹ Our task must be to ascertain whether, in fact, the Texas system has been shown to discriminate on any of these possible bases and, if so, whether the resulting classification may be regarded as suspect.

          The precedents of this Court provide the proper starting point. The individuals, or groups of individuals, who constituted the class discriminated against in our prior cases shared two distinguishing characteristics: because of their impecunity they were completely unable to pay for some desired benefit, and as a consequence, they sustained an absolute deprivation of a meaningful opportunity to enjoy that benefit. In Griffin v. Illinois,

Page 21

  351 U.S. 12, 76 S.Ct. 585, 100 L.Ed. 891 (1956), and its progeny,⁵² the Court invalidated state laws that prevented an indigent criminal defendant from acquiring a transcript, or an adequate substitute for a transcript, for use at several stages of the trial and appeal process. The payment requirements in each case were found to occasion de facto discrimination against those who, because of their indigency, were totally unable to pay for transcripts. And the Court in each case emphasized that no constitutional violation would have been shown if the State had provided some 'adequate substitute' for a full stenographic transcript. Britt v. North Carolina, 404 U.S. 226, 228, 92 S.Ct. 431, 434, 30 L.Ed.2d 400 (1971); Gardner v. California, 393 U.S. 367, 89 S.Ct. 580, 21 L.Ed.2d 601 (1969); Draper v. Washington, 372 U.S. 487, 83 S.Ct. 774, 9 L.Ed.2d 899 (1963); Eskridge v. Washington State Board of Prisons, 357 U.S. 214, 78 S.Ct. 1061, 2 L.Ed.2d 1269 (1958).

          Likewise, in Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963), a decision establishing an indigent defendant's right to court-appointed counsel on direct appeal, the Court dealt only with defendants who could not pay for counsel from their own resources and who had no other way of gaining representation. Douglas provides no relief for those on whom the burdens of paying for a criminal defense are relatively speaking, great but not insurmountable. Nor does it deal with relative differences in the quality of counsel acquired by the less wealthy.

          Williams v. Illinois, 399 U.S. 235, 90 S.Ct. 2018, 26 L.Ed.2d 586 (1970), and Tate v. Short, 401 U.S. 395, 91 S.Ct. 668, 28 L.Ed.2d 130 (1971), struck down criminal penalties that subjected indigents to incarceration simply be-

Page 22

cause of their inability to pay a fine. Again, the disadvantaged class was composed only of persons who were totally unable to pay the demanded sum. Those cases do not touch on the question whether equal protection is denied to persons with relatively less money on whom designated fines impose heavier burdens. The Court has not held that fines must be structured to reflect each person's ability to pay in order to avoid disproportionate burdens. Sentencing judges may, and often do, consider the defendant's ability to pay, but in such circumstances they are guided by sound judicial discretion rather than by constitutional mandate.

          Finally, in Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972), the Court invalidated the Texas filing-fee requirement for primary elections. Both of the relevant classifying facts found in the previous cases were present there. The size of the fee, often running into the thousands of dollars and, in at least one case, as high as $8,900, effectively barred all potential candidates who were unable to pay the required fee. As the system provided 'no reasonable alternative means of access to the ballot' (id., at 149, 92 S.Ct. at 859), inability to pay occasioned an absolute denial of a position on the primary ballot.

          Only appellees' first possible basis for describing the class disadvantaged by the Texas school-financing system—discrimination against a class of definably 'poor' persons—might arguably meet the criteria established in these prior cases. Even a cursory examination, however, demonstrates that neither of the two distinguishing characteristics of wealth classifications can be found here. First, in support of their charge that the system discriminates against the 'poor,' appellees have made no effort to demonstrate that it operates to the peculiar disadvantage of any class fairly definable as indigent, or as composed of persons whose incomes are beneath any

Page 23

designated poverty level. Indeed, there is reason to believe that the poorest families are not necessarily clustered in the poorest property districts. A recent and exhaustive study of school districts in Connecticut concluded that '(i)t is clearly incorrect . . . to contend that the 'poor' live in 'poor' districts . . .. Thus, the major factual assumption of Serrano—that the educational financing system discriminates against the 'poor'—is simply false in Connecticut.'⁵³ Defining 'poor' families as those below the Bureau of the Census 'poverty level,'⁵⁴ the Connecticut study found, not surprisingly, that the poor were clustered around commercial and industrial areas—those same areas that provide the most attractive sources of property tax income for school districts.⁵⁵ Whether a similar pattern would be discovered in Texas is not known, but there is no basis on the record in this case for assuming that the poorest people—defined by reference to any level of absolute impecunity—are concentrated in the poorest districts.

          Second, neither appellees nor the District Court addressed the fact that, unlike each of the foregoing cases, lack of personal resources has not occasioned an absolute deprivation of the desired benefit. The argument here is not that the children in districts having relatively low assessable property values are receiving no public education; rather, it is that they are receiving a poorer quality education than that available to children in districts having more assessable wealth. Apart from the unsettled and disputed question whether the quality of education may be determined by the amount of money

Page 24

expended for it,⁵⁶ a sufficient answer to appellees' argument is that, at least where wealth is involved, the Equal Protection Clause does not require absolute equality or precisely equal advantages.⁵⁷ Nor indeed, in view of the infinite variables affecting the educational process, can any system assure equal quality of education except in the most relative sense. Texas asserts that the Minimum Foundation Program provides an 'adequate' education for all children in the State. By providing 12 years of free public-school education, and by assuring teachers, books, transportation, and operating funds, the Texas Legislature has endeavored to 'guarantee, for the welfare of the state as a whole, that all people shall have at least an adequate program of education. This is what is meant by 'A Minimum Foundation Program of Education."⁵⁸ The State repeatedly asserted in its briefs in this Court that it has fulfilled this desire and that it now assures 'every child in every school district an adequate education.'⁵⁹ No proof was offered at trial persuasively discrediting or refuting the State's assertion.

Page 25

          For these two reasons—the absence of any evidence that the financing system discriminates against any definable category of 'poor' people or that it results in the absolute deprivation of education—the disadvantaged class is not susceptible of identification in traditional terms.⁶⁰

          As suggested above, appellees and the District Court may have embraced a second or third approach, the second of which might be characterized as a theory of relative or comparative discrimination based on family income. Appellees sought to prove that a direct correlation exists between the wealth of families within each district and the expenditures therein for education. That is, along a continuum, the poorer the family the lower the dollar amount of education received by the family's children.

          The principal evidence adduced in support of this comparative-discrimination claim is an affidavit submitted by Professor Joele S. Berke of Syracuse University's Educational Finance Policy Institute. The District Court, relying in major part upon this affidavit and apparently accepting the substance of appellees' theory,

Page 26

noted, first, a positive correlation between the wealth of school districts, measured in terms of assessable property per pupil, and their levels of per-pupil expenditures. Second, the court found a similar correlation between district wealth and the personal wealth of its residents, measured in terms of median family income. 337 F.Supp., at 282 n. 3.

          If, in fact, these correlations could be sustained, then it might be argued that expenditures on education—equated by appellees to the quality of education—are dependent on personal wealth. Appellees' comparative-discrimination theory would still face serious unanswered questions, including whether a bare positive correlation or some higher degree of correlation⁶¹ is necessary to provide a basis for concluding that the financing system is designed to operate to the peculiar disadvantage of the comparatively poor, ⁶² and whether a class of this size and diversity could ever claim the special protection accorded 'suspect' classes. These questions need not be addressed in this case, however, since appellees' proof fails to support their allegations or the District Court's conclusions.

          Professor Berke's affidavit is based on a survey of approximately 10% of the school districts in Texas. His findings, previously set out in the margin, ⁶³ show only

Page 27

that the wealthiest few districts in the sample have the highest median family incomes and spend the most on education, and that the several poorest districts have the lowest family incomes and devote the least amount of money to education. For the remainder of the districts—96 districts composing almost 90% of the sample the correlation is inverted, i.e., the districts that spend next to the most money on education are populated by families having next to the lowest median family incomes while the districts spending the least have the highest median family incomes. It is evident that, even if the conceptual questions were answered favorably to appellees, no factual basis exists upon which to found a claim of comparative wealth discrimination.⁶⁴

          This brings us, then, to the third way in which the classification scheme might be defined—district wealth discrimination. Since the only correlation indicated by the evidence is between district property wealth and expenditures, it may be argued that discrimination might be found without regard to the individual income characteristics of district residents. Assuming a perfect correlation between district property wealth and expenditures from top to to bottom, the disadvantaged class might be

Page 28

viewed as encompassing every child in every district except the district that has the most assessable wealth and spends the most on education.⁶⁵ Alternatively, as suggested in Mr. Justice MARSHALL's dissenting opinion, post, at 96, the class might be defined more restrictively to include children in districts with assessable property which falls below the statewide average, or median, or below some other artificially defined level.

          However described, it is clear that appellees' suit asks this Court to extend its most exacting scrutiny to review a system that allegedly discriminates against a large, diverse, and amorphous class, unified only by the common factor of residence in districts that happen to have less taxable wealth than other districts.⁶⁶ The system of alleged discrimination and the class it defines have none of the traditional indicia of suspectness: the class is not saddled with such disabilities, or subjected to such a history of purposeful unequal treatment, or relegated to such a position of political powerlessness as to command extraordinary protection from the majoritarian political process.

          We thus conclude that the Texas system does not operate to the peculiar disadvantage of any suspect class.

Page 29

But in recognition of the fact that this Court has never heretofore held that wealth discrimination alone provides an adequate basis for invoking strict scrutiny, appellees have not relied solely on this contention.⁶⁷ They also assert that the State's system impermissibly interferes with the exercise of a 'fundamental' right and that accordingly the prior decisions of this Court require the application of the strict standard of judicial review. Graham v. Richardson, 403 U.S. 365, 375—376, 91 S.Ct. 1848, 1853—1854, 29 L.Ed.2d 534 (1971); Kramer v. Union Free School District, 395 U.S. 621, 89 S.Ct. 1886, 23 L.Ed.2d 583 (1969); Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969). It is this question—whether education is a fundamental right, in the sense that it is among the rights and liberties protected by the Constitution—which has so consumed the attention of courts and commentators in recent years.⁶⁸

          In Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), a unanimous Court recognized that 'education is perhaps the most important function of state and local governments.' Id., at 493, 74 S.Ct., at 691. What was said there in the context of racial discrimination has lost none of its vitality with the passage of time:

          'Compulsory school attendance laws and the great expenditures for education both demonstrate our

Page 30

          recognition of the importance of education to our democratic society. It is required in the performance of our most basic public responsibilities, even service in the armed forces. It is the very foundation of good citizenship. Today it is a principal instrument in awakening the child to cultural values, in preparing him for later professional training, and in helping him to adjust normally to his environment. In these days, it is doubtful that any child may reasonably be expected to succeed in life if he is denied the opportunity of an education. Such an opportunity, where the state has undertaken to provide it, is a right which must be made available to all on equal terms.' Ibid.

          This theme, expressing an abiding respect for the vital role of education in a free society, may be found in numerous opinions of Justices of this Court writing both before and after Brown was decided. Wisconsin v. Yoder, 406 U.S. 205, 213, 92 S.Ct. 1526, 1532, 32 L.Ed.2d 234 (Burger, C.J.), 237, 238—239, 92 S.Ct. 1544 1545 (White, J.), (1972); Abington School Dist. v. Schempp, 374 U.S. 203, 230, 83 S.Ct. 1560, 1575, 10 L.Ed.2d 844 (1963) (Brennan, J.); People of State of Illinois ex rel. McCollum v. Board of Education, 333 U.S. 203, 212, 68 S.Ct. 461, 465, 92 L.Ed. 649 (1948) (Frankfurter, J.); Pierce v. Society of Sisters, 268 U.S. 510, 45 S.Ct. 571, 69 L.Ed. 1070 (1925); Meyer v. Nebraska, 262 U.S. 390, 43 S.Ct. 625, 67 L.Ed. 1042 (1923); Interstate Consolidated Street R. Co. v. Massachusetts, 207 U.S. 79, 28 S.Ct. 26, 52 L.Ed. 111 (1907).

          Nothing this Court holds today in any way detracts from our historic dedication to public education. We are in complete agreement with the conclusion of the three-judge panel below that 'the grave significance of education both to the individual and to our society' cannot be doubted.⁶⁹ But the importance of a service performed by the State does not determine whether it must be regarded as fundamental for purposes of examination under the Equal Protection Clause. Mr. Justice

Page 31

Harlan, dissenting from the Court's application of strict scrutiny to a law impinging upon the right of interstate travel, admonished that '(v)irtually every state statute affects important rights.' Shapiro v. Thompson, 394 U.S., at 655, 661, 89 S.Ct., at 1342, 1345. In his view, if the degree of judicial scrutiny of state legislation fluctuated, depending on a majority's view of the importance of the interest affected, we would have gone 'far toward making this Court a 'super-legislature." Ibid. We would, indeed, then be assuming a legislative role and one for which the Court lacks both authority and competence. But Mr. Justice Stewart's response in Shapiro to Mr. Justice Harlan's concern correctly articulates the limits of the fundamental-rights rationale employed in the Court's equal protection decisions:

          'The Court today does not 'pick out particular human activities, characterize them as 'fundamental,' and give them added protection . . ..' To the contrary, the Court simply recognizes, as it must, an established constitutional right, and gives to that right no less protection than the Constitution itself demands.' Id., at 642, 89 S.Ct., at 1335. (Emphasis in original.)

          Mr. Justice Stewart's statement serves to underline what the opinion of the Court in Shapiro makes clear. In subjecting to strict judicial scrutiny state welfare eligibility statutes that imposed a one-year durational residency requirement as a precondition to receiving AFDC benefits, the Court explained:

          '(I)n moving from State to State . . . appellees were exercising a constitutional right, and any classification which serves to penalize the exercise of that right, unless shown to be necessary to promote a compelling governmental interest, is unconstitutional.' Id., at 634, 89 S.Ct., at 1331. (Emphasis in original.)

Page 32

          The right to interstate travel had long been recognized as a right of constitutional significance,⁷⁰ and the Court's decision, therefore, did not require an ad hoc determination as to the social or economic importance of that right.⁷¹

          Lindsey v. Normet, 405 U.S. 56, 92 S.Ct. 862, 31 L.Ed.2d 36 (1972), decided only last Term, firmly reiterates that social importance is not the critical determinant for subjecting state legislation to strict scrutiny. The complainants in that case, involving a challenge to the procedural limitations imposed on tenants in suits brought by landlords under Oregon's Forcible Entry and Wrongful Detainer Law, urged the Court to examine the operation of the statute under 'a more stringent standard than mere rationality.' Id., at 73, 92 S.Ct., at 874. The tenants argued that the statutory limitations implicated 'fundamental interests which are particularly important to the poor,' such as the "need for decent shelter" and the "right to retain peaceful possession of one's home." Ibid. Mr. Justice White's analysis, in his opinion for the Court is instructive:

          'We do not denigrate the importance of decent, safe and sanitary housing. But the Constitution does not provide judicial remedies for every social and economic ill. We are unable to perceive in that document any constitutional guarantee of access

Page 33

          to dwellings of a particular quality or any recognition of the right of a tenant to occupy the real property of his landlord beyond the term of his lease, without the payment of rent . . .. Absent constitutional mandate, the assurance of adequate housing and the definition of landlord-tenant relationships are legislative, not judicial, functions.' Id., at 74, 92 S.Ct., at 874. (Emphasis supplied.)

          Similarly, in Dandridge v. Williams, 397 U.S. 471, 90 S.Ct. 1153, 25 L.Ed.2d 491 (1970), the Court's explicit recognition of the fact that the 'administration of public welfare assistance . . . involves the most basic economic needs of impoverished human beings,' id., at 485, 90 S.Ct., at 1162, ⁷² provided no basis for departing from the settled mode of constitutional analysis of legislative classifications involving questions of economic and social policy. As in the case of housing, the central importance of welfare benefits to the poor was not an adequate foundation for requiring the State to justify its law by showing some compelling state interest. See also Jefferson v. Hackney, 406 U.S. 535, 92 S.Ct. 1724, 32 L.Ed.2d 285 (1972); Richardson v. Belcher, 404 U.S. 78, 92 S.Ct. 254, 30 L.Ed.2d 231 (1971).

          The lesson of these cases in addressing the question now before the Court is plain. It is not the province of this Court to create substantive constitutional rights in the name of guaranteeing equal protection of the laws. Thus, the key to discovering whether education is 'fundamental' is not to be found in comparisons of the relative societal significance of education as opposed to subsistence or housing. Nor is it to be found by weighing whether education is as important as the right to travel. Rather, the answer lies in assessing whether there is a right to education explicitly or implicitly guaranteed by the Con-

Page 34

stitution. Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349 (1972);⁷³ Dunn v. Blumstein, 405 U.S. 330, 92 S.Ct. 995, 31 L.Ed.2d 274 (1972);⁷⁴ Police Dept. of City of Chicago v. Mosley, 408 U.S. 92, 92 S.Ct. 2286, 33 L.Ed.2d 212 (1972);⁷⁵ Skinner v. Oklahoma ex rel. Williamson, 316 U.S. 535, 62 S.Ct. 1110, 86 L.Ed. 1655 (1942). ⁷⁶

Page 35

          Education, of course, is not among the rights afforded explicit protection under our Federal Constitution. Nor do we find any basis for saying it is implicitly so protected. As we have said, the undisputed importance of education will not alone cause this Court to depart from the usual standard for reviewing a State's social and economic legislation. It is appellees' contention, however, that education is distinguishable from other services and benefits provided by the State because it bears a peculiarly close relationship to other rights and liberties accorded protection under the Constitution. Specifically, they insist that education is itself a fundamental personal right because it is essential to the effective exercise of First Amendment freedoms and to intelligent utilization of the right to vote. In asserting a nexus between speech and education, appellees urge that the right to speak is meaningless unless the speaker is capable of articulating his thoughts intelligently and persuasively. The 'marketplace of ideas' is an empty forum for those lacking basic communicative tools. Likewise, they argue that the corollary right to receive information⁷⁷ becomes little more than a hollow privilege when the recipient has not been taught to read, assimilate, and utilize available knowledge.

          A similar line of reasoning is pursued with respect to the right to vote. ⁷⁸ Exercise of the franchise, it is contended, cannot be divorced from the educational foun-

Page 36

dation of the voter. The electoral process, if reality is to conform to the democratic ideal, depends on an informed electorate: a voter cannot cast his ballot intelligently unless his reading skills and thought processes have been adequately developed.

          We need not dispute any of these propositions. The Court has long afforded zealous protection against unjustifiable governmental interference with the individual's rights to speak and to vote. Yet we have never presumed to possess either the ability or the authority to guarantee to the citizenry the most effective speech or the most informed electoral choice. That these may be desirable goals of a system of freedom of expression and of a representative form of government is not to be doubted.⁷⁹ These are indeed goals to be pursued by a people whose thoughts and beliefs are freed from governmental interference. But they are not values to be implemented by judicial instrusion into otherwise legitimate state activities.

          Even if it were conceded that some identifiable quantum of education is a constitutionally protected prerequisite to the meaningful exercise of either right, we have no indication that the present levels of

Page 37

educational expendi tures in Texas provide an education that falls short. Whatever merit appellees' argument might have if a State's financing system occasioned an absolute denial of educational opportunities to any of its children, that argument provides no basis for finding an interference with fundamental rights where only relative differences in spending levels are involved and where—as is true in the present case—no charge fairly could be made that the system fails to provide each child with an opportunity to acquire the basic minimal skills necessary for the enjoyment of the rights of speech and of full participation in the political process.

          Furthermore, the logical limitations on appellees' nexus theory are difficult to perceive. How, for instance, is education to be distinguished from the significant personal interests in the basics of decent food and shelter? Empirical examination might well buttress an assumption that the ill-fed, ill-clothed, and ill-housed are among the most ineffective participants in the political process, and that they derive the least enjoyment from the benefits of the First Amendment.⁸⁰ If so, appellees' thesis would cast serious doubt on the authority of Dandridge v. Williams, supra and Lindsey v. Normer, supra.

          We have carefully considered each of the arguments supportive of the District Court's finding that education is a fundamental right or liberty and have found those arguments unpersuasive. In one further respect we find this a particularly inappropriate case in which to subject state action to strict judicial scrutiny. The present case, in another basic sense, is significantly different from any of the cases in which the Court has

Page 38

applied strict scrutiny to state or federal legislation touching upon constitutionally protected rights. Each of our prior cases involved legislation which 'deprived,' 'infringed,' or 'interfered' with the free exercise of some such fundamental personal right or liberty. See Skinner v. Oklahoma, ex rel. Williamson, supra, 316 U.S. at 536, 62 S.Ct. at 1111; Shapiro v. Thompson, supra, 394 U.S. at 634, 89 S.Ct. at 1331; Dunn v. Blumstein, supra, 405 U.S. at 338—343, 92 S.Ct. at 1001—1004. A critical distinction between those cases and the one now before us lies in what Texas is endeavoring to do with respect to education. Mr. Justice Brennan, writing for the Court in Katzenbach v. Morgan, 384 U.S. 641, 86 S.Ct. 1717, 16 L.Ed.2d 828 (1966), expresses well the salient point:⁸¹

          'This is not a complaint that Congress . . . has unconstitutionally denied or diluted anyone's right to vote but rather that Congress violated the Constitution by not extending the relief effected (to others similarly situated) . . ..

          '(The federal law in question) does not restrict or deny the franchise but in effect extends the franchise to persons who otherwise would be denied it by state law. . . . We need only decide whether the challenged limitation on the relief effected . .. was permissible. In deciding that question, the principle that calls for the closest scrutiny of distinctions in laws denying fundamental rights . . . is

Page 39

          inapplicable; for the distinction challenged by appellees is presented only as a limitation on a reform measure aimed at eliminating an existing barrier to the exercise of the franchise. Rather, in deciding the constitutional propriety of the limitations in such a reform measure we are guided by the familiar principles that a 'statute is not invalid under the Constitution because it might have gone farther than it did,' . . . that a legislature need not 'strike at all evils at the same time,' . . . and that 'reform may take one step at a time, addressing itself to the phase of the problem which seems most acute to the legislative mind . . .." Id., at 656—657, 86 S.Ct., at 1727. (Emphasis in original.)

          The Texas system of school financing is not unlike the federal legislation involved in Katzenbach in this regard. Every step leading to the establishment of the system Texas utilizes today—including the decisions permitting localities to tax and expend locally, and creating and continuously expanding the state aid—was implemented in an effort to extend public education and to improve its quality.⁸² Of course, every reform that benefits some more than others may be criticized for what it fails to accomplish. But we think it plain that, in substance, the thrust of the Texas system is affirmative and reformatory and, therefore, should be scrutinized under judicial principles sensitive to the nature of the State's efforts and to the rights reserved to the States under the Constitution.⁸³

Page 40

          It should be clear, for the reasons stated above and in accord with the prior decisions of this Court, that this is not a case in which the challenged state action must be subjected to the searching judicial scrutiny reserved for laws that create suspect classifications or impinge upon constitutionally protected rights.

          We need not rest our decision, however, solely on the inappropriateness of the strict-scrutiny test. A century of Supreme Court adjudication under the Equal Protection Clause affirmatively supports the application of the traditional standard of review, which requires only that the State's system be shown to bear some rational relationship to legitimate state purposes. This case represents far more than a challenge to the manner in which Texas provides for the education of its children. We have here nothing lass than a direct attack on the way in which Texas has chosen to raise and disburse state and local tax revenues. We are asked to condemn the State's judgment in conferring on political subdivisions the power to tax local property to supply revenues for local interests. In so doing, appellees would have the Court intrude in an area in which it has traditionally deferred to state legislatures.⁸⁴ This Court has often admonished against such interferences with the State's fiscal policies under the Equal Protection Clause:

          'The broad discretion as to classification possessed by a legislature in the field of taxation has long been recognized. . . . (T)he passage of time has only served to underscore the wisdom of that recognition of the large area of discretion which is needed by a legislature in formulating sound tax poli-

Page 41

          cies. . . . It has . . . been pointed out that in taxation, even more than in other fields, legislatures possess the greatest freedom in classification. Since the members of a legislature necessarily enjoy a familiarity with local conditions which this Court cannot have, the presumption of constitutionality can be overcome only by the most explicit demonstration that a classification is a hostile and oppressive discrimination against particular persons and classes. . . .' Madden v. Kentucky, 309 U.S. 83, 87—88, 60 S.Ct. 406, 408, 84 L.Ed. 590 (1940).

          See also Lehnhausen v. Lake Shore Auto Parts Co., 410 U.S. 356, 93 S.Ct. 1001, 35 L.Ed.2d 351 (1973); Wisconsin v. J. C. Penney Co., 311 U.S. 435, 445, 61 S.Ct. 246, 250, 85 L.Ed. 267 (1940).

          Thus, we stand on familiar grounds when we continue to acknowledge that the Justices of this Court lack both the expertise and the familiarity with local problems so necessary to the making of wise decisions with respect to the raising and disposition of public revenues. Yet, we are urged to direct the States either to alter drastically the present system or to throw out the property tax altogether in favor of some other form of taxation. No scheme of taxation, whether the tax is imposed on property, income, or purchases of goods and services, has yet been devised which is free of all discriminatory impact. In such a complex arena in which no perfect alternatives exist, the Court does well not to impose too rigorous a standard of scrutiny lest all local fiscal schemes become subjects of criticism under the Equal Protection Clause.⁸⁵

Page 42

          In addition to matters of fiscal policy, this case also involves the most persistent and difficult questions of educational policy, another area in which this Court's lack of specialized knowledge and experience counsels against premature interference with the informed judgments made at the state and local levels. Education, perhaps even more than welfare assistance, presents a myriad of 'intractable economic, social, and even philosophical problems.' Dandridge v. Williams, 397 U.S., at 487, 90 S.Ct. at 1163. The very complexity of the problems of financing and managing a statewide public school system suggests that 'there will be more than one constitutionally permissible method of solving them,' and that, within the limits of rationality, 'the legislature's efforts to tackle the problems' should be entitled to respect. Jefferson v. Hackney, 406 U.S., at 546—547, 92 S.Ct., at 1731. On even the most basic questions in this area the scholars and educational experts are divided. Indeed, one of the major

Page 43

sources of controversy concerns the extent to which there is a demonstrable correlation between educational expenditures and the quality of education⁸⁶—an assumed correlation underlying virtually every legal conclusion drawn by the District Court in this case. Related to the questioned relationship between cost and quality is the equally unsettled controversy as to the proper goals of a system of public education.⁸⁷ And the question regarding the most effective relationship between state boards of education and local school boards, in terms of their respective responsibilities and degrees of control, is now undergoing searching re-examination. The ultimate wisdom as to these and related problems of education is not likely to be divined for all time even by the scholars who now so earnestly debate the issues. In such circumstances, the judiciary is well advised to refrain from imposing on the States inflexible constitutional restraints that could circumscribe or handicap the continued research and experimentation so vital to finding even partial solutions to educational problems and to keeping abreast of ever-changing conditions.

Page 44

          It must be remembered, also, that every claim arising under the Equal Protection Clause has implications for the relationship between national and state power under our federal system. Questions of federalism are always inherent in the process of determining whether a State's laws are to be accorded the traditional presumption of constitutionality, or are to be subjected instead to rigorous judicial scrutiny. While '(t)he maintenance of the principles of federalism is a foremost consideration in interpreting any of the pertinent constitutional provisions under which this Court examines state action,'⁸⁸ it would be difficult to imagine a case having a greater potential impact on our federal system than the one now before us, in which we are urged to abrogate systems of financing public education presently in existence in virtually every State.

          The foregoing considerations buttress our conclusion that Texas' system of public school finance is an inappropriate candidate for strict judicial scrutiny. These same considerations are relevant to the determination whether that system, with its conceded imperfections, nevertheles bears some rational relationship to a legitimate state purpose. It is to this question that we next turn our attention.

          The basic contours of the Texas school finance system have been traced at the outset of this opinion. We will now describe in more detail that system and how it operates, as these facts bear directly upon the demands of the Equal Protection Clause.

          Apart from federal assistance, each Texas school receives its funds from the State and from its local school

Page 45

district. On a statewide average, a roughly comparable amount of funds is derived from each source.⁸⁹ The State's contribution, under the Minimum Foundation Program, was designed to provide an adequate minimum educational offering in every school in the State. Funds are distributed to assure that there will be one teacher—compensated at the statesupported minimum salary—for every 25 students.⁹⁰ Each school district's other supportive personnel are provided for: one principal for every 30 teachers;⁹¹ one 'special service' teacher—librarian, nurse, doctor, etc.—for every 20 teachers;⁹² superintendents, vocational instructors, counselors, and educators for exceptional children are also provided.⁹³ Additional funds are earmarked for current operating expenses, for student transportation,⁹⁴ and for free textbooks.⁹⁵

          The program is administered by the State Board of Education and by the Central Education Agency, which also have responsibility for school accreditation⁹⁶ and for monitoring the statutory teacher-qualification standards.⁹⁷ As reflected by the 62% increase in funds allotted to the Edgewood School District over the last three years,⁹⁸ the State's financial contribution to education is steadily increasing. None of Texas' school districts, how-

Page 46

ever, has been content to rely alone on funds from the Foundation Program.

          By virtue of the obligation to fulfill its Local Fund Assignment, every district must impose an ad valorem tax on property located within its borders. The Fund Assignment was designed to remain sufficiently low to assure that each district would have some ability to provide a more enriched educational program.⁹⁹ Every district supplements its Foundation grant in this manner. In some districts, the local property tax contribution is insubstantial, as in Edgewood where the supplement was only $26 per pupil in 1967. In other districts, the local share may far exceed even the total Foundation grant. In part, local differences are attributable to differences in the rates of taxation or in the degree to which the market value for any category of property varies from its assessed value.¹⁰⁰ The greatest interdistrict disparities, however, are attributable to differences in the amount of assessable property available within any district. Those districts that have more property, or more valuable property, have a greater capability for supplementing state funds. In large measure, these additional local revenues are devoted to paying higher salaries to more teachers. Therefore, the primary distinguishing attributes of schools in property-affluent districts are lower pupil-teacher ratios and higher salary schedules.¹⁰¹

Page 47

                    This, then, is the basic outline of the Texas school financing structure. Because of differences in expenditure levels occasioned by disparities in property tax income, appellees claim that children in less affluent districts have been made the subject of invidious discrimination. The District Court found that the State had failed even 'to establish a reasonable basis' for a system that results in different levels of per-pupil expenditure. 337 F.Supp., at 284. We disagree.

          In its reliance on state as well as local resources, the Texas system is comparable to the systems employed

Page 48

in virtually every other State.¹⁰² The power to tax local property for educational purposes has been recognized in Texas at least since 1883.¹⁰³ When the growth of commercial and industrial centers and accompanying shifts in population began to create disparities in local resources, Texas undertook a program calling for a considerable investment of state funds.

          The 'foundation grant' theory upon which Texas legislators and educators based the Gilmer-Aikin bills, was a product of the pioneering work of two New York educational reformers in the 1920's, George D. Strayer and Robert M. Haig. ¹⁰⁴ Their efforts were devoted to establishing a means of guaranteeing a minimum statewide educational program without sacrificing the vital element of local participation. The Strayer-Haig thesis

Page 49

represented an accommodation between these two competing forces. As articulated by Professor Coleman:

          'The history of education since the industrial revolution shows a continual struggle between two forces: the desire by members of society to have educational opportunity for all children, and the desire of each family to provide the best education it can afford for its own children.'¹⁰⁵

          The Texas system of school finance is responsive to these two forces. While assuring a basis education for every child in the State, it permits and encourages a large measure of participation in and control of each district's schools at the local level. In an era that has witnessed a consistent trend toward centralization of the functions of government, local sharing of responsibility for public education has survived. The merit of local control was recognized last Term in both the majority and dissenting opinions in Wright v. Council of the City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972). Mr. Justice Stewart stated there that '(d)irect control over decisions vitally affecting the education of one's children is a need that is strongly felt in our society.' Id., at 469, 92 S.Ct., at 2206. The Chief Justice, in his dissent, agreed that '(l)ocal control is not only vital to continued public support of the schools, but it is of overriding importance from an educational standpoint as well.' Id., at 478, 92 S.Ct., at 2211.

          The persistence of attachment to government at the lowest level where education is concerned reflects the depth of commitment of its supporters. In part, local control means, as Professor Coleman suggests, the freedom to devote more money to the education of one's children. Equally important, however, is the opportunity

Page 50

it offers for participation in the decisionmaking process that determines how those local tax dollars will be spent. Each locality is free to tailor local programs to local needs. Pluralism also affords some opportunity for experimentation, innovation, and a healthy competition for educational excellence. An analogy to the Nation-State relationship in our federal system seems uniquely appropriate. Mr. Justice Brandeis identified as one of the peculiar strengths of our form of government each State's freedom to 'serve as a laboratory; and try novel social and economic experiments.'¹⁰⁶ No area of social concern stands to profit more from a multiplicity of viewpoints and from a diversity of approaches than does public education.

          Appellees do not question the propriety of Texas' dedication to local control of education. To the contrary, they attack the school-financing system precisely because, in their view, it does not provide the same level of local control and fiscal flexibility in all districts. Appellees suggest that local control could be preserved and promoted under other financing systems that resulted in more equality in education expenditures. While it is no doubt true that reliance on local property taxation for school revenues provides less freedom of choice with respect to expenditures for some districts than for others,¹⁰⁷

Page 51

the existence of 'some inequality' in the manner in which the State's rationale is achieved is not alone a sufficient basis for striking down the entire system. McGowan v. Maryland, 366 U.S. 420, 425—426, 81 S.Ct. 1101, 1104—1105, 6 L.Ed.2d 393 (1961). It may not be condemned simply because it imperfectly effectuates the State's goals. Dandridge v. Williams, 397 U.S., at 485, 90 S.Ct. at 1161. Nor must the financing system fail because, as appellees suggest, other methods of satisfying the State's interest, which occasion 'less drastic' disparities in expenditures, might be conceived. Only where state action impinges on the exercise of fundamental constitutional rights or liberties must it be found to have chosen the least restrictive alternative. Cf. Dunn v. Blumstein, 405 U.S., at 343, 92 S.Ct. at 1003; Shelton v. Tucker, 364 U.S. 479, 488, 81 S.Ct. 247, 252, 5 L.Ed.2d 231 (1960). It is also well to remember that even those districts that have reduced ability to make free decisions with respect to how much they spend on education still retain under the present system a large measure of authority as to how available funds will be allocated. They further enjoy the power to make numerous other decisions with respect to the operation of the schools.¹⁰⁸ The people of Texas may be

Page 52

justified in believing that other systems of school financing, which place more of the financial responsibility in the hands of the State, will result in a comparable lessening of desired local autonomy. That is, they may believe

Page 53

that along with increased control of the purse strings at the state level will go increased control over local policies.¹⁰⁹

          Appellees further urge that the Texas system is unconstitutionally arbitrary because it allows the availability of local taxable resources to turn on 'happenstance.' They see no justification for a system that allows, as they contend, the quality of education to fluctuate on the basis of the fortuitous positioning of the boundary lines of political subdivisions and the location of valuable commercial and industrial property. But any scheme of

Page 54

local taxation—indeed the very existence of identifiable local governmental units—requires the establishment of jurisdictional boundaries that are inevitably arbitrary. It is equally inevitable that some localities are going to be blessed with more taxable assets than others.¹¹⁰ Nor is local wealth a static quantity. Changes in the level of taxable wealth within any district may result from any number of events, some of which local residents can and do influence. For instance, commercial and industrial enterprises may be encouraged to locate within a district by various actions—public and private.

          Moreover, if local taxation for local expenditures were an unconstitutional method of providing for education then it might be an equally impermissible means of providing other necessary services customarily financed largely from local property taxes, including local police and fire protection, public health and hospitals, and public utility facilities of various kinds. We perceive no justification for such a severe denigration of local property taxation and control as would follow from appellees' contentions. It has simply never been within the constitutional prerogative of this Court to nullify statewide measures for financing public services merely because the burdens or benefits thereof fall unevenly depending upon the relative wealth of the political subdivisions in which citizens live.

          In sum, to the extent that the Texas system of school financing results in unequal expenditures between chil-

Page 55

dren who happen to reside in different districts, we cannot say that such disparities are the product of a system that is so irrational as to be invidiously discriminatory. Texas has acknowledged its shortcomings and has persistently endeavored—not without some success—to ameliorate the differences in levels of expenditures without sacrificing the benefits of local participation. The Texas plan is not the result of hurried, ill-conceived legislation. It certainly is not the product of purposeful discrimination against any group or class. On the contrary, it is rooted in decades of experience in Texas and elsewhere, and in major part is the product of responsible studies by qualified people. In giving substance to the presumption of validity to which the Texas system is entitled, Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78, 31 S.Ct. 337, 340, 55 L.Ed. 369 (1911), it is important to remember that at every stage of its development it has constituted a 'rough accommodation' of interests in an effort to arrive at practical and workable solutions. Metropolis Theatre Co. v. City of Chicago, 228 U.S. 61, 69—70, 33 S.Ct. 441, 443, 57 L.Ed. 730 (1913). One also must remember that the system here challenged is not peculiar to Texas or to any other State. In its essential characteristics, the Texas plan for financing public education reflects what many educators for a half century have thought was an enlightened approach to a problem for which there is no perfect solution. We are unwilling to assume for ourselves a level of wisdom superior to that of legislators, scholars, and educational authorities in 50 States, especially where the alternatives proposed are only recently conceived and nowhere yet tested. The constitutional standard under the Equal Protection Clause is whether the challenged state action rationally furthers a legitimate state purpose or interest. McGinnis v. Royster, 410 U.S. 263, 270, 93 S.Ct. 1055, 1059, 35 L.Ed.2d 282 (1973). We hold that the Texas plan abundantly satisfies this standard.

Page 56

          In light of the considerable attention that has focused on the District Court opinion in this case and on its California predecessor, Serrano v. Priest, 5 Cal.3d 584, 96 Cal.Rptr. 601, 487 P.2d 1241 (1971), a cautionary postscript seems appropriate. It cannot be questioned that the constitutional judgment reached by the District Court and approved by our dissenting Brothers today would occasion in Texas and elsewhere an unprecedented upheaval in public education. Some commentators have concluded that, whatever the contours of the alternative financing programs that might be devised and approved, the result could not avoid being a beneficial one. But, just as there is nothing simple about the constitutional issues involved in these cases, there is nothing simple or certain about predicting the consequences of massive change in the financing and control of public education. Those who have devoted the most thoughtful attention to the practical ramifications of these cases have found no clear or dependable answers and their scholarship reflects no such unqualified confidence in the desirability of completely uprooting the existing system.

          The complexity of these problems is demonstrated by the lack of consensus with respect to whether it may be said with any assurance that the poor, the racial minorities, or the children in over-burdened core-city school districts would be benefited by abrogation of traditional modes of financing education. Unless there is to be a substantial increase in state expenditures on education across the board—an event the likelihood of which is open to considerable question ¹¹¹—these groups stand to

Page 57

realize gains in terms of increased per-pupil expenditures only if they reside in districts that presently spend at relatively low levels, i.e., in those districts that would benefit from the redistribution of existing resources. Yet, recent studies have indicated that the poorest families are not invariably clustered in the most impecunious school districts.¹¹² Nor does it now appear that there is any more than a random chance that racial minorities are concentrated in property-poor districts.¹¹³ Additionally,

Page 58

several research projects have concluded that any financing alternative designed to achieve a greater equality of expenditures is likely to lead to higher taxation and lower educational expenditures in the major urban centers,¹¹⁴ a result that would exacerbate rather than ameliorate existing conditions in those areas.

          These practical considerations, of course, play no role in the adjudication of the constitutional issues presented here. But they serve to highlight the wisdom of the traditional limitations on this Court's function. The consideration and initiation of fundamental reforms with respect to state taxation and education are matters reserved for the legislative processes of the various States, and we do no violence to the values of federalism and separation of powers by staying our hand. We hardly need add that this Court's action today is not to be viewed as placing its judicial imprimatur on the status quo. The need is apparent for reform in tax systems which may well have relied too long and too heavily on the local property tax. And certainly innovative thinking as to public education, its methods, and its funding is necessary to assure both a higher level of quality and greater uniformity of opportunity. These matters merit the continued attention of the scholars who already

Page 59

have contributed much by their challenges. But the ultimate solutions must come from the lawmakers and from the democractic pressures of those who elect them.

          Reversed.

           Mr. Justice STEWART, concurring.

          The method of financing public schools in Texas, as in almost every other State, has resulted in a system of public education that can fairly be described as chaotic and unjust.¹ It does not follow, however, and I cannot find, that this system violates the Constitution of the United States. I join the opinion and judgment of the Court because I am convinced that any other course would mark an extraordinary departure from principled adjudication under the Equal Protection Clause of the Fourteenth Amendment. The unchartered directions of such a departure are suggested, I think, by the imaginative dissenting opinion my Brother MARSHALL has filed today.

          Unlike other provisions of the Constitution, the Equal Protection Clause confers no substantive rights and creates no substantive liberties.² The function of the Equal Protection Clause, rather, is simply to measure the validity of classifications created by state laws.

Page 60

          There is hardly a law on the books that does not affect some people differently from others. But the basic concern of the Equal Protection Clause is with state legislation whose purpose or effect is to create discrete and objectively identifiable classes.³ And with respect to such legislation, it has long been settled that the Equal Protection Clause is offended only by laws that are invidiously discriminatory—only by classifications that are wholly arbitrary or capricious. See, e.g., Rinaldi v. Yeager, 384 U.S. 305, 86 S.Ct. 1497, 16 L.Ed.2d 577. This settled principle of constitutional law was compendiously stated in Mr. Chief Justice Warren's opinion for the Court in McGowan v. Maryland, 366 U.S. 420, 425—426, 81 S.Ct. 1101, 1105, 6 L.Ed.2d 393, in the following words:

          'Although no precise formula has been developed, the Court has held that the Fourteenth Amendment permits the States a wide scope of discretion in enacting laws which affect some groups of citizens differently than others. The constitutional safeguard is offended only if the classification rests on grounds wholly irrelevant to the achievement of the State's objective. State legislatures are presumed to have acted within their constitutional power despite the fact that, in practice, their laws result in some inequality. A statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it.'

          This doctrine is no more than a specific application of one of the first principles of constitutional adjudication—the basic presumption of the constitutional validity of a duly enacted state or federal law. See Thayer, The Origin and Scope of the American Doctrine of Constitutional Law, 7 Harv.L.Rev. 129 (1893).

Page 61

          Under the Equal Protection Clause, this presumption of constitutional validity disappears when a State has enacted legislation whose purpose or effect is to create classes based upon criteria that, in a constitutional sense, are inherently 'suspect.' Because of the historic purpose of the Fourteenth Amendment, the prime example of such a 'suspect' classification is one that is based upon race. See, e.g., Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873; McLaughlin v. Florida, 379 U.S. 184, 85 S.Ct. 283, 13 L.Ed.2d 222. But there are other classifications that, at least in some settings, are also 'suspect'—for example, those based upon national origin, ⁴ alienage,⁵ indigency,⁶ or illegitimacy.⁷

          Moreover, quite apart from the Equal Protection Clause, a state law that impinges upon a substantive right or liberty created or conferred by the Constitution is, of course, presumptively invalid, whether or not the law's purpose or effect is to create any classifications. For example, a law that provided that newspapers could be published only by people who had resided in the State for five years could be superficially viewed as invidiously discriminating against an identifiable class in violation of the Equal Protection Clause. But, more basically, scuch a law would be invalid simply because it abridged the freedom of the press. Numerous cases in this Court illustrate this principle.⁸

Page 62

          In refusing to invalidate the Texas system of financing its public schools, the Court today applies with thoughtfulness and understanding the basic principles I have so sketchily summarized. First, as the Court points out, the Texas system has hardly created the kind of objectively identifiable classes that are cognizable under the Equal Protection Clause.⁹ Second, even assuming the existence of such discernible categories, the classifications are in no sense based upon constitutionally 'suspect' criteria. Third, the Texas system does not rest 'on grounds wholly irrelevant to the achievement of the State's objective.' Finally, the Texas system impinges upon no substantive constitutional rights or liberties. It follows, therefore, under the established principle reaffirmed in Mr. Chief Justice Warren's opinion for the Court in McGowan v. Maryland, supra, that the judgment of the District Court must be reversed.

           Mr. Justice BRENNAN, dissenting.

          Although I agree with my Brother WHITE that the Texas statutory scheme is devoid of any rational basis, and for that reason is violative of the Equal Protection Clause, I also record my disagreement with the Court's rather distressing assertion that a right may be deemed 'fundamental' for the purposes of equal protection analysis only if it is 'explicitly or implicitly guaranteed by the Constitution.' Ante, at 33—34. As my Brother MARSHALL convincingly demonstrates, our prior cases stand for the proposition that 'fundamentality' is, in large measure, a function of the right's importance in terms of the effectuation of those rights which are in fact constitutionally guaranteed. Thus, '(a)s the nexus between the specific constitutional guarantee and the nonconstitutional interest draws closer, the non-

Page 63

constitutional interest becomes more fundamental and the degree of judicial scrutiny applied when the interest is infringed on a discriminatory basis must be adjusted accordingly.' Post, at 102 103.

          Here, there can be no doubt that education is inextricably linked to the right to participate in the electoral process and to the rights of free speech and association guaranteed by the First Amendment. See post, at 111—115. This being so, any classification affecting education must be subjected to strict judicial scrutiny, and since even the State concedes that the statutory scheme now before us cannot pass constitutional muster under this stricter standard of review, I can only conclude that the Texas school-financing scheme is constitutionally invalid.

           Mr. Justice WHITE, with whom Mr. Justice DOUGLAS and Mr. Justice BRENNAN join, dissenting.

          The Texas public schools are financed through a combination of state funding, local property tax revenue, and some federal funds.¹ Concededly, the system yields wide disparity in per-pupil revenue among the various districts. In a typical year, for example, the Alamo Heights district had total revenues of $594 per pupil, while the Edgewood district had only $356 per pupil.² The majority and the State concede, as they must, the existence

Page 64

of major disparities in spendable funds. But the State contends that the disparities do not invidiously discriminate against children and families in districts such as Edgewood, because the Texas scheme is designed 'to provide an adequate education for all, with local autonomy to go beyond that as individual school districts desire and are able . . .. It leaves to the people of each district the choice whether to go beyond the minimum and, if so, by how much.'³ The majority advances this rationalization: 'While assuring a basic education for every child in the State, it permits and encourages a large measure of participation in and control of each district's schools at the local level.'

          I cannot disagree with the proposition that local control and local decisionmaking play an important part in our democratic system of government. Cf. James v. Valtierra, 402 U.S. 137, 91 S.Ct. 1331, 28 L.Ed.2d 678 (1971). Much may be left to local option, and this case would be quite different if it were true that the Texas system, while insuring minimum educational expenditures in every district through state funding, extended a meaningful option to all local districts to increase their per-pupil expenditures and so to improve their children's education to the extent that increased funding would achieve that goal. The system would then arguably provide a rational and sensible method of achieving the stated aim of preserving an area for local initiative and decision.

          The difficulty with the Texas system, however, is that it provides a meaningful option to Alamo Heights and like school districts but almost none to Edgewood and those other districts with a low per-pupil real estate tax base. In these latter districts, no matter how desirous parents are of supporting their schools with greater revenues, it is impossible to do so through the use of the

Page 65

real estate property tax. In these districts, the Texas system utterly fails to extend a realistic choice to parents because the property tax, which is the only revenue-raising mechanism extended to school districts, is practically and legally unavailable. That this is the situation may be readily demonstrated.

          Local school districts in Texas raise their portion of the Foundation School Program—the Local Fund Assignment—by levying ad valorem taxes on the property located within their boundaries. In addition, the districts are authorized, by the state constitution and by statute, to levy ad valorem property taxes in order to raise revenues to support educational spending over and above the expenditure of Foundation School Program funds.

          Both the Edgewood and Alamo Heights districts are located in Bexar County, Texas. Student enrollment in Alamo Heights is 5,432, in Edgewood 22,862. The per-pupil market value of the taxable property in Alamo Heights is $49,078, in Edgewood $5,960. In a typical relevant year, Alamo Heights had a maintenance tax rate of $1.20 and a debt service (bond) tax rate of 20¢ per $100 assessed evaluation, while Edgewood had a maintenance rate of 52¢ and a bond rate of 67¢. These rates, when applied to the respective tax bases, yielded Alamo Heights $1,433,473 in maintenance dollars and $236,074 in bond dollars, and Edgewood $223,034 in maintenance dollars and $279,023 in bond dollars. As is readily apparent, because of the variance in tax bases between the districts, results, in terms of revenues, do not correlate with effort, in terms of tax rate. Thus, Alamo Heights, with a tax base approximately twice the size of Edgewood's base, realized approximately six times as many maintenance dollars as Edgewood by using a tax rate only approximately two and one-half times larger. Similarly, Alamo Heights realized slightly fewer bond

Page 66

dollars by using a bond tax rate less than one-third of that used by Edgewood.

          Nor is Edgewood's revenue-raising potential only deficient when compared with Alamo Heights. North East District has taxable property with a per-pupil market value of approximately $31,000, but total taxable property approximately four and one-half times that of Edgewood. Applying a maintenance rate of $1, North East yielded $2,818,148. Thus, because of its superior tax base, North East was able to apply a tax rate slightly less than twice that applied by Edgewood and yield more than 10 times the maintenance dollars. Similarly, North East, with a bond rate of 45¢, yielded $1,249,159—more than four times Edgewood's yield with two-thirds the rate.

          Plainly, were Alamo Heights or North East to apply the Edgewood tax rate to its tax base, it would yield far greater revenues than Edgewood is able to yield applying those same rates to its base. Conversely, were Edgewood to apply the Alamo Heights or North East rates to its base, the yield would be far smaller than the Alamo Heights or North East yields. The disparity is, therefore, currently operative and its impact on Edgewood is undeniably serious. It is evident from statistics in the record that show that, applying an equalized tax rate of 85¢ per $100 assessed valuation, Alamo Heights was able to provide approximately $330 per pupil in local revenues over and above the Local Fund Assignment. In Edgewood, on the other hand, with an equalized tax rate of $1.05 per $100 of assessed valuation, $26 per pupil was raised beyond the Local Fund Assignment.⁴ As previously noted in Alamo Heights,

Page 67

total per-pupil revenues from local, state, and federal funds was $594 per pupil, in Edgewood $356.⁵

          In order to equal the highest yield in any other Bexar County district, Alamo Heights would be required to tax at the rate of 68 per $100 of assessed valuation. Edgewood would be required to tax at the prohibitive rate of $5.76 per $100. But state law places a $1.50 per $100 ceiling on the maintenance tax rate, a limit that would surely be reached long before Edgewood attained an equal yield. Edgewood is thus precluded in law, as well as in fact, from achieving a yield even close to that of some other districts.

          The Equal Protection Clause permits discriminations between classes but requires that the classification bear some rational relationship to a permissible object sought to be attained by the statute. It is not enough that the Taxas system before us seeks to achieve the valid, rational purpose of maximizing local initiative; the means chosen by the State must also be rationally ralated to the end sought to be achieved. As the Court stated just lat Term in Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 172, 92 S.Ct. 1400, 1405, 31 L.Ed.2d 768 (1972):

          'The tests to determine the validity of state statutes under the Equal Protection Clause have been variously expressed, but this Court requires, at a minimum, that a statutory classification bear some rational relationship to a legitimate state purpose. Morey v. Doud, 354 U.S. 457, 77 S.Ct. 1344, 1 L.Ed.2d 1485 (1957); Williamson v. Lee Optical Co., 348 U.S. 483, 75 S.Ct. 461, 99 L.Ed. 563 (1955); Gulf Colorado & Santa Fe Ry. v. Ellis, 165 U.S. 150, 17 S.Ct. 255, 41 L.Ed. 666 (1897); Yick Wo v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064, 30 L.Ed. 220 (1886).'

Page 68

          Neither Taxas nor the majority heeds this rule. If the State aims at maximizing local initiative and local choice, by permitting school districts to resort to the real property tax if they choose to do so, it utterly fails in achieving its purpose in districts with property tax bases so low that there is little if any opportunity for interested parents, rich or poor, to augment school district revenues. Requiring the State to establish only that unequal treatment is in furtherance of a permissible goal, without also requiring the State to show that the means chosen to effectuate that goal are rationally related to its achievement, makes equal protection analysis no more than an empty gesture.⁶ In my view, the parents and children in Edgewood, and in like districts, suffer from an invidious discrimination violative of the Equal Protection Clause.

          This does not, of course, mean that local control may not be a legitimate goal of a school-financing system. Nor does it mean that the State must guarantee each district an equal per-pupil revenue from the state school-financing system. Nor does it mean, as the majority appears to believe, that, by affirming the decision below,

Page 69

this Court would be 'imposing on the States inflexible constitutionl restraints that could circumscribe or handicap the continued research and experimentation so vital to finding even partial solutions to educational problems and to keeping abreast of ever-changing conditions.' On the contrary, it would merely mean that the State must fashion a financing scheme which provides a rational basis for the maximization of local control, if local control is to remain a goal of the system, and not a scheme with 'different treatment be(ing) accorded to persons placed by a statute into different classes on the basis of criteria wholly unrelated to the objective of that statute.' Reed v. Reed, 404 U.S. 71, 75—76, 92 S.Ct. 251, 254, 30 L.Ed.2d 225 (1971).

          Perhaps the majority believes that the major disparity in revenues provided and permitted by the Texas system is inconsequential. I cannot agree, however, that the difference of the magnitude appearing in this case can sensibly be ignored, particularly since the State itself considers it so important to provide opportunities to exceed the minimum state educational expenditures.

          There is no difficulty in identifying the class that is subject to the alleged discrimination and that is entitled to the benefits of the Equal Protection Clause. I need go no further than the parents and children in the Edgewood district, who are plaintiffs here and who assert that they are entitled to the same choice as Alamo Heights to augment local expenditures for schools but are denied that choice by state law. This group constitutes a class sufficiently definite to invoke the protection of the Constitution. They are as entitled to the protection of the Equal Protection Clause as were the voters in allegedly underrepresented counties in the reapportionment cases. See, e.g., Baker v. Carr, 369 U.S. 186, 204—208, 82 S.Ct. 691, 703—705, 7 L.Ed.2d 663 (1962); Gray v. Sanders, 372 U.S. 368, 375, 83 S.Ct. 801, 805, 9 L.Ed.2d 821 (1963); Reynolds v. Sims, 377 U.S. 533, 554—556, 84 S.Ct. 1362, 1377—1379, 12 L.Ed. 506 (1964). And in Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972), where a challenge to the

Page 70

Texas candidate filing fee on equal protection grounds was upheld, we noted that the victims of alleged discrimination wrought by the filing fee 'cannot be described by reference to discrete and precisely defined segments of the community as is typical of inequities challenged under the Equal Protection Clause,' but concluded that 'we would ignore reality were we not to recognize that this system falls with unequal weight on voters, as well as candidates, according to their economic status.' Id., at 144, 92 S.Ct., at 856. Similarly, in the present case we would blink reality to ignore the fact that school districts, and students in the end, are differentially affected by the Texas school-financing scheme with respect to their capability to supplement the Minimum Foundation School Program. At the very least, the law discriminates against those children and their parents who live in districts where the per-pupil tax base is sufficiently low to make impossible the provision of comparable school revenues by resort to the real property tax which is the only device the State extends for this purpose.

           Mr. Justice MARSHALL, with whom Mr. Justice DOUGLAS concurs, dissenting.

          The Court today decides, in effect, that a State may constitutionally vary the quality of education which it offers its children in accordance with the amount of taxable wealth located in the school districts within which they reside. The majority's decision represents an abrupt departure from the mainstream of recent state and federal court decisions concerning the unconstitutionality of state educational financing schemes dependent upon taxable local wealth.¹ More unfortunately, though, the

Page 71

majority's holding can only be seen as a retreat from our historic commitment to equality of educational opportunity and as unsupportable acquiescence in a system which deprives children in their earliest years of the chance to reach their full potential as citizens. The Court does this despite the absence of any substantial justification for a scheme which arbitrarily channels educational resources in accordance with the fortuity of the amount of taxable wealth within each district.

          In my judgment, the right of every American to an equal start in life, so far as the provision of a state service as important as education is concerned, is far too vital to permit state discrimination on grounds as tenuous as those presented by this record. Nor can I accept the notion that it is sufficient to remit these appellees to the vagaries of the political process which, contrary to the majority's suggestion, has proved singularly unsuited to the task of providing a remedy for this discrimination.² I, for one, am unsatisfied with the hope of an ultimate 'political' solution sometime in the indefinite future while, in the meantime, countless children unjustifiably receive inferior educations that may affect their hearts

Page 72

and minds in a way unlikely ever to be undone.' Brown v. Board of Education, 347 U.S. 483, 494, 74 S.Ct. 686, 691, 98 l.Ed. 873 (1954). I must therefore respectfully dissent.

          The Court acknowledges that 'substantial interdistrict disparities in school expenditures' exist in Texas, ante, at 15, and that these disparities are 'largely attributable to differences in the amounts of money collected through local property taxation,' ante, at 16. But instead of closely examining the seriousness of these disparities and the invidiousness of the Texas financing scheme, the Court undertakes an elaborate exploration of the efforts Texas has purportedly made to close the gaps between its districts in terms of levels of district wealth and resulting educational funding. Yet, however praiseworthy Texas' equalizing efforts, the issue in this case is not whether Texas is doing its best to ameliorate the worst features of a discriminatory scheme but, rather, whether the scheme itself is in fact unconstitutionally discriminatory in the face of the Fourteenth Amendment's guarantee of equal protection of the laws. When the Texas financing scheme is taken as a whole, I do not think it can be doubted that it produces a discriminatory impact on substantial numbers of the schoolage children of the State of Texas.

          Funds to support public education in Texas are derived from three sources: local ad valorem property taxes; the Federal Government; and the state government.³ It is enlightening to consider these in order.

Page 73

          Under Texas law, the only mechanism provided the local school district for raising new, unencumbered revenues is the power to tax property located within its boundaries.⁴ At the same time, the Texas financing scheme effectively restricts the use of monies raised by local property taxation to the support of public education within the boundaries of the district in which they are raised, since any such taxes must be approved by a majority of the property-taxpaying voters of the district.⁵

          The significance of the local property tax element of the Texas financing scheme is apparent from the fact that it provides the funds to meet some 40% of the cost of public education for Texas as a whole.⁶ Yet the amount of revenue that any particular Texas district can raise is dependent on two factors—its tax rate and its amount of taxable property. The first factor is determined by the property-taxpaying voters of the district.⁷ But, regardless of the enthusiasm of the local voters for public

Page 74

education, the second factor—the taxable property wealth of the district—necessarily restricts the district's ability to raise funds to support public education. ⁸ Thus, even though the voters of two Texas districts may be willing to make the same tax effort, the results for the districts will be substantially different if one is property rich while the other is property poor. The necessary effect of the Texas local property tax is, in short, to favor property-rich districts and to disfavor property-poor ones.

          The seriously disparate consequences of the Texas local property tax, when that tax is considered alone, are amply illustrated by data presented to the District Court by appellees. These data included a detailed study of a sample of 110 Texas school districts⁹ for the 1967—1968 school year conducted by Professor Joel S. Berke of Syracuse University's Educational Finance Policy Institute. Among other things, this study revealed that the 10 richest districts examined, each of which had more than $100,000 in taxable property per pupil, raised through local effort an average of $610 per pupil, whereas the four poorest districts studied, each of which had less than $10,000 in taxable property per pupil, were able

Page 75

to raise only an average of $63 per pupil.¹⁰ And, as the Court effectively recognizes, ante, at 27, this correlation between the amount of taxable property per pupil and the amount of local revenues per pupil holds true for the 96 districts in between the richest and poorest districts.¹¹

          It is clear, moreover, that the disparity of per-pupil revenues cannot be dismissed as the result of lack of local effort that is, lower tax rates—by property-poor districts. To the contrary, the data presented below indicate that the poorest districts tend to have the highest tax rates and the richest districts tend to have the lowest tax rates.¹² Yet, despite the apparent extra effort being made by the poorest districts, they are unable even to begin to match the richest districts in terms of the production of local revenues. For example, the 10 richest districts studied by Professor Berke were able to produce $585 per pupil with an equalized tax rate of 31¢

Page 76

on $100 of equalized valuation, but the four poorest districts studied, with an equalized rate of 70¢ on $100 of equalized valuation, were able to produce only $60 per pupil. ¹³ Without more, this stateimposed system of educational funding presents a serious picture of widely varying treatment of Texas school districts, and thereby of Texas schoolchildren, in terms of the amount of funds available for public education.

          Nor are these funding variations corrected by the other aspects of the Texas financing scheme. The Federal Government provides funds sufficient to cover only some 10% of the total cost of public education in Texas.¹⁴ Furthermore, while these federal funds are not distributed in Texas solely on a per-pupil basis, appellants do not here contend that they are used in such a way as to ameliorate signiticantly the widely varying consequences for Texas school districts and schoolchildren of the local property tax element of the state financing scheme.¹⁵

          State funds provide the remaining some 50% of the monies spent on public education in Texas.¹⁶ Technically, they are distributed under two programs. The first is the Available School Fund, for which provision is made in the Texas Constitution.¹⁷ The Available

Page 77

School Fund is composed of revenues obtained from a number of sources, including receipts from the state ad valorem property tax, one-fourth of all monies collected by the occupation tax, annual contributions by the legislature from general revenues, and the revenues derived from the Permanent School Fund.¹⁸ For the 1970—1971 school year the Available School Fund contained $296,000,000. The Texas Constitution requires that this money be distributed annually on a per capita basis¹⁹ to the local school districts. Obviously, such a flat grant could not alone eradicate the funding differentials atrributable to the local property tax. Moreover, today the Available School Fund is in reality simply one facet of the second state financing program, the Minimum Foundation School Program,²⁰ since each district's annual share of the Fund is deducted from the sum to which the district is entitled under the Foundation Program.²¹

          The Minimum Foundation School Program provides funds for three specific purposes: professional salaries, current operating expenses, and transportation expenses.²² The State pays, on an overall basis, for approximately 80% of the cost of the Program; the remaining 20% is distributed among the local school districts under the

Page 78

Local Fund Assignment.²³ Each district's share of the Local Fund Assignment is determined by a complex 'economic index' which is designed to allocate a larger share of the costs to property-rich districts than to property-poor districts.²⁴ Each district pays its share with revenues derived from local property taxation.

          The stated purpose of the Minimum Foundation School Program is to provide certain basic funding for each local Texas school district.²⁵ At the same time, the Program was apparently intended to improve, to some degree, the financial position of property-poor districts relative to property-rich districts, since through the use of the economic index—an effort is made to charge a disproportionate share of the costs of the Program to rich districts. ²⁶ It bears noting, however, that substantial criticism has been leveled at the practical effectiveness of the economic index system of local cost allocation.²⁷ In theory, the index is designed to ascertain the relative ability of each district to contribute to the Local Fund Assignment from local property taxes. Yet the index is not developed simply on the basis of each district's taxable wealth. It also takes into account the district's relative income from manufacturing, mining, and agriculture, its payrolls, and its scholastic population.²⁸

Page 79

It is difficult to discern precisely how these latter factors are predictive of a district's relative ability to raise revenues through local property taxes. Thus, in 1966, one of the consultants who originally participated in the development of the Texas economic index adopted in 1949 told the Governor's Committee on Public School Education: 'The Economic Index approach to evaluating local ability offers a little better measure than sheer chance, but not much.'²⁹

          Moreover, even putting aside these criticisms of the economic index as a device for achieving meaningful district wealth equalization through cost allocation, poor districts still do not necessarily receive more state aid than property-rich districts. For the standards which currently determine the amount received from the Foundation School Program by any particular district ³⁰ favor property-rich districts.³¹ Thus, focusing on the same

Page 80

Edgewood Independent and Alamo Heights School Districts which the majority uses for purposes of illustration, we find that in 1967 1968 property-rich Alamo Heights,³² which raised $333 per pupil on an equalized tax rate of 85¢ per $100 valuation, received $225 per pupil from the Foundation School Program, while property-poor Edgewood,³³ which raised only $26 per pupil with an equalized tax rate of $1.05 per $100 valuation, received only $222 per pupil from the Foundation School Program.³⁴ And, more recent data, which indicate that for the 1970—1971 school year Alamo Heights received $491 per pupil from

Page 81

the Program while Edgewood received only $356 per pupil, hardly suggest that the wealth gap between the districts is being narrowed by the State Program. To the contrary, whereas in 1967 1968 Alamo Heights received only $3 per pupil, or about 1%, more than Edgewood in state aid, by 1970—1971 the gap had widened to a difference of $135 per pupil, or about 38%.³⁵ It was data of this character that prompted the District Court to observe that 'the current (state aid) system tends to subsidize the rich at the expense of the poor, rather than the other way around.'³⁶ 337 F.Supp. 280, 282. And even the appellants go no further here than to venture that the Minimum Foundation School Program has 'a mildly equalizing effect.'³⁷

          Despite these facts, the majority continually emphasized how much state aid has, in recent years, been given

Page 82

to property-poor Texas school districts. What the Court fails to emphasize is the cruel irony of how much more state aid is being given to property-rich Texas school districts on top of their already substantial local property tax revenues.³⁸ Under any view, then, it is apparent that the state aid provided by the Foundation School Program fails to compensate for the large funding variations attributable to the local property tax element of the Texas financing scheme. And it is these stark differences in the treatment of Texas school districts and school children inherent in the Texas financing schement, not the absolute amount of state aid provided to any particular school district, that are the crux of this case. There can, moreover, be no escaping the conclusion that the local property tax which is dependent upon taxable district property wealth is an essential feature of the Texas scheme for financing public education.³⁹

          The appellants do not deny the disparities in educational funding caused by variations in taxable district property wealth. They do contend, however, that whatever the differences in per-pupil spending among Texas districts, there are no discriminatory consequences for the children of the disadvantaged districts. They recognize that what is at stake in this case is the quality of the

Page 83

public education provided Texas children in the districts in which they live. But appellants reject the suggestion that the quality of education in any particular district is determined by money beyond some minimal level of funding which they believe to be assured every Texas district by the Minimum Foundation School Program. In their view, there is simply no denial of equal educational opportunity to any Texas school children as a result of the widely varying per-pupil spending power provided districts under the current financing scheme.

          In my view, though, even an unadorned restatement of this contention is sufficient to reveal its absurdity. Authorities concerned with educational quality no doubt disagree as to the significance of variations in per-pupil spending.⁴⁰ Indeed, conflicting expert testimony was presented to the District Court in this case concerning the effect of spending variations on educational achievement.⁴¹ We sit, however, not to resolve disputes over educational theory but to enforce our Constitution. It is an inescapable fact that if one district has more funds available per pupil than another district, the

Page 84

former will have greater choice in educational planning than will the latter. In this regard, I believe the question of discrimination in educational quality must be deemed to be an objective one that looks to what the State provides its children, not to what the children are able to do with what they receive. That a child forced to attend an underfunded school with poorer physical facilities, less experienced teachers, larger classes, and a narrower range of courses than a school with substantially more funds—and thus with greater choice in educational planning may nevertheless excel is to the credit of the child, not the State, cf. Missouri ex rel. Gaines v. Canada, 305 U.S. 337, 349, 59 S.Ct. 232, 236, 86 L.Ed. 208 (1938). Indeed, who can ever measure for such a child the opportuntiies lost and the talents wasted for want of a broader, more enriched education? Discrimination in the opportunity to learn that is afforded a child must be our standard.

          Hence, even before this Court recognized its duty to tear down the barriers of state-enforced racial segregation in public education, it acknowledged that inequality in the educational facilities provided to students may be discriminatory state action as contemplated by the Equal Protection Clause. As a basis for striking down state-enforced segregation of a law school, the Court in Sweatt v. Painter, 339 U.S. 629, 633—634, 70 S.Ct. 848, 850, 94 L.Ed. 1114 (1950), stated:

          '(W)e cannot find substantial equality in the educational opportunities offered white and Negro law students by the State. In terms of number of the faculty, variety of courses and opportunity for specialization, size of the student body, scope of the library, availability of law review and similar activities, the (whites only) Law School is superior. . . . It is difficult to believe that one who had a free choice between these law schools would consider the question close.'

Page 85

          See also McLaurin v. Oklahoma State Regents for Higher Education, 339 U.S. 637, 70 S.Ct. 851, 94 L.Ed. 1149 (1950). Likewise, it is difficult to believe that if the children of Texas had a free choice, they would choose to be educated in districts with fewer resources, and hence with more antiquated plants, less experienced teachers, and a less diversified curriculum. In fact, if financing variations are so insignificant to educational quality, it is difficult to understand why a number of our country's wealthiest school districts, which have no legal obligation to argue in support of the constitutionality of the Texas legislation, have nevertheless zealously pursued its cause before this Court.⁴²

          The consequences, in terms of objective educational input, of the variations in district funding caused by the Texas financing scheme are apparent from the data introduced before the District Court. For example, in 1968—1969, 100% of the teachers in the property-rich Alamo Heights School District had college degrees.⁴³ By contrast, during the same school year only 80.02% of the teachers had college degrees in the property poor Edgewood Independent School District.⁴⁴ Also, in 1968—1969, approximately 47% of the teachers in the Edgewood District were on emergency teaching permits, whereas only 11% of the teachers in Alamo Heights were on such permits.⁴⁵ This is undoubtedly a reflection of the fact that the top of Edgewood's teacher salary scale was

Page 86

approximately 80% of Alamo Heights.⁴⁶ And, not surprisingly, the teacher-student ratio varies significantly between the two districts.⁴⁷ In other words, as might be expected, a difference in the funds available to districts results in a difference in educational inputs available for a child's public education in Texas. For constitutional purposes, I believe this situation, which is directly attributable to the Texas financing scheme, raises a grave question of state-created discrimination in the provision of public education. Cf. Gaston County v. United States, 395 U.S. 285, 293—294, 89 S.Ct. 1720, 1724—1725, 23 L.Ed.2d 309 (1969).

          At the very least, in view of the substantial interdistrict disparities in funding and in resulting educational inputs shown by appellees to exist under the Texas financing scheme, the burden of proving that these disparities do not in fact affect the quality of children's education must fall upon the appellants. Cf. Hobson v. Hansen, 327 F.Supp. 844, 860—861 (D.C.D.C.1971). Yet appellants made no effort in the District Court to demonstrate that educational quality is not affected by variations in funding and in resulting inputs. And, in this Court, they have argued no more than that the relationship is ambiguous. This is hardly sufficient to overcome appellees' prima facie showing of state-created discrimination between the schoolchildren of Texas with respect to objective educational opportunity.

          Nor can I accept the appellants' apparent suggestion that the Texas Minimum Foundation School Program effectively eradicates any discriminatory effects otherwise resulting from the local property tax element of the

Page 87

Texas financing scheme. Appellants assert that, despite its imperfections, the Program 'does guarantee an adequate education to every child.'⁴⁸ The majority, in considering the constitutionality of the Texas financing scheme, seems to find substantial merit in this contention, for it tells us that the Foundation Program 'was designed to provide an adequate minimum educational offering in every school in the State,' ante, at 45, and that the Program 'assur(es) a basic education for every child,' ante, at 49. But I fail to understand how the constitutional problems inherent in the financing scheme are eased by the Foundation Program. Indeed, the precise thrust of the appellants' and the Court's remarks are not altogether clear to me.

          The suggestion may be that the state aid received via the Foundation Program sufficiently improves the position of property-poor districts vis-a-vis property-rich districts—in terms of educational funds—to eliminate any claim of interdistrict discrimination in available educational resources which might otherwise exist if educational funding were dependent solely upon local property taxation. Certainly the Court has recognized that to demand precise equality of treatment is normally unrealistic, and thus minor differences inherent in any practical context usually will not make out a substantial equal protection claim. See, e.g., Mayer v. City of Chicago, 404 U.S. 189, 194—195, 92 S.Ct. 410, 414—415, 30 L.Ed.2d 372 (1971); Draper v. Washington, 372 U.S. 487, 495—496, 83 S.Ct. 774, 778—779, 9 L.Ed.2d 899 (1963); Bain Peanut Co. v. Pinson, 282 U.S. 499, 501, 51 S.Ct. 228, 229, 75 L.Ed. 482 (1931). But, as has already been seen, we are hardly presented here with some de minimis claim of discrimination resulting from the play necessary in any functioning system; to the contrary, it is clear that the Foundation Program utterly fails to

Page 88

ameliorate the seriously discriminatory effects of the local property tax. ⁴⁹

          Alternatively, the appellants and the majority may believe that the Equal Protection Clause cannot be offended by substantially unequal state treatment of persons who are similarly situated so long as the State provides everyone with some unspecified amount of education which evidently is 'enough.'⁵⁰ The basis for such a novel view is far from clear. It is, of course, true that the Constitution does not require precise equality in the treatment of all persons. As Mr. Justice Frankfurter explained:

          'The equality at which the 'equal protection' clause aims is not a disembodied equality. The Fourteenth Amendment enjoins 'the equal protection of the laws', and laws are not abstract propositions. . . . The Constitution does not require things which are different in fact or opinion to be treated in law as though they were the same.' Tigner v. Texas, 310 U.S. 141, 147, 60 S.Ct. 879, 882, 84 L.Ed. 1124 (1940).

          See also Douglas v. California, 372 U.S. 353, 357, 83 S.Ct. 814, 816, 9 L.Ed.2d 811 (1963); Goesaert v. Cleary, 335 U.S. 464, 466, 69 S.Ct. 198, 199, 93 L.Ed. 163 (1948).

Page 89

But this Court has never suggested that because some 'adequate' level of benefits is provided to all, discrimination in the provision of services is therefore constitutionally excusable. The Equal Protection Clause is not addressed to the minimal sufficiency but rather to the unjustifiable inequalities of state action. It mandates nothing less than that 'all persons similarly circumstanced shall be treated alike.' F. S. Royster Guano Co. v. Virginia, 253 U.S. 412, 415, 40 S.Ct. 560, 562, 64 L.Ed. 989 (1920).

          Even if the Equal Protection Clause encompassed some theory of constitutional adequacy, discrimination in the provision of educational opportunity would certainly seem to be a poor candidate for its application. Neither the majority nor appellants inform us how judicially manageable standards are to be derived for determining how much education is 'enough' to excuse constitutional discrimination. One would think that the majority would heed its own fervent affirmation of judicial self-restraint before undertaking the complex task of determining at large what level of education is constitutionally sufficient. Indeed, the majority's apparent reliance upon the adequacy of the educational opportunity assured by the Texas Minimum Foundation School Program seems fundamentally inconsistent with its own recognition that educational authorities are unable to agree upon what makes for educational quality, see ante, at 42—43, and n. 86 and at 47 n. 101. If, as the majority stresses, such authorities are uncertain as to the impact of various levels of funding on educational quality, I fail to see where it finds the expertise to divine that the particular levels of funding provided by the Program assure an adequate educational opportunity—much less an education substantially equivalent in quality to that which a higher level of funding might provide. Certainly appellants' mere assertion before this Court of the adequacy of the education guaranteed by the Minimum

Page 90

Foundation School Program cannot obscure the constitutional implications of the discrimination in educational funding and objective educational inputs resulting from the local property tax particularly since the appellees offered substantial uncontroverted evidence before the District Court impugning the now muchtouted 'adequacy' of the education guaranteed by the Foundation Program.⁵¹

          In my view, then, it is inequality—not some notion of gross inadequacy—of educational opportunity that raises a question of denial of equal protection of the laws. I find any other approach to the issue unintelligible and without directing principle. Here, appellees have made a substantial showing of wide variations in educational funding and the resulting educational opportunity afforded to the schoolchildren of Texas. This discrimination is, in large measure, attributable to significant disparities in the taxable wealth of local Texas school districts. This is a sufficient showing to raise a substantial question of discriminatory state action in violation of the Equal Protection Clause.⁵²

Page 91

          Despite the evident discriminatory effect of the Texas financing scheme, both the appellants and the majority raise substantial questions concerning the precise character of the disadvantaged class in this case. The District Court concluded that the Texas financing scheme draws 'distinction betwen groups of citizens depending upon the wealth of the district in which they live' and thus creates a disadvantaged class composed of persons living in property-poor districts. See 337 F.Supp., at 282. See also id., at 281. In light of the data introduced before the District Court, the conclusion that the schoolchildren of property-poor districts constitute a sufficient class for our purposes seems indisputablet to me.

          Appellants contend, however, that in constitutional terms this case involves nothing more than discrimination against local school districts, not against individuals, since on its face the state scheme is concerned only with the provision of funds to local districts. The result of the Texas financing scheme, appellants suggest, is merely that some local districts have more available revenues for education; others have less. In that respect,

Page 92

they point out, the States have broad discretion in drawing reasonable distinctions between their political subdivisions. See Griffin v. County School Board of Prince Edward County, 377 U.S. 218, 231, 84 S.Ct. 1226, 1233, 12 L.Ed.2d 256 (1964); McGowan v. Maryland, 366 U.S. 420, 427, 81 S.Ct. 1101, 1105, 6 L.Ed.2d 393 (1961); Salsburg v. Maryland, 346 U.S. 545, 550—554, 74 S.Ct. 280, 282—285, 98 L.Ed. 281 (1954).

          But this Court has consistently recognized that where there is in fact discrimination against individual interests, the constitutional guarantee of equal protection of the laws is not inapplicable simply because the discrimination is based upon some group characteristic such as geographic location. See Gordon v. Lance, 403 U.S. 1, 4, 91 S.Ct. 1889, 1891, 29 L.Ed.2d 273 (1971); Reynolds v. Sims, 377 U.S. 533, 565—566, 84 S.Ct. 1362, 1383—1384, 12 L.Ed.2d 506 (1964); Gray v. Sanders, 372 U.S. 368, 379, 83 S.Ct. 801, 807, 9 L.Ed.2d 821 (1963). Texas has chosen to provide free public education for all its citizens, and it has embodied that decision in its constitution. ⁵³ Yet, having established public education for its citizens, the State, as a direct consequence of the variations in local property wealth endemic to Texas' financing scheme, has provided some Texas schoolchildren with substantially less resources for their education than others. Thus, while on its face the Texas scheme may merely discriminate between local districts, the impact of that discrimination falls directly upon the children whose educational opportunity is dependent upon where they happen to live. Consequently, the District Court correctly concluded that the Texas financing scheme discriminates, from a constitutional perspective, between school children on the basis of the amount of taxable property located within their local districts.

          In my Brother STEWART's view, however, such a description of the discrimination inherent in this case is apparently not sufficient, for it fails to define the 'kind of objectively identifiable classes' that he evidentlyperceives

Page 93

to be necessary for a claim to be 'cognizable under the Equal Protection Clause,' ante, at 62. He asserts that this is also the view of the majority, but he is unable to cite, nor have I been able to find, any portion of the Court's opinion which remotely suggests that there is no objectively identifiable or definable class in this case. In any event, if he means to suggest that an essential predicate to equal protection analysis is the precise identification of the particular individuals who compose the disadvantaged class, I fail to find the source from which he derives such a requirement. Certainly such precision is not analytically necessary. So long as the basis of the discrimination is clearly identified, it is possible to test it against the State's purpose for such discrimination—whatever the standard of equal protection analysis employed.⁵⁴ This is clear from our decision only last Term in Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972), where the Court, in striking down Texas' primary filing fees as violative of equal protection, found no impediment to equal protection analysis in the fact that the members of the disadvantaged class could not be readily identified. The Court recognized that the filing-fee system tended 'to deny some voters the opportunity to vote for a candidate of their choosing; at the same time it gives the affluent the power to place on the ballot their own names or the names of persons they favor.' Id., at 144, 92 S.Ct., at 856. The

Page 94

Court also recognized that '(t)his disparity in voting power based on wealth cannot be described by reference to discrete and precisely defined segments of the community as is typical of inequities challenged under the Equal Protection Clause . . ..' Ibid. Nevertheless, it concluded that 'we would ignore reality were we not to recognize that this system falls with unequal weight on voters . . . according to their economic status.' Ibid. The nature of the classification in Bullock was clear, although the precise membership of the disadvantaged class was not. This was enough in Bullock for purposes of equal protection analysis. It is enough here.

          It may be, though, that my Brother STEWART is not in fact demanding precise identification of the membership of the disadvantaged class for purposes of equal protection analysis, but is merely unable to discern with sufficient clarity the nature of the discrimination charged in this case. Indeed, the Court itself displays some uncertainty as to the exact nature of the discrimination and the resulting disadvantaged class alleged to exist in this case. See ante, at 19—20. It is, of course, essential to equal protection analysis to have a firm grasp upon the nature of the discrimination at issue. In fact, the absence of such a clear, articulable understanding of the nature of alleged discrimination in a particular instance may well suggest the absence of any real discrimination. But such is hardly the case here.

          A number of theories of discrimination have, to be sure, been considered in the course of this litigation. Thus, the District Court found that in Texas the poor and minority group members tend to live in property-poor districts, suggesting discrimination on the basis of both personal wealth and race. See 337 F.Supp., at 282 and n. 3. The Court goes to great lengths to discredit the data upon which the District Court relied, and thereby its conclusion that poor people live in property-poor dis-

Page 95

tricts.⁵⁵ Although I have serious doubts as to the correctness of the Court's analysis in rejecting the data submitted below,⁵⁶ I have no need to join issue on these factual disputes.

Page 96

          I believe it is sufficient that the overarching form of discrimination in this case is between the schoolchildren of Texas on the basis of the taxable property wealth of the districts in which they happen to live. To understand both the precise nature of this discrimination and the parameters of the disadvantaged class it is sufficient to consider the constitutional principle which appellees contend is controlling in the context of educational financing. In their complaint appellees asserted that the Constitution does not permit local district wealth to be determinative of educational opportunity.⁵⁷ This is simply another way of saying, as the District Court concluded, that consistent with the guarantee of equal protection of the laws, 'the quality of public education may not be a function of wealth, other than the wealth of the state as a whole.' 337 F.Supp., at 284. Under such a principle, the children of a district are excessively advantaged if that district has more taxable property per pupil than the average amount of taxable property per pupil considering the State as a whole. By contrast, the children of a district are disadvantaged if that district has less taxable property per pupil than the state average. The majority attempts to disparage such a definition of the disadvantaged class as the product of an 'artificially defined level' of district wealth. Ante, at 28. But such is clearly not the case, for this is the

Page 97

definition unmistakably dictated by the constitutional principle for which appellees have argued throughout the course of this litigation. And I do not believe that a clearer definition of either the disadvantaged class of Texas schoolchildren or the allegedly unconstitutional discrimination suffered by the members of that class under the present Texas financing scheme could be asked for, much less needed.⁵⁸ Whether this discrimination, against the schoolchildren of property-poor districts, inherent in the Texas financing scheme, is violative of the Equal Protection Clause is the question to which the must now turn.

          To avoid having the Texas financing scheme struck down because of the interdistrict variations in taxable property wealth, the District Court determined that it was insufficient for appellants to show merely that the State's scheme was rationally related to some legitimate state purpose; rather, the discrimination inherent in the scheme had to be shown necessary to promote a 'compelling state interest' in order to withstand constitutional scrutiny. The basis for this determination was twofold: first, the financing scheme divides citizens on a wealth basis, a classification which the District Court viewed as highly suspect; and second, the discriminatory scheme directly affects what it considered to be a 'fundamental interest,' namely, education.

          This Court has repeatedly held that state discrimination which either adversely affects a 'fundamental interest,' see, e.g., Dunn v. Blumstein, 405 U.S. 330, 336—342, 92 S.Ct. 995, 999 1003, 31 L.Ed.2d 274 (1972); Shapiro v. Thompson, 394 U.S. 618, 629—631, 89 S.Ct. 1322, 1328—1330, 22 L.Ed.2d 600 (1969), or is based on a distinction of a suspect character, see, e.g., Graham v. Richardson, 403 U.S. 365, 372, 91 S.Ct. 1848, 1852, 29 L.Ed.2d 534

Page 98

(1971); McLaughlin v. Florida, 379 U.S. 184, 191—192, 85 S.Ct. 283, 287—289, 13 L.Ed.2d 222 (1964), must be carefully scrutinized to ensure that the scheme is necessary to promote a substantial, legitimate state interest. See, e.g., Dunn v. Blumstein, supra, 405 U.S., at 342—343, 92 S.Ct., at 1003—1004; Shapiro v. Thompson, supra, 394 U.S., at 634, 89 S.Ct., at 1331. The majority today concludes, however, that the Texas scheme is not subject to such a strict standard of review under the Equal Protection Clause. Instead, in its view, the Texas scheme must be tested by nothing more than that lenient standard of rationality which we have traditionally applied to discriminatory state action in the context of economic and commercial matters. See, e.g., McGowan v. Maryland, 366 U.S., at 425—426, 81 S.Ct., at 1104—1105; Morey v. Doud, 354 U.S. 457, 465—466, 77 S.Ct. 1344, 1349—1351, 1 L.Ed.2d 1485 (1957); F. S. Royster Guano Co. v. Virginia, 253 U.S., at 415, 40 S.Ct., at 561; Lindsley v. Natural Carbonic Gas Co., 220 U.S. 61, 78—79, 31 S.Ct. 337, 340—341, 55 L.Ed. 369 (1911). By so doing, the Court avoids the telling task of searching for a substantial state interest which the Texas financing scheme, with its variations in taxable district property wealth, is necessary to further. I cannot accept such an emasculation of the Equal Protection Clause in the context of this case.

          To begin, I must once more voice my disagreement with the Court's rigidified approach to equal protection analysis. See Dandridge v. Williams, 397 U.S. 471, 519—521, 90 S.Ct. 1153, 1178 1180, 25 L.Ed.2d 491 (1970) (dissenting opinion); Richardson v. Belcher, 404 U.S. 78, 90, 92 S.Ct. 254, 261, 30 L.Ed.2d 231 (1971) (dissenting opinion). The Court apparently seeks to establish today that equal protection cases fall into one of two neat categories which dictate the appropriate standard of review—strict scrutiny or mere rationality. But this Court's decisions in the field of equal protection defy such easy categorization. A principled reading of what this Court has done reveals that it has applied a spectrum of standards in reviewing discrimination allegedly violative of the Equal Protec-

Page 99

tion Clause. This spectrum clearly comprehends variations in the degree of care with which the Court will scrutinize particular classifications, depending, I believe, on the constitutional and societal importance of the interest adversely affected and the recognized invidiousness of the basis upon which the particular classification is drawn. I find in fact that many of the Court's recent decisions embody the very sort of reasoned approach to equal protection analysis for which I previously argued—that is, an approach in which 'concentration (is) placed upon the character of the classification in question, the relative importance to individuals in the class discriminated against of the governmental benefits that they do not receive, and the asserted state interests in support of the classification.' Dandridge v. Williams, supra, 397 U.S., at 520—521, 90 S.Ct., at 1180 (dissenting opinion).

          I therefore cannot accept the majority's labored efforts to demonstrate that fundamental interests, which call for strict scrutiny of the challenged classification, encompass only established rights which we are somehow bound to recognize from the text of the Constitution itself. To be sure, some interests which the Court has deemed to be fundamental for purposes of equal protection analysis are themselves constitutionally protected rights. Thus, discrimination against the guaranteed right of freedom of speech has called for strict judicial scrutiny. See Police Dept. of City of Chicago v. Mosley, 408 U.S. 92, 92 S.Ct. 2286, 33 L.Ed.2d 212 (1972). Further, every citizen's right to travel interstate, although nowhere expressly mentioned in the Constitution, has long been recognized as implicit in the premises underlying that document: the right 'was conceived from the beginning to be a necessary concomitant of the stronger Union the Constitution created.' United States v. Guest, 383 U.S. 745, 758, 86 S.Ct. 1170, 1178, 16 L.Ed.2d 239 (1966). See also Crandall v. Nevada, 6 Wall. 35, 48, 18 L.Ed. 744 (1868). Consequently, the Court has required that a state classification affecting theconstitutionally

Page 100

protected right to travel must be 'shown to be necessary to promote a compelling governmental interest.' Shapiro v. Thompson, 394 U.S., at 634, 89 S.Ct., at 1331. But it will not do to suggest that the 'answer' to whether an interest is fundamental for purposes of equal protection analysis is always determined by whether that interest 'is a right . . . explicitly or implicitly guaranteed by the Constitution,' ante, at 33—34.⁵⁹

          I would like to know where the Constitution guarantees the right to procreate, Skinner v. Oklahoma ex rel. Williamson, 316 U.S. 535, 541, 62 S.Ct. 1110, 1113, 86 L.Ed. 1655 (1942), or the right to vote in state elections, e.g., Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964), or the right to an appeal from a criminal conviction, e.g., Griffin v. Illinois, 351 U.S. 12, 76 S.Ct. 585, 100 L.Ed. 891 (1956). These are instances in which, due to the importance of the interests at stake, the Court has displayed a strong concern with the existence of discriminatory state treatment. But the Court has never said or indicated that these are interests which independently enjoy fullblown constitutional protection.

          Thus, in Buck v. Bell, 274 U.S. 200, 47 S.Ct. 584, 71 L.Ed. 1000 (1927), the Court refused to recognize a substantive constitutional guarantee of the right to procreate. Nevertheless, in Skinner v. Oklahoma ex rel. Williamson, supra, 316 U.S., at 541, 62 S.Ct., at 1113, the Court, without impugning the continuing validity of Buck v. Bell, held that 'strict scrutiny' of state discrimination affecting procreation 'is essential' for '(m)arriage and procreation are fundamental to the very existence and survival of the race.' Recently, in Roe v. Wade, 410 U.S. 113, 152—154, 93 S.Ct. 705, 726—727, 35 L.Ed.2d 147 (1973),

Page 101

the importance of procreation has indeed been explained on the basis of its intimate relationship with the constitutional right of privacy which we have recognized. Yet the limited stature thereby accorded any 'right' to procreate is evident from the fact that at the same time the Court reaffirmed its initial decision in Buck v. Bell. See Roe v. Wade, supra, at 154, 93 S.Ct., at 727.

          Similarly, the right to vote in state elections has been recognized as a 'fundamental political right,' because the Court concluded very early that it is 'preservative of all rights.' Yick Wo v. Hopkins, 118 U.S. 356, 370, 6 S.Ct. 1064, 1071, 30 L.Ed. 220 (1886); see, e.g., Reynolds v. Sims, supra, 377 U.S., at 561—562, 84 S.Ct. at 1381—1382. For this reason, 'this Court has made clear that a citizen has a constitutionally protected right to participate in elections on an equal basis with other citizens in the jurisdiction.' Dunn v. Blumstein, 405 U.S., at 336, 92 S.Ct., at 1000 (emphasis added). The final source of such protection from inequality in the provision of the state franchise is, of course, the Equal Protection Clause. Yet it is clear that whatever degree of importance has been attached to the state electoral process when unequally distributed, the right to vote in state elections has itself never been accorded the statute of an independent constitutional guarantee. ⁶⁰ See Oregon v. Mitchell, 400 U.S. 112, 91 S.Ct. 260, 27 L.Ed.2d 272 (1970); Kramer v. Union Free School District No. 15, 395 U.S. 621, 626—629, 89 S.Ct. 1886, 1889—1891, 23 L.Ed.2d 583 (1969); Harper v. Virginia Board of Elections, 383 U.S. 663, 665, 86 S.Ct. 1079, 1080, 16 L.Ed.2d 169 (1966).

Page 102

          Finally, it is likewise 'true that a State is not required by the Federal Constitution to provide appellate courts or a right to appellate review at all.' Griffin v. Illinois, 351 U.S., at 18, 76 S.Ct., at 590. Nevertheless, discrimination adversely affecting access to an appellate process which a State has chosen to provide has been considered to require close judicial scrutiny. See, e.g., Griffin v. Illinois, supra; Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963).⁶¹

          The majority is, of course, correct when it suggests that the process of determining which interests are fundamental is a difficult one. But I do not think the problem is insurmountable. And I certainly do not accept the view that the process need necessarily degenerate into an unprincipled, subjective 'picking-and-choosing' between various interests or that it must involve this Court in creating 'substantive constitutional rights in the name of guaranteeing equal protection of the laws,' ante, at 33. Although not all fundamental interests are constitutionally guaranteed, the determination of which interests are fundamental should be firmly rooted in the text of the Constitution. The task in every case should be to determine the extent to which constitutionally guaranteed rights are dependent on interests not mentioned in the Constitution. As the nexus between the specific constitutional guarantee and the nonconstitutional interest draws closer, the nonconstitutional interest becomes

Page 103

more fundamental and the degree of judicial scrutiny applied when the interest is infringed on a discriminatory basis must be adjusted accordingly. Thus, it cannot be denied that interests such as procreation, the exercise of the state franchise, and access to criminal appellate processes are not fully guaranteed to the citizen by our Constitution. But these interests have nonetheless been afforded special judicial consideration in the face of discrimination because they are, to some extent, interrelated with constitutional guarantees. Procreation is now understood to be important because of its interaction with the established constitutional right of privacy. The exercise of the state franchise is closely tied to basic civil and political rights inherent in the First Amendment. And access to criminal appellate processes enhances the integrity of the range of rights⁶² implicit in the Fourteenth Amendment guarantee of due process of law. Only if we closely protect the related interests from state discrimination do we ultimately ensure the integrity of the constitutional guarantee itself. This is the real lesson that must be taken from our previous decisions involving interests deemed to be fundamental.

          The effect of the interaction of individual interests with established constitutional guarantees upon the degree of care exercised by this Court in reviewing state discrimination affecting such interests is amply illustrated by our decision last Term in Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349 (1972). In Baird, the Court struck down as violative of the Equal Protection Clause a state statute which denied unmarried persons access to contraceptive devices on the same basis as married persons. The Court

Page 104

purported to test the statute under its traditional standard whether there is some rational basis for the discrimination effected. Id., at 446—447, 92 S.Ct. at 1034—1035. In the context of commercial regulation, the Court has indicated that the Equal Protection Clause 'is offended only if the classification rests on grounds wholly irrelevant to the achievement of the State's objective.' See, e.g., McGowan v. Maryland, 366 U.S., at 425, 81 S.Ct., at 1105; Kotch v. Board of River Port Pilot Comm'rs, 330 U.S. 552, 557, 67 S.Ct. 910, 912, 91 L.Ed. 1093 (1947). And this lenient standard is further weighted in the State's favor by the fact that '(a) statutory discrimination will not be set aside if any state of facts reasonably may be conceived (by the Court) to justify it.' McGowan v. Maryland, supra, 366 U.S., at 426, 81 S.Ct. at 1105. But in Baird the Court clearly did not adhere to these highly tolerant standards of traditional rational review. For although there were conceivable state interests intended to be advanced by the statute—e.g., deterrence of premarital sexual activity and regulation of the dissemination of potentially dangerous articles—the Court was not prepared to accept these interests on their face, but instead proceeded to test their substantiality by independent analysis. See 405 U.S., at 449—454, 92 S.Ct., at 1036—1039. Such close scrutiny of the State's interests was hardly characteristic of the deference shown state classifications in the context of economic interests. See, e.g., Goesaert v. Cleary, 335 U.S. 464, 69 S.Ct. 198, 93 L.Ed. 163 (1948); Kotch v. Board of River Port Pilot Comm'rs, supra. Yet I think the Court's action was entirely appropriate, for access to and use of contraceptives bears a close relationship to the individual's constitutional right of privacy. See 405 U.S., at 453 454; id., at 463—464, 92 S.Ct. 1038—1039; Id., at 1043—1044 (White, J., concurring in result). See also Roe v. Wade, 410 U.S., at 152—153, 93 S.Ct., at 726—727.

          A similar process of analysis with respect to the invidiousness of the basis on which a particular classification is drawn has also influenced the Court as to the

Page 105

appropriate degree of scrutiny to to accorded any particular case. The highly suspect character of classifications based on race,⁶³ nationality, ⁶⁴ or alienage⁶⁵ is well established. The reasons why such classifications call for close judicial scrutiny are manifold. Certain racial and ethnic groups have frequently been recognized as 'discrete and insular minorities' who are relatively powerless to protect their interests in the political process. See Graham v. Richardson, 403 U.S., at 372, 91 S.Ct., at 1852; United States v. Carolene Products Co., 304 U.S. 144, 152—153, n. 4, 58 S.Ct. 778, 783—784, 82 L.Ed. 1234 (1938). Moreover, race, nationality, or alienage is "in most circumstances irrelevant' to any constitutionally acceptable legislative purpose, Kiyoshi Hirabayashi v. United States, 320 U.S. 81, 100, 63 S.Ct. 1375, 87 L.Ed. 1774.' McLaughlin v. Florida, 379 U.S., at 192, 85 S.Ct., at 288. Instead, lines drawn on such bases are frequently the reflection of historic prejudices rather than legislative rationality. It may be that all of these considerations, which make for particular judicial solicitude in the face of discrimination on the basis of race, nationality, or alienage, do not coalesce—or at least not to the same degree—in other forms of discrimination. Nevertheless, these considerations have undoubtedly influenced the care with which the Court has scrutinized other forms of discrimination.

          In James v. Strange, 407 U.S. 128, 92 S.Ct. 2027, 32 L.Ed.2d 600 (1972), the Court held unconstitutional a state statute which provided for recoupment from indigent convicts of legal defense fees paid by the State. The Court found that the statute impermissibly differentiated between indigent criminals in debt to the State and civil judgment debtors, since criminal debtors were denied various protective exemptions

Page 106

afforded civil judgment debtors.⁶⁶ The Court suggested that in reviewing the statute under the Equal Protection Clause, it was merely applying the traditional requirement that there be "some rationality" in the line drawn between the different types of debtors. Id., at 140, 92 S.Ct., at 2034. Yet it then proceeded to scrutinize the statute with less than traditional deference and restraint. Thus, the Court recognized 'that state recoupment statutes may betoken legitimate state interests' in recovering expenses and discouraging fraud. Nevertheless, Mr. Justice Powell, speaking for the Court, concluded that

          'these interests are not thwarted by requiring more even treatment of indigent criminal defendants with other classes of debtors to whom the statute itself repeatedly makes reference. State recoupment laws, notwithstanding the state interests they may serve, need not blight in such discriminatory fashion the hopes of indigents for self sufficiency and self respect.' Id., at 141—142, 92 S.Ct., at 2034.

          The Court, in short, clearly did not consider the problems of fraud and collection that the state legislature might have concluded were peculiar to indigent criminal defendants to be either sufficiently important or at least sufficiently substantiated to justify denial of the protective exemptions afforded to all civil judgment debtors, to a class composed exclusively of indigent criminal debtors.

          Similarly, in Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971), the Court, in striking down a state statute which gave men

Page 107

preference over women when persons of equal entitlement apply for assignment as an administrator of a particular estate, resorted to a more stringent standard of equal protecting review than that employed in cases involving commercial matters. The Court indicated that it was testing the claim of sex discrimination by nothing more than whether the line drawn bore 'a rational relationship to a state objective,' which it recognized as a legitimate effort to reduce the work of probate courts in choosing between competing applications for letters of administration. Id., at 76, 92 S.Ct., at 254. Accepting such a purpose, the Idaho Supreme Court had thought the classification to be sustainable on the basis that the legislature might have reasonably concluded that, as a rule, men have more experience than women in business matters relevant to the administration of an estate. 93 Idaho 511, 514, 465 P.2d 635, 638 (1970). This Court, however, concluded that '(t)o give a mandatory preference to members of either sex over members of the other, merely to accomplish the elimination of hearings on the merits, is to make the very kind of arbitrary legislative choice forbidden by the Equal Protection Clause of the Fourteenth Amendment . . ..' 404 U.S., at 76, 92 S.Ct., at 254. This Court, in other words, was unwilling to consider a theoretical and unsubstantiated basis for distinction—however reasonable it might appear—sufficient to sustain a statute discriminating on the basis of sex.

          James and Reed can only be understood as instances in which the particularly invidious character of the classification caused the Court to pause and scrutinize with more than traditional care the rationality of state discrimination. Discrimination on the basis of past criminality and on the basis of sex posed for the Court the spector of forms of discrimination which it implicitly recognized to have deep social and legal roots without necessarily having any basis in actual differences. Still,

Page 108

the Court's sensitivity to the invidiousness of the basis for discrimination is perhaps most apparent in its decisions protecting the interests of children born out of wedlock from discriminatory state action. See Weber v. Aetna Casualty & Surety Co., 406 U.S., 164, 92 S.Ct. 1400, 31 L.Ed.2d 768 (1972); Levy v. Louisiana, 391 U.S. 68, 88 S.Ct. 1509, 20 L.Ed.2d 436 (1968).

          In Weber, the Court struck down a portion of a state workmen's compensation statute that relegated unacknowledged illegitimate children of the deceased to a lesser status with respect to benefits than that occupied by legitimate children of the deceased. The Court acknowledged the true nature of its inquiry in cases such as these: 'What legitimate state interest does the classification promote? What fundamental personal rights might the classification endanger?' Id., 406 U.S. at 173, 92 S.Ct., at 1405. Embarking upon a determination of the relative substantiality of the State's justifications for the classification, the Court rejected the contention that the classifications reflected what might be presumed to have been the deceased's preference of beneficiaries as 'not compelling . . . where dependency on the deceased is a prerequisite to anyone's recovery . . ..' Ibid. Likewise, it deemed the relationship between the State's interest in encouraging legitimate family relationships and the burden placed on the illegitimates too tenuous to permit the classification to stand. Ibid. A clear insight into the basis of the Court's action is provided by its conclusion:

          '(I)mposing disabilities on the illegitimate child is contrary to the basic concept of our system that legal burdens should bear some relationship to individual responsibility or wrongdoing. Obviously, no child is responsible for his birth and penalizing the illegitimate child is an ineffectual—as well as an unjust—way of deterring the parent. Courts are powerless to prevent the social opprobrium suffered by these hapless children, but the Equal Protection

Page 109

          Clause does enable us to strike down discriminatory laws relating to status of birth . . ..' Id., at 175—176, 92 S.Ct., at 1407 (footnote omitted).

          Status of birth, like the color of one's skin, is something which the individual cannot control, and should generally be irrelevant in legislative considerations. Yet illegitimacy has long been stigmatized by our society. Hence, discrimination on the basis of birth—particularly when it affects innocent children warrants special judicial consideration.

          In summary, it seems to me inescapably clear that this Court has consistently adjusted the care with which it will review state discrimination in light of the constitutional significance of the interests affected and the invidiousness of the particular classification. In the context of economic interests, we find that discriminatory state action is almost always sustained, for such interests are generally far removed from constitutional guarantees. Moreover, '(t)he extremes to which the Court has gone in dreaming up rational bases for state regulation in that area may in many instances be ascribed to a healthy revulsion from the Court's earlier excesses in using the Constitution to protect interests that have more than enough power to protect themselves in the legislative halls.' Dandridge v. Williams, 397 U.S., at 520, 90 S.Ct., at 1179 (dissenting opinion). But the situation differs markedly when discrimination against important individual interests with constitutional implications and against particularly disadvantaged or powerless classes is involved. The majority suggests, however, that a variable standard of review would give this Court the appearance of a 'super-legislature.' Ante, at 31. I cannot agree. Such an approach seems to me a part of the guarantees of our Constitution and of the historic experiences with oppression of and discrimination against discrete, powerless minorities which underlie that document. In truth,

Page 110

the Court itself will be open to the criticism raised by the majority so long as it continues on its present course of effectively selecting in private which cases will be afforded special consideration without acknowledging the true basis of its action.⁶⁷

          Opinions such as those in Reed and James seem drawn more as efforts to shield rather than to reveal the true basis of the Court's decisions. Such obfuscated action may be appropriate to a political body such as a legislature, but it is not appropriate to this Court. Open debate of the bases for the Court's action is essential to the rationality and consistency of our decisionmaking process. Only in this way can we avoid the label of legislature and ensure the integrity of the judicial process.

          Nevertheless, the majority today attempts to force this case into the same category for purposes of equal protection analysis as decisions involving discrimination affecting commercial interests. By so doing, the majority singles this case out for analytic treatment at odds with what seems to me to be the clear trend of recent decisions in this Court, and thereby ignores the constitutional importance of the interest at stake and the invidiousness of the particular classification, factors that call for far more than the lenient scrutiny of the Texas financing scheme which the majority pursues. Yet if the discrimination inherent in the Texas scheme is scrutinized with the care demanded by the interest and classification present in this case, the unconstitutionality of that scheme is unmistakable.

          Since the Court now suggests that only interests guaranteed by the Constitution are fundamental for purposes of equal protection analysis, and since it rejects

Page 111

the contention that public education is fundamental, it follows that the Court concludes that public education is not constitutionally guaranteed. It is true that this Court has never deemed the provision of free public education to be required by the Constitution. Indeed, it has on occasion suggested that state-supported education is a privilege bestowed by a State on its citizens. See Missouri ex rel. Gaines v. Canada, 305 U.S., at 349, 59 S.Ct., at 236. Nevertheless, the fundamental importance of education is amply indicated by the prior decisions of this Court, by the unique status accorded public education by our society, and by the close relationship between education and some of our most basic constitutional values.

          The special concern of this Court with the educational process of our country is a matter of common knowledge. Undoubtedly, this Court's most famous statement on the subject is that contained in Brown v. Board of Education, 347 U.S., at 493, 74 S.Ct., at 691:

          'Today, education is perhaps the most important function of state and local governments. Compulsory school attendance laws and the great expenditures for education both demonstrate our recognition of the importance of education to our democratic society. It is required in the performance of our most basic public responsibilities, even service in the armed forces. It is the very foundation of good citizenship. Today it is a principal instrument in awakening the child to cultural values, in preparing him for later professional training, and in helping him to adjust normally to his environment. . . .'

          Only last Term, the Court recognized that '(p)roviding public schools ranks at the very apex of the function of a State.' Wisconsin v. Yoder, 406 U.S. 205, 213, 92 S.Ct., 1526, 1532, 32 L.Ed.2d 15 (1972). This is clearly borne out by the fact that in 48

Page 112

of our 50 States the provision of public education is mandated by the state constitution.⁶⁸ No other state function is so uniformly recognized⁶⁹ as an essential element of our society's well-being. In large measure, the explanation for the special importance attached to education must rest, as the Court recognized in Yoder, id., at 221, 92 S.Ct., at 1536, on the facts that 'some degree of education is necessary to prepare citizens to participate effectively and intelligently in our open political system . . .,' and that 'education prepares individuals to be self-reliant and self-sufficient participants in society.' Both facets of this observation are suggestive of the substantial relationship which education bears to guarantees of our Constitution.

          Education directly affects the ability of a child to exercise his First Amendment rights, both as a source and as a receiver of information and ideas, whatever interests he may pursue in life. This Court's decision in Sweezy v. New Hampshire, 354 U.S. 234, 250, 77 S.Ct. 1203, 1212, 1 L.Ed.2d 1311 (1957), speaks of the right of students 'to inquire, to study and to evaluate, to gain new maturity and understanding . . .' Thus, we have not casually described the classroom as the "marketplace of ideas." Keyishian v. Board of Regents, 385 U.S. 589, 603, 87 S.Ct. 675, 683, 17 L.Ed.2d 629 (1967). The opportunity for formal education may not necessarily be the essential determinant of an individual's ability to enjoy throughout his life the rights of free speech andassociation

Page 113

guaranteed to him by the First Amendment. But such an opportunity may enhance the individual's enjoyment of those rights, not only during but also following school attendance. Thus, in the final analysis, 'the pivotal position of education to success in American society and its essential role in opening up to the individual the central experiences of our culture lend it an importance that is undeniable.'⁷⁰

          Of particular importance is the relationship between education and the political process. 'Americans regard the public schools as a most vital civic institution for the preservation of a democratic system of government.' School District of Abington Township v. Schempp, 374 U.S. 203, 230, 83 S.Ct. 1560, 1576, 10 L.Ed.2d 844 (1963) (Brennan, J., concurring). Education serves the essential function of instilling in our young an understanding of and appreciation for the principles and operation of our governmental processes. ⁷¹ Education may instill the interest and provide the tools necessary for political discourse and debate. Indeed, it has frequently been suggested that education is the dominant factor affecting political consciousness and participation.⁷² A system of '(c)ompetition in ideas andgovernmental

Page 114

policies is at the core of our electoral process and of the First Amendment freedoms.' Williams v. Rhodes, 393 U.S. 23, 32, 89 S.Ct. 5, 11, 21 L.Ed.2d 24 (1968). But of most immediate and direct concern must be the demonstrated effect of education on the exercise of the franchise by the electorate. The right to vote in federal elections is conferred by Art. I, § 2, and the Seventeenth Amendment of the Constitution, and access to the state franchise has been afforded special protection because it is 'preservative of other basic civil and political rights,' Reynolds v. Sims, 377 U.S., at 562, 84 S.Ct., at 1381. Data from the Presidential Election of 1968 clearly demonstrate a direct relationship between participation in the electoral process and level of educational attainment;⁷³ and, as this Court recognized in Gaston County v. United States, 395 U.S. 285, 296, 89 S.Ct. 1720, 1725, 23 L.Ed.2d 309 (1969), the quality of education offered may

Page 115

influence a child's decision to 'enter or remain in school.' It is this very sort of intimate relationship between a particular personal interest and specific constitutional guarantees that has heretofore caused the Court to attach special significance, for purposes of equal protection analysis, to individual interests such as procreation and the exercise of the state franchise.⁷⁴

          While ultimately disputing little of this, the majority seeks refuge in the fact that the Court has 'never presumed to possess either the ability or the authority to guarantee to the citizenry the most effective speech or the most informed electoral choice.' Ante at 36. This serves only to blur what is in fact at stake. With due respect, the issue is neither provision of the most effective speech nor of the most informed vote. Appellees

Page 116

do not now seek the best education Texas might provide. They do seek, however, an end to state discrimination resulting from the unequal distribution of taxable district property wealth that directly impairs the ability of some districts to provide the same educational opportunity that other districts can provide with the same or even substantially less tax effort. The issue is, in other words, one of discrimination that affects the quality of the education which Texas has chosen to provide its children; and, the precise question here is what importance should attach to education for purposes of equal protection analysis of that discrimination. As this Court held in Brown v. Board of Education, 347 U.S., at 493, 74 S.Ct., at 691, the opportunity of education, 'where the state has undertaken to provide it, is a right which must be made available to all on equal terms.' The factors just considered, including the relationship between education and the social and political interests enshrined within the Constitution, compel us to recognize the fundamentality of education and to scrutinize with appropriate care the bases for state discrimination affecting equality of educational opportunity in Texas' school districts⁷⁵—aconclusion

Page 117

which is only strengthened when we consider the character of the classification in this case.

          The District Court found that in discriminating between Texas schoolchildren on the basis of the amount of taxable property wealth located in the district in which they live, the Texas financing scheme created a form of wealth discrimination. This Court has frequently recognized that discrimination on the basis of wealth may create a classification of a suspect character and thereby call for exacting judicial scrutiny. See, e.g., Griffin v. Illinois, 351 U.S. 12, 76 S.Ct. 585, 100 L.Ed. 891 (1956); Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963); McDonald v. Board of Election Comm'rs of Chicago, 394 U.S. 802, 807, 89 S.Ct. 1404, 1407, 22 L.Ed.2d 739 (1969). The majority, however, considers any wealth classification in this case to lack certain essential characteristics which it contends are common to the instances of wealth discrimination that this Court has heretofore recognized. We are told that in every prior case involving a wealth classification, the members of the disadvantaged class have 'shared two distinguishing characteristics: because

Page 118

of their impecunity they were completely unable to pay for some desired benefit, and as a consequence, they sustained an absolute deprivation of a meaningful opportunity to enjoy that benefit.' Ante, at 20. I cannot agree. The Court's distinctions may be sufficient to explain the decisions in Williams v. Illinois, 399 U.S. 235, 90 S.Ct. 2018, 26 L.Ed.2d 586 (1970); Tate v. Short, 401 U.S. 395, 91 S.Ct. 668, 28 L.Ed.2d 130 (1971); and even Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972). But they are not in fact consistent with the decisions in Harper v. Virginia Board of Elections, 383 U.S. 663, 86 S.Ct. 1079, 16 L.Ed.2d 169 (1966), or Griffin v. Illinois, supra, or Douglas v. California, supra.

          In Harper, the Court struck down as violative of the Equal Protection Clause an annual Virginia poll tax of $1.50, payment of which by persons over the age of 21 was a prerequisite to voting in Virginia elections. In part, the Court relied on the fact that the poll tax interfered with a fundamental interest—the exercise of the state franchise. In addition, though, the Court emphasized that '(l)ines drawn on the basis of wealth or property . . . are traditionally disfavored.' 383 U.S., at 668, 86 S.Ct., at 1082. Under the first part of the theory announced by the majority, the disadvantaged class in Harper, in terms of a wealth analysis, should have consisted only of those too poor to afford the $1.50 necessary to vote. But the Harper Court did not see it that way. In its view, the Equal Protection Clause 'bars a system which excludes (from the franchise) those unable to pay a fee to vote or who fail to pay.' Ibid. (Emphasis added.) So far as the Court was concerned, the 'degree of the discrimination (was) irrelevant.' Ibid. Thus, the Court struck down the poll tax in toto; it did not order merely that those too poor to pay the tax be exempted; complete impecunity clearly was not determinative of the limits of the disadvantaged class, nor was it essential to make an equal protection claim.

Page 119

                    Similarly, Griffin and Douglas refute the majority's contention that we have in the past required an absolute deprivation before subjecting wealth classifications to strict scrutiny. The Court characterizes Griffin as a case concerned simply with the denial of a transcript or an adequate substitute therefor, and Douglas as involving the denial counsel. But in both cases the question was in fact whether 'a State that (grants) appellate review can do so in a way that discriminates against some convicted defendants on account of their proverty.' Griffin v. Illinois, supra, 351 U.S., at 18, 76 S.Ct., at 590 (emphasis added). In that regard, the Court concluded that inability to purchase a transcript denies 'the poor an adequate appellate review accorded to all who have money enough to pay the costs in advance,' ibid. (emphasis added), and that 'the type of an appeal a person is afforded . . . hinges upon whether or not he can pay for the assistance of counsel,' Douglas v. California, supra, 372 U.S., at 355—356, 83 S.Ct., at 816 (emphasis added). The right of appeal itself was not absolutely denied to those too poor to pay; but because of the cost of a transcript and of counsel, the appeal was a substantially less meaningful right for the poor than for the rich.⁷⁶ It was on these terms that the Court a denial of equal protection, and those terms clearly encompassed degrees of discrimination on the

Page 120

basis of wealth which do not amount to outright denial of the affected right or interest.⁷⁷

          This is not to say that the form of wealth classification in this case does not differ significantly from those recognized in the previous decisions of this Court. Our prior cases have dealt essentially with discrimination on the basis of personal wealth.⁷⁸ Here, by contrast, the

Page 121

children of the disadvantaged Texas school districts are being discriminated against not necessarily because of their personal wealth or the wealth of their families, but because of the taxable property wealth of the residents of the district in which they happen to live. The appropriate question, then, is whether the same degree of judicial solicitude and scrutiny that has previously been afforded wealth classifications is warranted here.

          As the Court points out, ante, at 28—29, no previous decision has deemed the presence of just a wealth classification to be sufficient basis to call forth rigorous judicial scrutiny of allegedly discriminatory state action. Compare, e.g., Harper v. Virginia Board of Elections, supra, with, e.g., James v. Valtierra, 402 U.S. 137, 91 S.Ct. 1331, 28 L.Ed.2d 678 (1971). That wealth classifications alone have not necessarily been considered to bear the same high degree of suspectness as have classifications based on, for instance, race or alienage may be explainable on a number of grounds. The 'poor' may not be seen as politically powerless as certain discrete and insular minority groups. ⁷⁹ Personal proverty may entail much the same social stigma as historically attached to certain racial or ethnic groups.⁸⁰ But personal poverty is not a permanent disability; its shackles may be escaped. Perhaps most importantly, though, personal wealth may not necessarily share the general irrelevance as a basis for legislative action that race or nationality is recognized to have. While the 'poor' have frequently been a

Page 122

legally disadvantaged group,⁸¹ it cannot be ignored that social legislation must frequently take cognizance of the economic status of our citizens. Thus, we have generally gauged the invidiousness of wealth classifications with an awareness of the importance of the interests being affected and the relevance of personal wealth to those interests. See Harper v. Virginia Board of Elections, supra.

          When evaluated with these considerations in mind, it seems to me that discrimination on the basis of group wealth in this case likewise calls for careful judicial scrutiny. First, it must be recognized that while local district wealth may serve other interests,⁸² it bears no relationship whatsoever to the interest of Texas schoolchildren in the educational opportunity afforded them by the State of Texas. Given the importance of that interest, we must be particularly sensitive to the invidious characteristics of any form of discrimination that is not clearly intended to serve it, as opposed to some other distinct state interest. Discrimination on the basis of group wealth may not, to be sure, reflect the social stigma frequently attached to personal poverty. Nevertheless, insofar as group wealth discrimination involves wealth over which the disadvantaged individual has no significant control,⁸³ it represents in fact a more serious basis of discrimination than does personal wealth. For such discrimination

Page 123

is no reflection of the individual's characteristics or his abilities. And thus—particularly in the context of a disadvantaged class composed of children—we have previously treated discrimination on a basis which the individual cannot control as constitutionally disfavored. Cf. Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 92 S.Ct. 1400, 31 L.Ed.2d 768 (1972); Levy v. Louisiana, 391 U.S. 68, 88 S.Ct. 1509, 20 L.Ed.2d 436 (1968).

          The disability of the disadvantaged class in this case extends as well into the political processes upon which we ordinarily rely as adequate for the protection and promotion of all interests. Here legislative reallocation of the State's property wealth must be sought in the face of inevitable opposition from significantly advantaged districts that have a strong vested interest in the preservation of the status quo, a problem not completely dissimilar to that faced by underrepresented districts prior to the Court's intervention in the process of reapportionment,⁸⁴ see Baker v. Carr, 369 U.S. 186, 191—192, 82 S.Ct. 691, 695—697, 7 L.Ed.2d 663 (1962).

          Nor can we ignore the extent to which, in contrast to our prior decisions, the State is responsible for the wealth discrimination in this instance. Griffin, Douglas, Williams, Tate, and our other prior cases have dealt with discrimination on the basis of indigency which was attributable to the operation of the private sector. But we have no such simple de facto wealth discrimination here. The means for financing public education in Texas are selected and specified by the State. It is the State that has created local school districts, and tied educational funding to the local property tax and thereby to local district wealth. At the same time, governmentally

Page 124

imposed land use controls have undoubtedly encouraged and rigidified natural trends in the allocation of particular areas for residential or commercial use,⁸⁵ and thus determined each district's amount of taxable property wealth. In short, this case, in contrast to the Court's previous wealth discrimination decisions, can only be seen as 'unusual in the extent to which governmental action is the cause of the wealth classifications.'⁸⁶

          In the final anaylsis, then The invidious characteristics of the group wealth classification present in this case merely serve to emphasize the need for careful judicial scrutiny of the State's justifications for the resulting interdistrict discrimination in the educational opportunity afforded to the schoolchildren of Texas.

          The nature of our inquiry into the justifications for state discrimination is essentially the same in all equal protection cases: We must consider the substantiality of the state interests sought to be served, and we must scrutinize the reasonableness of the means by which the State has sought to advance its interests. See Police Dept. of City of Chicago v. Mosley, 408 U.S., at 95, 92 S.Ct., at 2289. Differences in the application of this test are, in my view, a function of the constitutional importance of the interests at stake and the invidiousness of the particular classification. In terms of the asserted state interests, the Court has indicated that it will require, for instance, a 'compelling,' Shapiro v. Thompson, 394 U.S., at 634, 89 S.Ct., at 1331, or a 'substantial'

Page 125

or 'important,' Dunn v. Blumstein, 405 U.S., at 343, 92 S.Ct., at 1003, state interest to justify discrimination affecting individual interests of constitutional significance. Whatever the differences, if any, in these descriptions of the character of the state interest necessary to sustain such discrimination, basic to each is, I believe, a concern with the legitimacy and the reality of the asserted state interests. Thus, when interests of constitutional importance are at stake, the Court does not stand ready to credit the State's classification with any conceivable legitimate purpose,⁸⁷ but demands a clear showing that there are legitimate state interests which the classification was in fact intended to serve. Beyond the question of the adequacy of the State's purpose for the classification, the Court traditionally has become increasingly sensitive to the means by which a State chooses at act as its action affects more directly interests of constitutional significance. See, e.g., United States v. Robel, 389 U.S. 258, 265, 88 S.Ct. 419, 424, 19 L.Ed.2d 508 (1967); Shelton v. Tucker, 364 U.S. 479, 488, 81 S.Ct. 247, 252, 5 L.Ed.2d 231 (1961). Thus, by now, 'less restrictive alternatives' analysis is firmly established in equal protection jurisprudence. See Dunn v. Blumstein, supra, 405 U.S., at 343, 92 S.Ct., at 1003; Kramer v. Union Free School District No. 15, 395 U.S., at 627, 89 S.Ct., at 1889. It seems to me that the range of choice we are willing to accord the State in selecting the means by which it will act, and the care with which we scrutinize the effectiveness of the means which the State selects, also must reflect the constitutional importance of the interest affected and the invidiousness of the particular classification. Here, both the nature of the interest and the classification dictate close judicial scrutiny of the purposes which Texas seeks to serve with its present educational financing

Page 126

scheme and of the means it has selected to serve that purpose.

          The only justification offered by appellants to sustain the discrimination in educational opportunity caused by the Texas financing scheme is local educational control. Presented with this justification, the District Court concluded that '(n)ot only are defendants unable to demonstrate compelling state interests for their classifications based upon wealth, they fail even to establish a reasonable basis for these classifications.' 337 F.Supp., at 284. I must agree with this conclusion.

          At the outset, I do not question that local control of public education, as an abstract matter, constitutes a very substantial state interest. We observed only last Term that '(d)irect control over decisions vitally affecting the education of one's children is a need that is strongly felt in our society.' Wright v. Council of the City of Emporia, 407 U.S. 451, 469, 92 S.Ct. 2196, 2206, 33 L.Ed.2d 51 (1972). See also id., at 477—478, 92 S.Ct., at 2210 2211 (Burger, C.J., dissenting). The State's interest in local educational control—which certainly includes questions of educational funding—has deep roots in the inherent benefits of community support for public education. Consequently, true state dedication to local control would present, I think, a substantial justification to weigh against simply interdistrict variations in the treatment of a State's schoolchildren. But I need not now decide how I might ultimately strike the balance were we confronted with a situation where the State's sincere concern for local control inevitably produced educational inequality. For, on this record, it is apparent that the State's purported concern with local control is offered primarily as an excuse rather than as a justification for interdistrict inequality.

          In Texas, statewide laws regulate in fact the most minute details of local public education. For example,

Page 127

the State prescribes required courses.⁸⁸ All textbooks must be submitted for state approval,⁸⁹ and only approved textbooks may be used.⁹⁰ The State has established the qualifications necessary for teaching in Texas public schools and the procedures for obtaining certification.⁹¹ The State has even legislated on the length of the school day.⁹² Texas' own courts have said:

          'As a result of the acts of the Legislature our school system is not of mere local concern but it is statewide. While a school district is local in territorial limits, it is an integral part of the vast school system which is coextensive with the confines of the State of Texas.' Treadaway v. Whitney Independent School District, 205 S.W.2d 97, 99 (Tex.Civ.App.1947).

          See also El Dorado Independent School District v. Tisdale, 3 S.W.2d 420, 422 (Tex.Com.App. 1928).

          Moreover, even if we accept Texas' general dedication to local control in educational matters, it is difficult to find any evidence of such dedication with respect to fiscal matters. It ignores reality to suggest—as the Court does, ante, at 49—50—that the local property tax element of the Texas financing scheme reflects a conscious legislative effort to provide school districts with local fiscal control. If Texas had a system truly dedicated to local fiscal control, one would expect the quality of the educational opportunity provided in each district to vary with the decision of the voters in that district as

Page 128

to the level of sacrifice they wish to make for public education. In fact, the Texas scheme produces precisely the opposite result. Local school districts cannot choose to have the best education in the State by imposing the highest tax rate. Instead, the quality of the educational opportunity offered by any particular district is largely determined by the amount of taxable property located in the district—a factor over which local voters can exercise no control.

          The study introduced in the District Court showed a direct inverse relationship between equalized taxable district property wealth and district tax effort with the result that the property-poor districts making the highest tax effort obtained the lowest per-pupil yield.⁹³ The implications of this situation for local choice are illustrated by again comparing the Edgewood and Alamo Heights School Districts. In 1967—1968, Edgewood, after contributing its share to the Local Fund Assignment, raised only $26 per pupil through its local property tax, whereas Alamo Heights was able to raise $333 per pupil. Since the funds received through the Minimum Foundation School Program are to be used only for minimum professional salaries, transportation costs, and operating expenses, it is not hard to see the lack of local choice with respect to higher teacher salaries to attract more and better teachers, physical facilities, library books, and facilities, special courses, or participation in special state and federal matching funds programs—under which a property-poor district such as Edgewood is forced to labor.⁹⁴ In fact, because of the difference in taxable local property wealth, Edgewood would have to tax itself almost nine times as heavily to obtain the same

Page 129

yield as Alamo Heights.⁹⁵ At present, then, local control is a myth for many of the local school districts in Texas. As one district court has observed, 'rather than reposing in each school district the economic power to fix its own level of per pupil expenditure, the State has so arranged the structure as to guarantee that some districts will spend low (with high taxes) while others will spend high (with low taxes).' Van Dusartz v. Hatfield, 334 F.Supp. 870, 876 (D.C.Minn.1971).

          In my judgment, any substantial degree of scrutiny of the operation of the Texas financing scheme reveals that the State has selected means wholly inappropriate to secure its purported interest in assuring its school districts local fiscal control.⁹⁶ At the same time, appellees have pointed out a variety of alternative financing schemes which may serve the State's purported interest in local control as well as, if not better than, the present scheme without the current impairment of the educational opportunity of vast numbers of Texas schoolchildren.⁹⁷ I see no need, however, to explore the practical or constitutional merits of those suggested alternatives at this time for, whatever their positive or negative features, experience

Page 130

with the present financing scheme impugns any suggestion that it constitutes a serious effort to provide local fiscal control. If for the sake of local education control, this Court is to sustain interdistrict discrimination in the educational opportunity afforded Texas school children, it should require that the State present something more than the mere sham now before us.

          In conclusion, it is essential to recognize that an end to the wide variations in taxable district property wealth inherent in the Texas financing scheme would entail none of the untoward consequences suggested by the Court or by the appellants.

          First, affirmance of the District Court's decisions would hardly sound the death knell for local control of education. It would mean neither centralized decisionmaking nor federal court intervention in the operation of public schools. Clearly, this suit has nothing to do with local decisionmaking with respect to educational policy or even educational spending. It involves only a narrow aspect of local control—namely, local control over the raising of educational funds. In fact, in striking down interdistrict disparities in taxable local wealth, the District Court took the course which is most likely to make true local control over educational decision-making a reality for all Texas school districts.

          Nor does the District Court's decision even necessarily eliminate local control of educational funding. The District Court struck down nothing more than the continued interdistrict wealth discrimination inherent in the present property tax. Both centralized and decentralized plans for educational funding not involving such interdistrict discrimination have been put forward.⁹⁸ The choice

Page 131

among these or other alternatives would remain with the State, not with the federal courts. In this regard, it should be evident that the degree of federal intervention

Page 132

in matters of local concern would be substantially less in this context than in previous decisions in which we have been asked effectively to impose a particular scheme upon the States under the guise of the Equal Protection Clause. See, e.g., Dandridge v. Williams, 397 U.S. 471, 90 S.Ct. 1153, 25 L.Ed.2d 491 (1970); Cf. Richardson v. Belcher, 404 U.S. 78, 92 S.Ct. 254, 30 L.Ed.2d 231 (1971).

          Still, we are told that this case requires us 'to condemn the State's judgment in conferring on political subdivisions the power to tax local property to supply revenues for local interests.' Ante, at 40. Yet no one in the course of this entire litigation has ever questioned the constitutionality of the local property tax as a device for raising educational funds. The District Court's decision, at most, restricts the power of the State to make educational funding dependent exclusively upon local property taxation so long as there exists interdistrict disparities in taxable property wealth. But it hardly eliminates the local property tax as a source of educational funding or as a means of providing local fiscal control.⁹⁹

          The Court seeks solace for its action today in the possibility of legislative reform. The Court's suggestions of legislative redress and experimentation will doubtless be of great comfort to the schoolchildren of Texas' disadvantaged districts, but considering the vested interests of wealthy school districts in the preservation of the status quo, they are worth little more. The possibility of legislative action is, in all events, no answer to this Court's duty under the Constitution to eliminate unjustified state discrimination. In this case we have been presented with an instance of such discrimination, in a particularly invidious form, against an individual interest of large constitutional and practical importance. To support the demonstrated discrimination in the provision

Page 133

of educational opportunity the State has offered a justification which, on analysis, takes on at best an ephemeral character. Thus, I believe that the wide disparities in taxable district property wealth inherent in the local property tax element of the Texas financing scheme render that scheme violative of the Equal Protection Clause.¹⁰⁰

          I would therefore affirm the judgment of the District Court.

Page 134

                                    TO OPINION OF MARSHALL, J., DISSENTING
 
                                     REVENUES OF TEXAS SCHOOL DISTRICTS
                        CATEGORIZED BY EQUALIZED PROPERTY VALUES AND SOURCE OF FUNDS
                                             CATEGORIES
 
                                                                                                    Total Revenues
                                                               State and Local                        Per Pupil
 Market Value of                                                Revenues Per      Federal           (State-Local-
 Taxable Property         Local Revenues  State Revenues        Pupil (Columns    Revenues          Federal, Columns
  Per Pupil                 Per Pupil       Per Pupil             1 and 2)        Per Pupil           1, 2 and 4)
 
 
 Above $100,000                $610            $205                 $815             $41                $856
  (10 districts)
 $100,000-$50,000               287             257                  544              66                 610
  (26 districts)
 $50,000-$30,000                224             260                  484              45                 529
  (30 districts)
 $30,000-$10,000                166             295                  461              85                 546
  (40 districts)
 Below $10,000                   63             243                  306             135                 441
 (4 districts)
 

Page 135

      APPENDIX II TO OPINION OF MARSHALL, J., DISSENTING
 
                         TEXAS SCHOOL DISTRICTS CATEGORIZED BY
                          EQUALIZED PROPERTY VALUES, EQUALIZED
                            TAX RATES, AND YIELD OF RATES
 
   CATEGORIES                       EQUALIZED                  YIELD PER PUPIL
  Market Value of                      TAX                     (Equalized Rate
  Taxable Property                    RATES                   Applied to District
    Per Pupil                        ON $100                     Market Value)
 
  Above $100,000                      $.31                            $585
   (10 districts)
  $100,000-$50,000                     .38                             262
  (26 districts)
  $50,000-$30,000                      .55                             213
  (30 districts)
  $30,000-$10,000                      .72                             162
  (40 districts)
  Below $10,000                        .70                              60
  (4 districts)
 

Page 136

                              APPENDIX III TO
                   OPINION OF MARSHALL, J., DISSENTING
 
         SELECTED BEXAR COUNTY, TEXAS, SCHOOL DISTRICTS CATEGORIZED BY
             EQUALIZED PROPERTY VALUATION AND SELECTED INDICATORS
                          OF EDUCATIONAL QUALITY
 
  Selected Districts                Per Cent of   Per Cent of
  From High to Low by Professional  Teachers With          Total Staff   Student-
  Professional
  Market Valuation   Salaries   Per College     Masters   With Emerg- Counselor   Professional
  Per Pupil             Pupil       Degrees     Degrees   ency Permits Ratios    Per 100 Pupils
 
 ALAMO HIEIGHTS          $372          100%         40%         11%       645         4.80
 NORTH EAST               288           99          24           7      1,516         4.50
 SAN ANTONIO              251           98          29          17      2,320         4.00
 NORTH SIDE               258           99          20          17      1,493         4.30
 HARLANDALE               243           94          21          22      1,800         4.00
 EDGEWOOD                 209           96          15          47      3,098         4.06
 

Page 137

                APPENDIX IV TO OPINION OF MARSHALL, J., DISSENTING
 
                      BEXAR COUNTY, TEXAS, SCHOOL DISTRICTS
                       RANKED BY EQUALIZED PROPERTY VALUE
                       AND TAX RATE REQUIRED TO GENERATE
                        HIGHEST YIELD IN ALL DISTRICTS
 
    Districts Ranked from                                   Tax Rate Per $100
     High to Low Market                                      Needed to Equal
    Valuation Per Pupil                                       Highest Yield
 
       ALAMO HEIGHTS                                              $0.68
       JUDSON                                                      1.04
       EAST CENTRAL                                                1.17
       NORTH EAST                                                  1.21
       SOMERSET                                                    1.32
       SAN ANTONIO                                                 1.56
       NORTH SIDE                                                  1.65
       SOUTH WEST                                                  2.10
       SOUTH SIDE                                                  3.03
       HARLANDALE                                                  3.20
       SOUTH SAN ANTONIO                                           5.77
       EDGEWOOD                                                    5.76
 
 

^{1. Not all of the children of these complainants attend public school. One family's children are enrolled in private school 'because of the condition of the schools in the Edgewood Independent School District.' Third Amended Complaint, App. 14.}

^2. The San Antonio Independent School District, whose name this case still bears, was one of seven school districts in the San Antonio metropolitan area that were originally named as defendants. After a pretrial conference, the District Court issued an order dismissing the school districts from the case. Subsequently, the San Antonio Independent School District joined in the plaintiffs' challenge to the State's school finance system and filed an amicus curiae brief in support of that position in this Court.

^3. A three-judge court was properly convened and there are no questions as to the District Court's jurisdiction or the direct appealability of its judgment. 28 U.S.C. §§ 2281, 1253.

^4. The trial was delayed for two years to permit extensive pretrial discovery and to allow completion of a pending Texas legislative investigation concerning the need for reform of its public school finance system. 337 F.Supp. 280, 285 n. 11 (W.D.Tex.1971).

^5. 337 F.Supp. 280. The District Court stayed its mandate for two years to provide Texas an opportunity to remedy the inequities found in its financing program. The court, however, retained jurisdiction to fashion its own remedial order if the State failed to offer an acceptable plan. Id., at 286.

^6. Tex.Const., Art. X, § 1 (1845):

'A general diffusion of knowledge being essential to the preservation of the rights and liberties of the people, it shall be the duty of the legislature of this State to make suitable provision for the support and maintenance of public schools.'

Id., § 2:

'The Legislature shall, as early as practicable, establish free schools throughout the State, and shall furnish means for their support by taxation on property . . ..'

^7. Tex.Const. of 1876, Art. 7, § 3, as amended, Aug. 14, 1883, Vernon's Ann.Tex.St.

^8. Id., Art. 7, §§ 3, 4, 5.

^9. 3 Gammel's Laws of Texas 1847—1854, p. 1461. See Tex.Const. Art. 7, §§ 1, 2, 5 (interpretive commentaries); 1 Report of Governor's Committee on Public School Education, The Challenge and the Chance 27 (1969) (hereinafter Governor's Committee Report).

^10. Tex.Const., Art. 7, § 5 (see also the interpretive commentary); 5 Governor's Committee Report 11—12.

^11. The various sources of revenue for the Available School Fund are cataloged in A Report of the Adequacy of Texas Schools, prepared by Texas State Board of Education, 7—15 (1938) (hereinafter Texas State Bd. of Educ.).

^12. Tex.Const., Art. 7, § 3, as amended, Nov. 5, 1918 (see interpretive commentary).

^13. 1 Governor's Committee Report 35; Texas State Md. of Educ., supra, n. 11, at 5—7; J. Coons, W. Clune, & S. Sugarman, Private Wealth and Public Education 48—49 (1970); E. Cubberley, School Funds and Their Apportionment 21—27 (1905).

^14. By 1940, one-half of the State's population was clustered in its metropolitan centers. 1 Governor's Committee Report 35.

^15. Gilmer-Aikin Committee, To Have What We Must 13 (1948).

^16. Still, The Gilmer-Aikin Bills 11—13 (1950); Texas State Bd. of Educ., supra, n. 11.

^17. R. Still, supra, n. 16, at 12. It should be noted that during this period the median per-pupil expenditure for all schools with an enrollment of more than 200 was approximately $50 per year. During this same period, a survey conducted by the State Board of Education concluded that 'in Texas the best educational advantages offered by the State at present may be had for the median cost of $52.67 per year per pupil in average daily attendance.' Texas State Bd. of Educ., supra, n. 11, at 56.

^18. General Laws of Texas, 46th Legis., Reg.Sess.1939, c. 7, pp. 274—275 ($22.50 per student); General & Spec.Laws of Texas, 48th Legis., Reg.Sess.1943, c. 161, pp. 262—263 ($25 per student).

^19. General & Spec.Laws of Texas, 49th Legis., Reg.Sess.1945, c. 52, pp. 74—75; Still, supra, n. 16, at 12.

^20. For a complete history of the adoption in Texas of a foundation program, see Still, supra, n. 16. See also 5 Governor's Committee Report 14; Texas Research League, Public School Finance Problems in Texas 9 (Interim Report 1972).

^21. For the 1970—1971 school year this state aid program accounted for 48% of all public school funds. Local taxation contributed 41.1% and 10.9% was provided in federal funds. Texas Research League, supra, n. 20, at 9.

^22. 5 Governor's Committee Report 44—48.

^23. At present, there are 1,161 school districts in Texas. Texas Research League, supra, n. 20, at 12.

^24. In 1948, the Gilmer-Aikin Committee found that some school districts were not levying any local tax to support education. Gilmer-Aikin Committee, supra, n. 15, at 16. The Texas State Board of Education Survey found that over 400 common and independent school districts were levying no local property tax in 1935—1936. Texas State Bd. of Educ., supra n. 11, at 39—42.

^25. Gilmer-Aikin Committee, supra, n. 15, at 15.

^26. 1 Governor's Committee Report 51—53.

^27. Texas Research League, supra, n. 20, at 2.

^28. In the years between 1949 and 1967, the average per-pupil expenditure for all current operating expenses increased from $206 to $493. In that same period, capital expenditures increased from $44 to $102 per pupil. 1 Governor's Committee Report 53—54.

^29. Acts 1949, 51st Legis., p. 625, c. 334, Art. 4, Tex.Educ.Code Ann. § 16.302 (1972); see generally 3 Governor's Committee Report 113—146; Berke, Carnevale, Morgan & White, The Texas School Finance Case: A Wrong in Search of a Remedy, 1 J. of L. & Educ. 659, 681—682 (1972).

^30. The family income figures are based on 1960 census statistics.

^31. The Available School Fund, technically, provides a second source of state money. That Fund has continued as in years past (see text accompanying nn. 16—19, supra) to distribute uniform per-pupil grants to every district in the State. In 1968, this Fund allotted $98 per pupil However, because the Available School Fund contribution is always subtracted from a district's entitlement under the Foundation Program, it plays no siginficant role in educational finance today.

^32. While federal assistance has an ameliorating effect on the difference in school budgets between wealthy and poor disdistricts, the District Court rejected an argument made by the State in that court that it should consider the effect of the federal grant in assessing the discrimination claim. 337 F.Supp., at 284. The State has not renewed that contention here.

^33. A map of Bexar County included in the record shows that Edgewood and Alamo Heights are among the smallest districts in the county and are of approximately equal size. Yet, as the figures above indicate, Edgewood's student population is more than four times that of Alamo Heights. This factor obviously accounts for a significant percentage of the differences between the two districts in per-pupil property values and expenditures. If Alamo Heights had as many students to educate as Edgewood does (22,000) its per pupil assessed property value would be approximately $11,100 rather than $49,000, and its per-pupil expenditures would therefore have been considerably lower.

^34. The figures quoted above vary slightly from those utilized in the District Court opinion. 337 F.Supp., at 282. These trivial differences are apparently a product of that court's reliance on slightly different statistical data than we have relied upon.

^35. Although the Foundation Program has made significantly greater contributions to both school districts over the last several years, it is apparent that Alamo Heights has enjoyed a larger gain. The sizable difference between the Alamo Heights and Edgewood grants is due to the emphasis in the State's allocation formula on the guaranteed minimum salaries for teachers. Higher salaries are guaranteed to teachers having more years of experience and possessing more advanced degrees. Therefore, Alamo Heights, which has a greater percentage of experienced personnel with advanced degrees, receives more state support. In this regard, the Texas Program is not unlike that presently in existence in a number of other States. Coones, Clune, Sugarman, supra, n. 13, at 63—125. Because more dollars have been given to districts that already spend more per pupil, such Foundation formulas have been described as 'anti-equalizing.' Ibid. The formula, however, is anti-equalizing only if viewed in absolute terms. The percentage disparity between the two Texas districts is diminshed substantially by state aid. Alamo Heights derived in 1967—1968 almost 13 times as much money from local taxes as Edgewood did. The state aid grants to each district in 1970—1971 lowered the ratio to approximately two to one, i.e., Alamo Heights had a little more than twice as much money to spend per pupil from its combined state and local resources.

^36. Texas Research League, supra, n. 20, at 13.

^37. The Economic Index, which determines each county's share of the total Local Fund Assignment, is based on a complex formula conceived in 1949 when the Foundation Program was instituted. See text, supra, at 9—10. It has frequently been suggested by Texas researchers that the fomula be altered in several respects to provide a more accurate reflection of local taxpaying ability, especially of urban school districts. 5 Governor's Committee, Report 48; Texas Research League, Texas Public School Finance: A Majority of Exceptions 31—32 (2d Interim Report 1972); Berke, Carnevale, Morgan & White, supra, n. 29, at 680—681.

^38. The District Court relied on the findings presented in an affidavit submitted by Professor Berke of Syracuse University. His sampling of 110 Texas school districts demonstrated a direct correlation between the amount of a district's taxable property and its level of per-pupil expenditures. But this study found only a partial correlation between a district's median family income and per-pupil expenditures. The study also shows, in the relatively few districts at the extremes, an inverse correlation between percentage of minorities and expenditures.

      Categorized by Equalized Property Values, Median Family
      Income, and State-Local Revenue
 
    Market Value                 Median                        State &
    of Taxable                   Family        Per Cent         Local
    Property                     Income        Minority        Revenues
    Per Pupil                   From 1960       Pupils         Per Pupil
 
    Above $100,000               $5,900           8%             $815
     (10 districts)
    $100,000-$50,000             $4,425          32%             $544
     (26 districts)
    $50,000-$30,000              $4,900          23%             $483
     (30 districts)
    $30,000-$10,000              $5,050          31%             $462
     (40 districts)
    Below $10,000                $3,325          79%             $305
     (4 districts)
 

Although the correlations with respect to family income and race appear only to exist at the extremes, and although the affiant's methodology has been questioned (see Goldstein, Interdistrict Inequalities in School Financing: A Critical Analysis of Serrano v. Priest and its Progeny, 120 U.Pa.L.Rev. 504, 523—525, nn. 67, 71 (1972)), insofar as any of these correlations is relevant to the constitutional thesis presented in this case we may accept its basic thrust. But see infra, at 27—25. For a defense of the reliability of the affidavit, see Berke, Carnevale, Morgan & White, supra, n. 29.

^39. E.g., Police Dept. of the City of Chicago v. Mosley, 408 U.S. 92, 92 S.Ct. 2286, 33 L.Ed.2d 212 (1972); Dunn v. Blumstein, 405 U.S. 330, 92 S.Ct. 995, 31 L.Ed.2d 274 (1972); Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969).

^40. E.g., Graham v. Richardson, 403 U.S. 365, 91 S.Ct. 1848, 29 L.Ed.2d 534 (1971); Loving v. Virginia, 388 U.S. 1, 87 S.Ct. 1817, 18 L.Ed.2d 1010 (1967); McLaughlin v. Florida, 379 U.S. 184, 85 S.Ct. 283, 13 L.Ed.2d 222 (1964).

^41. See Dunn v. Blumstein, supra, 405 U.S., at 343, 92 S.Ct., at 1003, and the cases collected therein.

^42. Brief for Appellants 11.

^43. Ibid.

^44. Tr. of Oral Arg. 3; Reply Brief for Appellants 2.

^45. E.g., Griffin v. Illinois, 351 U.S. 12, 76 S.Ct. 585, 100 L.Ed. 891 (1956); Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963).

^46. Harper v. Virginia Bd. of Elections, 383 U.S. 663, 86 S.Ct. 1079, 16 L.Ed.2d 169 (1966); McDonald v. Board of Election Com'rs, 394 U.S. 802, 89 S.Ct. 1404, 22 L.Ed.2d 739 (1969); Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972); Goosby v. Osser, 409 U.S. 512, 93 S.Ct. 854, 35 L.Ed.2d 36 (1973).

^47. See cases cited in text, infra, at 29—30.

^48. Serrano v. Priest, 5 Cal.3d 584, 96 Cal.Rptr. 601, 487 P.2d 1241 (1971); Van Dusartz v. Hatfield, 334 F.Supp. 870 (D.C.Minn.1971); Robinson v. Cahill, 118 N.J.Super. 223, 287 A.2d 187 (1972); Milliken v. Green, 389 Mich. 1, 203 N.W.2d 457 (1972), rehearing granted, Jan. 1973.

^49. In their complaint, appellees purported to represent a class composed of persons who are 'poor' and who reside in school districts having a 'low value of . . . property.' Third Amended Complaint App. 15. Yet appellees have not defined the term 'poor' with reference to any absolute or functional level of impecunity. See text, infra, at 22—23. See also Brief for Appellees 1, 3; Tr. of Oral Arg. 20—21.

^50. Appellees' proof at trial focused on comparative differences in family incomes between residents of wealthy and poor districts. They endeavored, apparently, to show that there exists a direct correlation between personal family income and educational expenditures. See text, infra, at 25—27. The District Court may have been relying on this notion of relative discrimination based on family wealth. Citing appellees' statistical proof, the court emphasized that 'those districts most rich in property also have the highest median family income . . . while the poor property districts are poor in income . . ..' 337 F.Supp., at 282.

^51. At oral argument and in their brief, appellees suggest that description of the personal status of the residents in districts that spend less on education is not critical to their case. In their view, the Texas system is impermissibly discriminatory even if relatively poor districts do not contain poor people. Brief for Appellees 43—44; Tr. of Oral Arg. 20—21. There are indications in the District Court opinion that it adopted this theory of districts discrimination. The opinion repeatedly emphasizes the comparative financial status of districts and early in the opinion it describes appellees' class as being composed of 'all . . . children throughout Texas who live in school districts with low property valuations.' 337 F.Supp., at 281.

^52. Mayer v. City of Chicago, 404 U.S. 189, 92 S.Ct. 410, 30 L.Ed.2d 372 (1971); Williams v. Oklahoma City, 395 U.S. 458, 89 S.Ct. 1818, 23 L.Ed.2d 440 (1969); Gardner v. California, 393 U.S. 367, 89 S.Ct. 580, 21 L.Ed.2d 601 (1969); Roberts v. LaVallee, 389 U.S. 40, 88 S.Ct. 194, 19 L.Ed.2d 41 (1967); Long v. District Court of Iowa, 385 U.S. 192, 87 S.Ct. 362, 17 L.Ed.2d 290 (1966); Draper v. Washington, 372 U.S. 487, 83 S.Ct. 774, 9 L.Ed.2d 899 (1963); Eskridge v. Washington State Board of Prisons, 357 U.S. 214, 78 S.Ct. 1061, 2 L.Ed.2d 1269 (1958).

^53. Note, A Statistical Analysis of the School Finance Decisions: On Winning Battles and Losing Wars, 81 Yale L.J. 1303, 1328—1329 (1972).

^54. Id., at 1324 and n. 102.

^55. Id., at 1328.

^56. Each of appellees' possible theories of wealth discrimination is founded on the assumption that the quality of education varies directly with the amount of funds expended on it and that, therefore, the difference in quality between two schools can be determined simplistically by looking at the difference in per-pupil expenditures. This is a matter of considerable dispute among educators and commentators. See nn. 86 and 101, infra.

^57. E.g., Bullock v. Carter, 405 U.S., at 137, 149, 92 S.Ct., at 852, 858; Mayer v. City of Chicago, 404 U.S., at 194, 92 S.Ct., at 414; Draper v. Washington, 372 U.S., at 495—496, 83 S.Ct., at 778—779; Douglas v. California, 372 U.S., at 357, 83 S.Ct., at 816.

^58. Gilmer-Aikin Committee, supra, n. 15, at 13. Indeed, even though local funding has long been a significant aspect of educational funding, the State has always viewed providing an acceptable education as one of its primary functions. See Texas State Bd. of Educ., supra, n. 11, at 1, 7.

^59. Brief for Appellants 35; Reply Brief for Appellants 1.

^60. An educational financing system might be hypothesized, however, in which the analogy to the wealth discrimination cases would be considerably closer. If elementary and secondary education were made available by the State only to those able to pay a tuition assessed against each pupil, there would be a clearly defined class of 'poor' people—definable in terms of their inability to pay the prescribed sum—who would be absolutely precluded from receiving an education. That case would present a far more compelling set of circumstances for judicial assistance than the case before us today. After all, Texas has undertaken to do a good deal more than provide an education to those who can afford it. It has provided what it considers to be an adequate base education for all children and has attempted, though imperfectly, to ameliorate by state funding and by the local assessment program the disparities in local tax resources.

^61. Also, it should be recognized that median income statistics may not define with any precision the status of individual families within any given district. A more dependable showing of comparative wealth discrimination would also examine factors such as the average income, the mode, and the concentration of poor families in any district.

^62. Cf. Jefferson v. Hackney, 406 U.S. 535, 547—549, 92 S.Ct. 1724, 1723—1733, 32 L.Ed.2d 285 (1972); Ely, Legislative and Administrative Motivation in Constitutional Law, 79 Yale L.J. 1205, 1258—1259 (1970); Simon, The School Finance Decisions: Collective Bargaining and Future Finance Systems, 82 Yale L.J. 409, 439—440 (1973).

^63. Supra, at 15 n. 38.

^64. Studies in other States have also questioned the existence of any dependable correlation between a district's wealth measured in terms of assessable property and the collective wealth of families residing in the district measured in terms of median family income. Ridenour & Ridenour, Serrano v. Priest: Wealth and Kansas School Finance, 20 Kan.L. 213, 225 (1972) ('it can be argued that there exists in Kansas almost an inverse correlation: districts with highest income per pupil have low assessed value per pupil, and districts with high assessed value per pupil have low income per pupil'); Davis, Taxpaying Ability: A Study of the Relationship Between Wealth and Income in California Counties, in The Challenge of Change in School Finance, 10th Nat. Educational Assn. Conf. on School Finance 199 (1967). Note, 81 Yale L.J., supra, n. 53. See also Goldstein, supra, n. 38, at 522 527.

^65. Indeed, this is predisely how the plaintiffs in Serrano v. Priest defined the class they purported to represent: 'Plaintiff children claim to represent a class consisting of all public school pupils in California, 'except children in that school district . . . which . . . affords the greatest educational opportunity of all school districts within California." 5 Cal.3d, at 589, 96 Cal.Rptr., at 604, 487 P.2d, at 1244. See also Van Dusartz v. Hatfield, 334 F.Supp., at 873.

^66. Appellees, however, have avoided describing the Texas system as one resulting merely in discrimination between districts per se since this Court has never questioned the State's power to draw reasonable distinctions between political subdivisions within its borders. Griffin v. County School Board of Prince Edward County, 377 U.S. 218, 230—231, 84 S.Ct. 1226, 1232—1233, 12 L.Ed.2d 256 (1964); McGowan v. Maryland, 366 U.S. 420, 427, 81 S.Ct. 1101, 1105, 6 L.Ed.2d 393 (1961); Salsburg v. Maryland, 346 U.S. 545, 552, 74 S.Ct. 280, 284, 98 L.Ed. 281 (1954).

^67. E.g., Harper v. Virginia Bd. of Elections, 383 U.S. 663, 86 S.Ct. 1079, 16 L.Ed.2d 169 (1966); United States v. Kras, 409 U.S. 434, 93 S.Ct. 631, 34 L.Ed.2d 626 (1973). See Mr. Justice MARSHALL'S dissenting opinion, post, at 121.

^68. See Serrano v. Priest, supra; Van Dusartz v. Hatfield, supra; Robinson v. Cahill, 118 N.J.Super. 223, 287 A.2d 187, (1972); Coons, Clune & Sugarman, supra, n. 13, at 339—393; Goldstein, supra, n. 38, at 534—541; Vieira, Unequal Educational Expenditures: Some Minority Views on Serrano v. Priest, 37 Mo.L.Rev. 617, 618—624 (1972); Comment, Educational Financing, Equal Protection of the Laws, and the Supreme Court, 70 Mich.L.Rev. 1324, 1335—1342 (1972); Note, The Public School Financing Cases: Interdistrict Inequalities and Wealth Discrimination, 14 Ariz.L.Rev. 88, 120—124 (1972).

^69. 337 F.Supp., at 283.

^70. E.g., United States v. Guest, 383 U.S. 745, 757—759, 86 S.Ct. 1170, 1177—1179, 16 L.Ed.2d 239 (1966); Oregon v. Mitchell, 400 U.S. 112, 229, 237—238, 91 S.Ct. 260, 317, 321—322, 27 L.Ed.2d 272 (1970) (opinion of Brennan, White, and Marshall, JJ.).

^71. After Dandridge v. Williams, 397 U.S. 471, 90 S.Ct. 1153, 25 L.Ed.2d 491 (1970), there could be no lingering question about the constitutional foundation for the Court's holding in Shapiro. In Dandridge, the Court applied the rational-basis test in reviewing Maryland's maximum family grant provision under its AFDC program. A federal district court held the provision unconstitutional, applying a stricter standard of review. In the course of reversing the lower court, the Court distinguished Shapiro properly on the ground that in that case 'the Court found state interference with the constitutionally protected freedom of interstate travel.' Id., at 484 n. 16, 90 S.Ct., at 1161.

^72. The Court refused to apply the strict-scrutiny test despite its contemporaneous recognition in Goldberg v. Kelly, 397 U.S. 254, 264, 90 S.Ct. 1011, 1018, 25 L.Ed.2d 287 (1970) that 'welfare provides the means to obtain essential food, clothing, housing, and medical care.'

^73. In Eisenstadt, the Court struck down a Massachusetts statute that prohibited the distribution of contraceptive devices, finding that the law failed 'to satisfy even the more lenient equal protection standard.' 405 U.S., at 447 n. 7, 92 S.Ct., at 1035. Nevertheless, in dictum, the Court recited the correct form of equal protection analysis: '(I)f we were to conclude that the Massachusetts statute impinges upon fundamental freedoms under Griswold (v. Connecticut, 381 U.S. 479, 85 S.Ct. 1678, 14 L.Ed.2d 510 (1965)), the statutory classification would have to be not merely rationally related to a valid public purpose but necessary to the achievement of a compelling state interest.' Ibid. (emphasis in original).

^74. Dunn fully canvasses this Court's voting rights cases and explains that 'this Court has made clear that a citizen has a constitutionally protected right to participate in elections on an equal basis with other citizens in the jurisdiction.' 405 U.S., at 336, 92 S.Ct., at 1000 (emphasis supplied). The constitutional underpinnings of the right to equal treatment in the voting process can no longer be doubted even though, as the Court noted in Harper v. Virginia Bd. of Elections, 383 U.S., at 665, 86 S.Ct., at 1080, 'the right to vote in state elections is nowhere expressly mentioned.' See Oregon v. Mitchell, 400 U.S., at 135, 138—144, 91 S.Ct., at 270, 271—275 (Douglas, J.) 229, 241—242, 91 S.Ct. 317, 323—324 (Brennan, White, and Marshall, JJ.); Bullock v. Carter, 405 U.S., at 140—144, 92 S.Ct., at 854—856; Kramer v. Union Free School District, 395 U.S. 621, 625—630, 89 S.Ct. 1886, 1888—1889, 23 L.Ed.2d 583 (1969); Williams v. Rhodes, 393 U.S. 23, 29, 30—31, 89 S.Ct. 5, 9, 10—11, 21 L.Ed.2d 24 (1968); Reynolds v. Sims, 377 U.S. 533, 554—562, 84 S.Ct. 1362, 1377—1382, 12 L.Ed.2d 506 (1964); Gray v. Sanders, 372 U.S. 368, 379—381, 83 S.Ct. 801, 807—809, 9 L.Ed.2d 821 (1963).

^75. In Mosley, the Court struck down a Chicago antipicketing ordinance that exempted labor picketing from its prohibitions. The ordinance was held invalid under the Equal Protection Clause after subjecting it to careful scrutiny and finding that the ordinance was not narrowly drawn. The stricter standard of review was appropriately applied since the ordinance was one 'affecting First Amendment interests.' 408 U.S., at 101, 92 S.Ct., at 2293.

^76. Skinner applied the standard of close scrutiny to a state law permitting forced sterilization of 'habitual criminals.' Implicit in the Court's opinion is the recognition that the right of procreation is among the rights of personal privacy protected under the Constitution. See Roe v. Wade, 410 U.S. 113, 152, 93 S.Ct. 705, 726, 35 L.Ed.2d 147 (1973).

^77. See, e.g., Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 389—390, 89 S.Ct. 1794, 1806—1807, 23 L.Ed.2d 371 (1969); Stanley v. Georgia, 394 U.S. 557, 564, 89 S.Ct. 1243, 1247, 22 L.Ed.2d 542 (1969); Lamont v. Postmaster General, 381 U.S. 301, 306—307, 85 S.Ct. 1493, 1496—1497, 14 L.Ed.2d 398 (1965).

^78. Since the right to vote, per se, is not a constitutionally protected right, we assume that appellees' references to that right are simply shorthand references to the protected right, implicit in our constitutional system, to participate in state elections on an equal basis with other qualified voters whenever the State has adopted an elective process for determining who will represent any segment of the State's population. See n. 74, supra.

^79. The States have often pursued their entirely legitimate interest in assuring 'intelligent exercise of the franchise,' Katzenbach v. Morgan, 384 U.S. 641, 655, 86 S.Ct. 1717, 1726, 16 L.Ed.2d 828 (1966), through such devices as literacy tests and age restrictions on the right to vote. See ibid.; Oregon v. Mitchell, 400 U.S. 112, 91 S.Ct. 260, 27 L.Ed.2d 272 (1970). And, where those restrictions have been found to promote intelligent use of the ballot without discriminating against those racial and ethnic minorities previously deprived of an equal educational opportunity, this Court has upheld their use. Compare Lassiter v. Northampton County Bd. of Elections, 360 U.S. 45, 79 S.Ct. 985, 3 L.Ed.2d 1072 (1959), with Oregon v. Mitchell, supra, 400 U.S., at 133, 91 S.Ct., at 269 (Black, J.), 135, 144—147, 91 S.Ct. 270, 274 276 (Douglas, J.), 152, 216—217, 91 S.Ct. 279, 310—311 (Harlan, j.), 229, 231—236, 91 S.Ct. 317, 318—321 (Brennan, White, and Marshall, JJ.), 281, 282—284, 91 S.Ct. 343—344 (Stewart, J.), and Gaston County v. United States, 395 U.S. 285, 89 S.Ct. 1720, 23 L.Ed.2d 309 (1969).

^80. See Schoettle, The Equal Protection Clause in Public Education, 71 Col.L.Rev. 1355, 1389—1390 (1971); Vieira, supra, n. 68, at 622—623; Comment, Tenant Interest Representation: Proposal for a National Tenants' Association, 47 Tex.L.Rev. 1160, 1172 1173, n. 61 (1969).

^81. Katzenbach v. Morgan involved a challenge by registered voters in New York City to a provision of the Voting Rights Act of 1965 that prohibited enforcement of a state law calling for English literacy tests for voting. The law was suspended as to residents from Puerto Rico who had completed at least six years of education at an 'American-flag' school in that country even though the language of instruction was other than English. This Court upheld the questioned provision of the 1965 Act over the claim that it discriminated against those with a sixth-grade education obtained in non-English-speaking schools other than the ones designated by the federal legislation.

^82. Cf. Meyer v. Nebraska, 262 U.S. 390, 43 S.Ct. 625, 67 L.Ed. 1042 (1923); Pierce v. Society of Sisters, 268 U.S. 510, 45 S.Ct. 571, 69 L.Ed. 1070 (1925); Hargrave v. Kirk, 313 F.Supp. 944 (M.D.Fla.1970), vacated, 401 U.S. 476, 91 S.Ct. 856, 28 L.Ed.2d 196 (1971).

^83. See Schilb v. Kuebel, 404 U.S. 357, 92 S.Ct. 479, 30 L.Ed.2d 502 (1971); McDonald v. Board of Election Com'rs, 394 U.S. 802, 89 S.Ct. 1404, 22 L.Ed.2d 739 (1969).

^84. See, e.g., Bell's Gap R. Co. v. Pennsylvania, 134 U.S. 232, 10 S.Ct. 533, 33 L.Ed. 892 (1890); Carmichael v. Southern Coal & Coke Co., 301 U.S. 495, 508—509, 57 S.Ct. 868, 871—872, 81 L.Ed. 1245 (1937); Allied Stores of Ohio, Inc. v. Bowers, 358 U.S. 522, 79 S.Ct. 437, 3 L.Ed.2d 480 (1959).

^85. Those who urge that the present system be invalidated offer little guidance as to what type of school financing should replace it. The most likely result of rejection of the existing system would be state-wide financing of all public education with funds derived from taxation of property or from the adoption or expansion of sales and income taxes. See Simon, supra, n. 62. The authors of Private Wealth and Public Education, supra, n. 13, at 201—242, suggest an alternative scheme, known as 'district power equalizing.' In simplest terms, the State would guarantee that at any particular rate of property taxation the district would receive a stated number of dollars regardless of the district's tax base. To finance the subsidies to 'poorer' districts, funds would be taken away from the 'wealthier' districts that, because of their higher property values, collect more than the stated amount at any given rate. This is not the place to weigh the arguments for an against 'district power equalizing,' beyond noting that commentators are in disagreement as to whether it is feasible, how it would work, and indeed whether it would violate the equal protection theory underlying appellees' case. President's Commission on School Finance, Schools, People, & Money 32—33 (1972); Bateman & Brown. Some Reflections on Serrano v. Priest, 49 J. Urban L. 701, 706—708 (1972); Brest, Book Review, 23 Stan.L.Rev. 591, 594—596 (1971); Goldstein, supra, n. 38, at 542 543; Wise, School Finance Equalization Lawsuits: A Model Legislative Response, 2 Yale Rev. of L. & Soc. Action 123, 125 (1971); Silard & White, Intrastate Inequalities in Public Education: The Case for Judicial Relief Under the Equal Protection Clause, 1970 Wis.L.Rev. 7, 29—30.

^86. The quality-cost controversy has received considerable attention. Among the notable authorities on both sides are the following: C. Jencks, Inequality (1972); C. Silberman, Crisis in the Classroom (1970); U.S. Office of Education, Equality of Educational Opportunity (1966) (the Coleman Report); On Equality of Educational Opportunity (F. Mosteller & D. Moynihan eds. 1972); J. Guthrie, G. Kleindorfer, H. Levin & R. Stout, Schools and Inequality; President's Commission on School Finance, supra, n. 85; Swanson, The Cost-Quality Relationship, in The Challenge of Change in School Finance, 10th Nat. Educational Assn. Conf. on School Finance 151 (1967).

^87. See the results of the Texas Governor's Committee's statewide survey on the goals of education in that State. 1 Governor's Committee Report 59—68. See also Goldstein, supra, n. 38, at 519—522; Schoettle, supra, n. 80; authorities cited in n. 86, supra.

^88. Allied Stores of Ohio, Inc. v. Bowers, 358 U.S. 522, 530, 532, 79 S.Ct. 437, 442, 444, 3 L.Ed.2d 480 (1959) (Brennan, J., concurring); Katzenbach v. Morgan, 384 U.S., at 659, 661, 86 S.Ct., at 1731, 1732 (Harlan, J., dissenting).

^89. In 1970 Texas expended approximately.$2.1 billion for education and a little over one billion came from the Minimum Foundation Program. Texas Research League, supra, n. 20, at 2.

^90. Tex.Educ.Code Ann. § 16.13 (1972) V.T.C.A.

^91. Id., § 16.18.

^92. Id., § 16.15.

^93. Id., §§ 16.16, 16.17, 16.19.

^94. Id., §§ 16.45, 16.51—16.63.

^95. Id., §§ 12.01—12.04.

^96. Id., § 11.26(a)(5).

^97. Id., § 16.301 et seq.

^98. See supra, at 13—14.

^99. Gilmer-Aikin Committee, supra, n. 15, at 15.

^100. There is no uniform statewide assessment practice in Texas. Commercial property, for example, might be assessed at 30% of market value in one county and at 50% in another. 5 Governor's Committee Report 25—26; Berke, Carnevale, Morgan & White, supra, n. 29, at 666—667, n. 16.

^101. Texas Research League, supra, n. 20, at 18. Texas, in this regard, is not unlike most other States. One commentator has observed that 'disparities in expenditures appear to be largely explained

by variations in teacher salaries.' Simon, supra, n. 62, at 413.

As previously noted, see text accompanying n. 86, supra, the extent to which the quality of education varies with expenditure per pupil is debated inconclusively by the most thoughtful students of public education. While all would agree that there is a correlation up to the point of providing the recognized essentials in facilities and academic opportunities, the issues of greatest disagreement include the effect on the quality of education of pupil-teacher ratios and of higher teacher salary schedules. E.g., Office of Education, supra, n. 86, at 316—319. The state funding in Texas is designed to assure, on the average, one teacher for every 25 students, which is considered to be a favorable ratio by most standards. Whether the minimum salary of $6,000 per year is sufficient in Texas to attract qualified teachers may be more debatable, depending in major part upon the location of the school district. But there appear to be few empirical data that support the advantage of any particular pupil-teacher ratio or that document the existence of a dependable correlation between the level of public school teachers' salaries and the quality of their classroom instruction. An intractable problem in dealing with teachers' salaries is the absence, up to this time, of satisfactory techniques for judging their ability or performance. Relatively few school systems have merit plans of any kind, with the result that teachers' salaries are usually increased across the board in a way which tends to reward the least deserving on the same basis as the most deserving. Salaries are usually raised automatically on the basis of length of service and according to predetermined 'steps,' extending over 10- to 12-year periods.

^102. President's Commission on School Finance, supra, n. 85, at 9. Until recently, Hawaii was the only State that maintained a purely state-funded educational program. In 1968, however, that State amended its educational finance statute to permit counties to collect additional funds locally and spend those amounts on its schools. The rationale for that recent legislative choice is instructive on the question before the Court today:

'Under existing law, counties are precluded from doing anything in this area, even to spend their own funds if they so desire. This corrective legislation is urgently needed in order to allow counties to go above and beyond the State's standards and provide educational facilities as good as the people of the counties want and are willing to pay for. Allowing local communities to go above and beyond established minimums to provide for their people encourages the best features of democratic government.' Haw.Sess.Laws, 1968, Act 38, § 1.

^103. See text accompanying n. 7, supra.

^104. G. Strayer & R. Haig, The Financing of Education in the State of New York (1923). For a thorough analysis of the contribution of these reformers and of the prior and subsequent history of educational finance, see Coons, Clune & Sugarman, supra, n. 13, at 39—95.

^105. J. Coleman, Forward to Strayer & Haig, supra, at vii.

^106. New State Ice Co. v. Liebmann, 285 U.S. 262, 280, 311, 52 S.Ct. 371, 375, 387, 76 L.Ed. 747 (1932) (Brandeis, J., dissenting).

^107. Mr. Justice WHITE suggests in his dissent that the Texas system violates the Equal Protection Clause because the means it has selected to effectuate its interest in local autonomy fail to guarantee complete freedom of choice to every district. He places special emphasis on the statutory provision that establishes a maximum rate of $1.50 per $100 valuation at which a local school district may tax for school maintenance. Tex.Educ.Code Ann. § 20.04(d) (1972). The maintenance rate in Edgewood when this case was litigated in the District Court was $.55 per $100, barely one-third of the allowable rate. (The tax rate of $1.05 per $100, see supra, at 12, is the equalized

rate for maintenance and for the retirement of bonds.) Appellees do not claim that the ceiling presently bars desired tax increases in Edgewood or in any other Texas district. Therefore, the constitutionality of that statutory provision is not before us and must await litigation in a case in which it is properly presented. Cf. Hargrave v. Kirk, 313 F.Supp. 944 (M.D.Fla.1970), vacated, 401 U.S. 476, 91 S.Ct. 856, 28 L.Ed.2d 196 (1971).

^108. Mr. Justice MARSHALL states in his dissenting opinion that the State's asserted interest in local control is a 'mere sham,' post, at 130, and that it has been offered, not as a legitimate justification, but 'as an excuse . . . for interdistrict inequality.' Id., at 126. In addition to asserting that local control would be preserved and possibly better served under other systems—a consideration that we find irrelevant for the purpose of deciding whether the system may be said to be supported by a legitimate and reasonable basis—the dissent suggests that Texas' lack of good faith may be demonstrated

by examining the extent to which the State already maintains considerable control. The State, we are told, regulates 'the most minute details of local public education,' ibid., including textbook selection, teacher qualifications, and the length of the school day. This assertion, that genuine local control does not exist in Texas, simply cannot be supported. It is abundantly refuted by the elaborate statutory division of responsibilities set out in the Texas Education Code. Although policy decision-making and supervision in certain areas are reserved to the State, the day-to-day authority over the 'management and control' of all public elementary and secondary schools is squarely placed on the local school boards. Tex.Educ.Code Ann. §§ 17.01, 23.26 (1972). Among the innumerable specific powers of the local school authorities are the following: the power of eminent domain to acquire land for the construction of school facilities, id., §§ 17.26, 23.26; the power to hire and terminate teachers and other personnel, id., §§ 13.101—13.103; the power to designate conditions of teacher employment and to establish certain standards of educational policy, id., § 13.901; the power to maintain order and discipline, id., § 21.305, including the prerogative to suspend students for disciplinary reasons, id., § 21.301; the power to decide whether to offer a kindergarten program, id., §§ 21.131—21.135, or a vocational training program, id., § 21.111, or a program of special education for the handicapped, id., § 11.16; the power to control the assignment and transfer of students, id., §§ 21.074—21.080; and the power to operate and maintain a school bus program, id., § 16.52. See also Pervis v. LaMarque Ind. School Dist., 328 F.Supp. 638, 642—643 S.D.Tex.1971), reversed, 466 F.2d 1054 (CA5 1972); Nichols v. Aldine Ind. School Dist., 356 S.W.2d 182 (Tex.Civ.App.1962). Local school boards also determine attendance zones, location of new schools, closing of old ones, school attendance hours (within limits), grading and promotion policies subject to general guidelines, recreational and athletic policies, and a myriad of other matters in the routine of school administration. It cannot be seriously doubted that in Texas education remains largely a local function, and that the preponderating bulk of all decisions affecting the schools is made and executed at the local level, guaranteeing the greatest participation by those most directly concerned.

^109. This theme—that greater state control over funding will lead to greater state power with respect to local educational programs and policies—is a recurrent one in the liternature on financing public education. Professor Simon, in his thoughtful analysis of the political ramifications of this case, states that one of the most likely consequences of the District Court's decision would be an inerease in the centralization of school finance and an increase in the extent of collective bargaining by teacher unions at the state level. He suggests that the subjects for bargaining may include many 'non-salary' items, such as teaching loads, class size, curricular and program choices, questions of student discipline, and selection of administrative personnel—matters traditionally decided heretofore at the local level. Simon, supra, n. 62, at 434—436. See, e.g., Coleman, The Struggle for Control of Education, in Education and Social Policy: Local Control of Education 64, 77—79 (C. Bowers, I. Housego & D. Dyke eds. 1970); J Conant, The Child, The Parent, and The State 27 (1959) ('Unless a local community, through its school board, has some control over the purse, there can be little real feeling in the community that the schools are in fact local schools . . .'); Howe, Anatomy of a Revolution, in Saturday Review 84, 88 (Nov. 20, 1971) ('It is an axiom of American politics that control and power follow money . . .'); R. Hutchinson, State-Administered Locally-Shared Taxes 21 (1931) ('(S)tate administration of taxation is the first step toward state control of the functions supported by these taxes . . .'). Irrespective of whether one regards such prospects as detrimental, or whether he agrees that the consequence is inevitable, it certainly cannot be doubted that there is a rational basis for this concern on the part of parents, educators, and legislators.

^110. This Court has never doubted the propriety of maintaining political subdivisions within the States and has never found in the Equal Protection Clause any per se rule of 'territorial uniformity.' McGowan v. Maryland, 366 U.S., at 427, 81 S.Ct., at 1105. See also Griffin v. County School Board of Prince Edward County, 377 U.S., at 230—231, 84 S.Ct., at 1232 1233; Salsburg v. Maryland, 346 U.S. 545, 74 S.Ct. 280, 98 L.Ed. 281 (1954). Cf. Board of Education of, etc., Muskogee v. Oklahoma, 409 F.2d 665, 668 (CA10 1969).

^111. Any alternative that calls for significant increases in expenditures for education, whether financed through increases in property taxation or through other sources of tax dollars, such as income and sales taxes, is certain to encounter political barriers. At a time when nearly every State and locality is suffering from fiscal undernourishment, and with demands for services of all kinds burgeoning and with weary taxpayers already resisting tax increases, there is considerable reason to question whether a decision of this Court nullifying present state taxing systems would result in a marked increase in the financial commitment to education. See Senate Select Committee on Equal Educational Opportunity, 92d Cong., 2d Sess., Toward Equal Educational Opportunity 339—345 (Comm.Print 1972); Berke & Callahan, Serrano v. Priest: Milestone or Millstone for School Finance, 21 J.Pub.L. 23, 25—26 (1972); Simon, supra, n. 62, at 420 421. In Texas, it has been calculated that $2.4 billion of additional school funds would be required to bring all schools in that State up to the present level of expenditure of all but the wealthiest districts—an amount more than double that currently being spent on education. Texas Research League, supra, n. 20, at 16—18. An amicus curiae brief filed on behalf of almost 30 States, focusing on these practical consequences, claims with some justification that 'each of the undersigned states . . . would suffer severe financial stringency.' Brief of Amici Curiae in Support of Appellants 2 (filed by Montgomery county, Md., et al.).

^112. See Note, supra, n. 53. See also authorities cited n. 114, infra.

^113. See Goldstein, supra, n. 38, at 526; Jencks, supra, n. 86, at 27; U.S. Comm'n on Civil Rights, Inequality in School Financing: The Role of the Law 37 (1972). Coons, Clune & Sugarman, supra, n. 13, at 356—357, n. 47, have noted that in California, for example, (f)ifty-nine percent . . . of minority students live in districts above the median (average valuation per pupil.)' In Bexar County, the largest district by far—the San Antonio Independent School District—is above the local average in both the amount of taxable wealth per pupil and in median family income. Yet 72% of its students are Mexican-Americans. And, in 1967—1968 it spent only a very few dollars less per pupil than the North East and North Side Independent School Districts, which have only 7% and 18% Mexican—American enrollment respectively. Berke, Carnevale, Morgan & White, supra, n. 29, at 673.

^114. See Senate Select Committee on Equal Educational Opportunity, 92d Cong., 2d Sess., Issues in School Finance 129 (Comm.Print 1972) (monograph entitled Inequities in School Finance prepared by Professors Berke and Callhan); U.S. Office of Education, Finances of Large-City School Systems: A Comparative Analysis (1972) (HEW publication); U.S. Comm'n on Civil Rights, supra, n. 113, at 33—36; Simon, supra, n. 62, at 410—411, 418.

^1. See New York Times, Mar. 11, 1973, p. 1, col. 1.

^2. There is one notable exception to the above statement: It has been established in recent years that the Equal Protection Clause confers the substantive right to participate on an equal basis with other qualified voters whenever the State has adopted an electoral process for determining who will represent any segment of the State's population. See, e.g., Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506; Kramer v. Union Free School District, 395 U.S. 621, 89 S.Ct. 1886, 23 L.Ed.2d 583; Dunn v. Blumstein, 405 U.S. 330, 336, 92 S.Ct. 995, 999, 31 L.Ed.2d 274. But there is no constitutional right to vote, as such. Minor v. Happersett, 21 Wall. 162, 22 L.Ed. 627. If there were such a right, both the Fifteenth Amendment and the Nineteenth Amendment would have been wholly unnecessary.

^3. But see Bullock v. Carter, 405 U.S. 134, 92 S.Ct. 849, 31 L.Ed.2d 92.

^4. See Oyama v. California, 332 U.S. 633, 644—646, 68 S.Ct. 269, 274—275, 92 L.Ed. 249.

^5. See Graham v. Richardson, 403 U.S. 365, 372, 91 S.Ct. 1848, 1852, 29 L.Ed.2d 534.

^6. See Griffin v. Illinois, 351 U.S. 12, 76 S.Ct. 585, 100 L.Ed. 891. 'Indigency' means actual or functional indigency; it does not mean comparative poverty vis-a -vis comparative affluence. See James v. Valtierra, 402 U.S. 137, 91 S.Ct. 1331, 28 L.Ed.2d 678.

^7. See Gomez v. Perez, 409 U.S. 535, 93 S.Ct. 872, 35 L.Ed.2d 56; Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 92 S.Ct. 1400, 31 L.Ed.2d 768.

^8. See. e.g., Police Dept. of City of Chicago v. Mosley, 408 U.S. 92, 92 S.Ct. 2286, 33 L.Ed.2d 212 (free speech); Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (freedom of interstate travel); Williams v Rhodes, 393 U.S. 23, 89 S.Ct. 5, 21 L.Ed.2d 24 (freedom of association); Skinner v. Oklahoma, ex rel. Williamson, 316 U.S. 535, 62 S.Ct. 1110, 86 L.Ed. 1655 ('liberty' conditionally protected by Due Process Clause of Fourteenth Amendment).

^9. See Katzenbach v. Morgan, 384 U.S. 641, 660, 86 S.Ct. 1731, 1732, 16 L.Ed.2d 828 (Harlan, J., dissenting).

^1. The heart of the Texas system is embodied in an intricate series of statutory provisions which make up Chapter 16 of the Texas Education Code, Tex.Educ.Code Ann. § 16.01 et seq. See also Tex.Educ.Code Ann. § 15.01 et seq., and § 20.10 et seq.

^2. The figures discussed are from Plaintiffs' Exhibits 7, 8, and 12. The figures are from the 1967—1968 school year. Because the various exhibits relied upon different attendance totals, the per-pupil results do not precisely correspond to the gross figures quoted. The disparity between districts, rather than the actual figures, is the important factor.

^3. Brief for Appellants 11—13, 35.

^4. Variable assessment practices are also revealed in this record. Appellants do not, however, contend that this factor accounts, even to a small extent, for the interdistrict disparities.

^5. The per-pupil funds received from state, federal, and other sources, while not precisely equal, do not account for the large differential and are not directly attacked in the present case.

^6. The State of Texas appears to concede that the choice of whether or not to go beyond the state-provided minimum 'is easier for some districts than for others. Those districts with large amounts of taxable property can produce more revenue at a lower tax rate and will provide their children with a more expensive education.' Brief for Appellants 35. The State nevertheless insists that districts have a choice and that the people in each district have exercised that choice by providing some real property tax money over and above the mimimum funds guaranteed by the State. Like the majority, however, the State fails to explain why the Equal Protection Clause is not violated, or how its goal of providing local government with realistic choices as to how much money should be expended on education is implemented, where the system makes it much more difficult for some than for others to privide additional educational funds and where, as a practical and legal matter, it is impossible for some districts to provide the educational budgets that other districts can make available from real property tax revenues.

^1. See Van Dusartz v. Hatfield, 334 F.Supp. 870, (D.C.Minn.1971); Milliken v. Green, 389 Mich. 1, 203 N.W.2d 457 (1972), rehearing granted, Jan. 1973; Serrano v. Priest, 5 Cal.3d 584, 96 Cal.Rptr. 601, 487 P.2d 1241 (1971); Robinson v. Cahill, 118 N.J.Super. 223, 287 A.2d 187, 119 N.J.Super. 40, 289 A.2d 569 (1972); Hollins v. Shofstall, Civil No. C—253652 (Super.Ct.Maricopa County, Ariz., July 7, 1972). See also Sweetwater County Planning Com. for the Organization of School Districts v. Hinkle, 491 P.2d 1234 (Wyo. 1971), juris. relinquished, 493 P.2d 1050 (Wyo.1972).

^2. The District Court in this case postponed decision for some two years in the hope that the Texas Legislature would remedy the gross disparities in treatment inherent in the Texas financing scheme. It was only after the legislature failed to act in its 1971 Regular Session that the District Court, apparently recognizing the lack of hope for self-initiated legislative reform, rendered its decision. See Texas Research League, Public School Finance Problems in Texas 13 (Interim Report 1972). The strong vested interest of property-rich districts in the existing property tax scheme poses a substantial barrier to self-initiated legislative reform in educational financing. See N.Y. Times, Dec. 19, 1972, p. 1, col. 1.

^3. Texas provides its school districts with extensive bonding authority to obtain capital both for the acquisition of school sites and 'the construction and equipment of school buildings,' Tex.Educ.Code Ann. § 20.01 (1972), and for the acquisition, construction, and maintenance of 'gymnasia, stadia, or other recreational facilities,' id., §§ 20.21—20.22. While such private capital provides a fourth source of revenue, it is, of course, only temporary in nature since the principal and interest of all bonds must ultimately be paid out of the receipts of the local ad valorem property tax, see id., §§ 20.01, 20.04, except to the extent that outside revenues derived from the operation of certain facilities, such as gymnasia, are employed to repay the bonds issued thereon, see id., §§ 20.22, 20.25.

^4. See Tex.Const., Art. 7, § 3; Tex.Educ.Code Ann. §§ 20.01 20.02. As a part of the property tax scheme, bonding authority is conferred upon the local school districts, see n. 3, supra.

^5. See Tex.Educ.Code Ann. § 20.04.

^6. For the 1970—1971 school year, the precise figure was 41.1%. See Texas Research League, supra, n. 2, at 9.

^7. See Tex.Educ.Code Ann. § 20.04.

Theoretically, Texas law limits the tax rate for public school maintenance, see id., § 20.02, to $1.50 per $100 valuation, see id., § 20.04(d). However, it does not appear that any Texas district presently taxes itself at the highest rate allowable, although some poor districts are approaching it, see App. 174.

^8. Under Texas law local districts are allowed to employ differing bases of assessment—a fact that introduces a third variable into the local funding. See Tex.Educ.Code Ann. § 20.03. But neither party has suggested that this factor is responsible for the disparities in revenues available to the various districts. Consequently, I believe we must deal with this case on the assumption that differences in local methods of assessment do not meaningfully affect the revenue-raising power of local districts relative to one another. The Court apparently admits as much. See ante, at 46. It should be noted, moreover, that the main set of data introduced before the District Court to establish the disparities at issue here was based upon 'equalized taxable property' values which had been adjusted to correct for differing methods of assessment. See App. C to Affidavit of Professor Joel S. Berke.

^9. Texas has approximately 1,200 school districts.

^10. See Appendix I, post, p. 134.

^11. See Ibid. Indeed, appellants acknowledge that the relevant data from Professor Berke's affidavit show ' a very positive correlation, 0.973, between market value of taxable property per pupil and state and local revenues per pupil.' Reply Brief for Appellants 6 n. 9.

While the Court takes issue with much of Professor Berke's data and conclusions, ante, at 15—16, n. 38 and 25—27, I do not understand its criticisms to run to the basic finding of a correlation between taxable district property per pupil and local revenues per pupil. The critique of Professor Berke's methodology upon which the Court relies, see Goldstein, Interdistrict Inequalities in School Financing: A Critical Analysis of Serrano v. Priest and its Progeny, 120 U.Pa.L.Rev. 504, 523—525, nn. 67, 71 (1972), is directed only at the suggested correlations between fimily income and taxable district wealth and between race and taxable district wealth. Obviously, the appellants do not question the relationship in Texas between taxable district wealth and per-pupil expenditures; and there is no basis for the Court to do so, whatever the criticisms that may be leveled at other aspects of Professor Berke's study, see infra, n. 55.

^12. See Appendix II, post, p. 135.

^13. See ibid.

^14. For the 1970—1971 school year, the precise figure was 10.9%. See Texas Research League, supra, n. 2, at 9.

^15. Appellants made such a contention before the District Court but apparently have abandoned it in this Court. Indeed, data introduced in the District Court simply belie the argument that federal funds have a significant equalizing effect. See Appendix I, post, p. 134. And, as the District Court observed, it does not follow that remedial action by the Federal Government would excuse any unconstitutional discrimination effected by the state financing scheme. 337 F.Supp. 280, 284.

^16. For the 1970—1971 school year, the precise figure was 48%. See Texas Research League, supra, n. 2, at 9.

^17. See Tex.Const., Art. 7, § 5 (Supp.1972). See also Tex.Educ.Code Ann. § 15.01(b).

^18. See Tex.Educ.Code Ann. § 15.01(b).

The Permanent School Fund is, in essence, a public trust initially endowed with vast quantities of public land, the sale of which has provided an enormous corpus that in turn produces substantial annual revenues which are devoted exclusively to public education. See Tex.Const., Art. 7, § 5 (Supp.1972). See also 5 Report of Governor's Committee on Public School Education, The Challenge and the Chance 11 (1969) (hereinafter Governor's Committee Report).

^19. This is determined from the average daily attendance within each district for the preceding year. Tex.Educ.Code Ann. § 15.01(c).

^20. See id., §§ 16.01—16.975.

^21. See id., §§ 16.71(2), 16.79.

^22. See id., §§ 16.301—16.316, 16.45, 16.51—16.63.

^23. See id., §§ 16.72—16.73, 16.76—16.77.

^24. See id., §§ 16.74—19.76. The formula for calculating each district's share is described in 5 Governor's Committee Report 44 48.

^25. See Tex.Educ.Code Ann. § 16.01.

^26. See 5 Governor's Committee Report 40—41.

^27. See id., at 45—67; Texas Research League, Texas Public Schools Under the Minimum Foundation Program—An Evaluation: 1949 4954, pp. 67—68 (1954).

^28. Technically, the economic index involves a two-step calculation. First, on the basis of the factors mentioned above, each Texas county's share of the Local Fund Assignment is determined. Then each county's share is divided among its school districts on the basis of their relative shares of the county's assessable wealth. See Tex.Educ.Code Ann. §§ 16.74—16.76; 5 Governor's Committee Report 43—44; Texas Research League, Texas Public School Finance: A Majority of Exceptions 6—8 (2d Interim Report 1972).

^29. 5 Governor's Committee Report 48, quoting statement of Dr. Edgar Morphet.

^30. The extraordinarily complex standards are summarized in 5 Governor's Committee Report 41—43.

^31. The key element of the Minimum Foundation School Program is the provision of funds for professional salaries—more particularly, for teacher salaries. The Program provides each district with funds to pay its professional payroll as determined by certain state standards. See Tex.Educ.Code Ann. §§ 16.301 16.316. If the district fails to pay its teachers at the levels determined by the state standards it receives nothing from the Program. See id., § 16.301(c). At the same time, districts are free to pay their teachers salaries in excess of the level set by the state standards, using local revenues—that is, property tax revenue—to make up the difference, see id., § 16.301(a).

The state salary standards focus upon two factors: the educational level and the experience of the district's teachers. See id., §§ 16.301—16.316. The higher these two factors are, the more funds the district will receive from the Foundation Program for professional salaries.

It should be apparent that the net effect of this scheme is to provide more assistance to property-rich districts than to property-poor ones. For rich districts are able to pay their teachers, out of local funds, salary increments above the state minimum levels. Thus, the rich districts are able to attract the teachers with the best education and the most experience. To complete the circle, this then means, given the state standards, that the rich districts receive more from the Foundation Program for professional salaries than do poor districts. A portion of Professor Berke's study vividly illustrates the impact of the State's standards on districts of varying wealth. See Appendix III, post, p. 136.

^32. In 1967—1968, Alamo Heights School District had $49,478 in taxable property per pupil. See Berke Affidavit, Table VII, App. 216.

^33. In 1967—1968, Edgewood Independent School District had $5,960 in taxable property per pupil. Ibid.

^34. I fail to understand the relevance for this case of the Court's suggestion that if Alamo Heights School District, which is approximately the same physical size as Edgewood Independent School District but which has only one-fourth as many students, had the same number of students as Edgewood, the former's per-pupil expenditure would be considerably closer to the latter's. Ante, at 13, n. 33. Obviously, this is true, but it does not alter the simple fact that Edgewood does have four times as many students but not four times as much taxable property wealth. From the perspective of Edgewood's school children then—the perspective that ultimately counts here—Edgewood is clearly a much poorer district than Alamo Heights. The question here is not whether districts have equal taxable property wealth in absolute terms, but whether districts have differing taxable wealth given their respective school-age populations.

^35. In the face of these gross disparities in treatment which experience with the Texas financing scheme has revealed, I cannot accept the Court's suggestion that we are dealing here with a remedial scheme to which we should accord substantial deference because of its accomplishments rather than criticize it for its failures. Ante, at 38—39. Moreover, Texas' financing scheme is hardly remedial legislation of the type for which we have previously shown substantial tolerance. Such legislation may in fact extend the vote to 'persons who otherwise would be denied it by state law,' Katzenbach v. Morgan, 384 U.S. 641, 657, 86 S.Ct. 1717, 1727, 16 L.Ed.2d 828 (1966), or it may eliminate the evils of the private bail bondsman, Schilb v. Kuebel, 404 U.S. 357, 92 S.Ct. 479, 30 L.Ed.2d 502 (1971). But those are instances in which a legislative body has sought to remedy problems for which it cannot be said to have been directly responsible. By contrast, public education is the function of the State in Texas, and the responsibility for any defect in the financing scheme must ultimately rest with the State. It is the State's own scheme which has caused the funding problem, and, thus viewed, that scheme can hardly be deemed remedial.

^36. Cf. Appendix I, post, p. 134.

^37. Brief for Appellants 3.

^38. Thus, in 1967—1968, Edgewood had a total of $248 per pupil in state and local funds compared with a total of $558 per pupil for Alamo Heights. See Berke Affidavit, Table X, App. 219. For 1970—1971, the respective totals were $418 and $913. See Texas Research League, supra, n. 2, at 14.

^39. Not only does the local property tax provide approximately 40% of the funds expended on public education, but it is the only source of funds for such essential aspects of educational financing as the payment of school bonds, see n. 3, supra, and the payment of the district's share of the Local Fund Assignment, as well as for nearly all expenditures above the minimums established by the Foundation School Program.

^40. Compare, e.g., J. Coleman et al., Equality of Educational Opportunity 290—330 (1966); Jencks, The Coleman Report and the Conventional Wisdom, in On Equality of Educational Opportunity 69, 91—104 (F. Mosteller & D. Moynihan eds. 1972), with, e.g., Guthrie, G. Kleindorfer, H. Levin & R. Stout, Schools and inequality 79—90 (1971); Kiesling, Measuring a Local Government Service: A Study of School Districts in New York State, 49 Rev.Econ. & Statistics, 356 (1967).

^41. Compare Berke Answers to Interrogatories 10 ('Dollar expenditures are probably the best way of measuring the quality of education afforded students . . .'), with Graham Deposition 39 ('(I)t is not just necessarily the money, no. It is how wisely you spend it'). It warrants noting that even appellants' witness, Mr. Graham, qualified the importance of money only by the requirement of wise expenditure. Quite obviously, a district which is property poor is powerless to match the education provided by a proterty-rich district, assuming each district allocates its funds with equal wisdom.

^42. See Brief of amici curiae, inter alia, San Marino Unified School District; Beverly Hills Unified School District; Brief of amici curiae, inter alia, Bloomfield Hills, Michigan, School District; Dearborn City, Michigan School District; Grosse Pointe, Michigan, Public School System.

^43. Answers to Plaintiffs' Interrogatories, App. 115.

^44. Ibid. Moreover, during the same period, 37.17% of the teachers in Alamo Heights had advanced degrees, while only 14.98% of Edgewood's faculty had such degrees. See id., at 116.

^45. Id., at 117.

^46. Id., at 118.

^47. In the 1967—1968 school year, Edgewood had 22,862 students and 864 teachers, a ratio of 26.5 to 1. See id., at 110, 114. In Alamo Heights, for the same school year, there were 5,432 students and 265 teachers for a ratio of 20.5 to 1. Ibid.

^48. Reply Brief for Appellants 17. See also, id., at 5, 15 16.

^49. Indeed, even apart from the differential treatment inherent in the local property tax, the significant interdistrict disparties in state aid received under the Minimum Foundation School Program would seem to raise substantial equal protection questions.

^50. I find particularly strong intimations of such a view in the majority's efforts to denigrate the constitutional significance of children in property-poor districts 'receiving a poorer quality education than that available to children in districts having more assessable wealth' with the assertion 'that, at least where wealth is involved, the Equal Protection Clause does not require absolute equality or precisely equal advantages.' Ante, at 23, 24. The Court, to be sure, restricts its remark to 'wealth' discrimination. But the logical basis for such a restriction is not explained by the Court, nor is it otherwise apparent, see infra, at 117—120 and n. 77.

^51. See Answers to Interrogatories by Dr. Joel S. Berke, Ans. 17, p. 9; Ans. 48—51, pp. 22—24; Ans. 88—89, pp. 41—42; Deposition of Dr. Daniel C. Morgan, Jr., at 52—55; Affidavit of Dr. Daniel C. Morgan, Jr., App. 242—243.

^52. It is true that in two previous cases this Court has summarily affirmed district court dismissals of constitutional attacks upon other state educational financing schemes. See McInnis v. Shapiro, 293 F.Supp. 327 (N.D.Ill.1968), aff'd per curiam, sub nom. McInnis v. Ogilvie, 394 U.S. 322, 89 S.Ct. 1197, 22 L.Ed.2d 308 (1969); Burruss v. Wilkerson, 310 F.Supp. 572 (W.D.Va.1969), aff'd per curiam, 397 U.S. 44, 90 S.Ct. 812, 25 L.Ed.2d 37 (1970). But those decisions cannot be considered dispositive of this action, for the thrust of those suits differed materially from that of the present case. In McInnis, the plaintiffs asserted that 'only a financing system which apportions public funds according to the educational needs of the students satisfies the Fourteenth Amendment.' 293 F.Supp., at 331. The District Court concluded that '(1) the Fourteenth Amendment does not require that public school expenditures be made only on the basis of pupils' educational needs, and (2) the lack of judicially manageable standards makes this controversy nonjusticiable.' Id., at 329. The Burruss District Court dismissed that suit essentially in reliance on McInnis which it found to be 'scarcely distinguishable.' 310 F.Supp. at 574. This suit involves no effort to obtain an allocation of school funds that considers only educational need. The District Court rules only that the State must remedy the discrimination resulting from the distribution of taxable local district wealth which has heretofore prevented many districts from truly exercising local fiscal control. Furthermore, the limited holding of the District Court presents none of the problems of judicial management which would exist if the federal courts were to attempt to ensure the distribution of educational funds solely on the basis of educational need, see infra, at 130 132.

^53. Tex.Const., Art. 7, § 1.

^54. Problems of remedy may be another matter. If provision of the relief sought in a particular case required identification of each member of the affected class, as in the case of monetary relief, the need for clarity in defining the class is apparent. But this involves the procedural problems inherent in class action litigation, not the character of the elements essential to equal protection analysis. We are concerned here only with the latter. Moreover, it is evident that in cases such as this, provision of appropriate relief, which takes the injunctive form, is not a serious problem since it is enough to direct the action of appropriate officials. Cf. Potts v. Flax, 313 F.2d 284, 288—290 (CA5 1963).

^55. I assume the Court would lodge the same criticism against the validity of the finding of a correlation between poor districts and racial minorities.

^56. The Court rejects the District Court's finding of a correlation between poor people and poor districts with the assertion that 'there is reason to believe that the poorest families are not necessarily clustered in the poorest property districts' in Texas. Ante, at 23. In support of its conclusion the Court offers absolutely no data—which it cannot on this record concerning the distribution of poor people in Texas to refute the data introduced below by appellees; it relies instead on a recent law review note concerned solely with the State of Connecticut, Note, A Statistical Analysis of the School Finance Decisions: On Winning Battles and Losing Wars, 81 Yale L.J. 1303 (1972). Common sense suggests that the basis for drawing a demographic conclusion with respect to a geographically large, urban-rural, industrial-agricultural State such as Texas from a geographically small, densely populated, highly industrialized State such an Connecticut is doubtful at best.

Furthermore, the article upon which the Court relies to discredit the statistical procedures employed by Professor Berke to establish the correlation between poor people and poor districts, see n. 11, supra, based its criticism primarily on the fact that only four of the 110 districts studied were in the lowest of the five categories, which were determined by relative taxable property per pupil, and most districts clustered in the middle three groups. See Goldstein, Interdistrict Inequalities in School Financing: A Critical Analysis of Serrano v. Priest and its Progeny, 120 U.Pa.L.Rev. 504, 524 n. 67 (1972). See also ante, at 26—27. But the Court fails to note that the four poorest districts in the sample had over 50,000 students which constituted 10% of the students in the entire sample. It appears, moreover, that even when the richest and the poorest categories are enlarged to include in each category 20% of the students in the sample, the correlation between district and individual wealth holds true. See Brief for the Governors of Minnesota, Maine, South Dakota, Wisconsin, and Michigan as amici curiae 17 n. 21.

Finally, it cannot be ignored that the data introduced by appellees went unchallenged in the District Court. The majority's willingness to permit appellants to litigate the correctness of those data for the first time before this tribunal—where effective response by appellees is impossible—is both unfair and judicially unsound.

^57. Third Amended Complaint App. 23. Consistent with this theory, appellees purported to represent, among others, a class composed of 'all . . . school children in independent school districts . . . who . . . have been deprived of the equal protection of the law under the Fourteenth Amendment with regard to public school education because of the low value of the property lying within the independent school districts in which they reside.' Id., at 15.

^58. The degree of judicial scrutiny that this particular classification demands is a distinct issue which I consider in Part II, C, infra.

^59. Indeed, the Court's theory would render the established concept of fundamental interests in the context of equal protection analysis superfluous, for the substantive constitutional right itself requires that this Court strictly scrutinize any asserted state interest for restricting or denying access to any particular guaranteed right, see, e.g., United States v. O'Brien, 391 U.S. 367, 377, 88 S.Ct. 1673, 1679, 20 L.Ed.2d 672 (1968); Cox v. Louisiana, 379 U.S. 536, 545—551, 85 S.Ct. 453, 459—463, 13 L.Ed.2d 471 (1965).

^60. It is interesting that in its effort to reconcile the state voting rights cases with its theory of fundamentality the majority can muster nothing more than the contention that '(t)he constitutional underpinnings of the right to equal treatment in the voting process can no longer be doubted . . ..' Ante, at 34 n. 74 (emphasis added). If, by this, the Court intends to recognize a substantive constitutional 'right to equal treatment in the voting process' independent of the Equal Protection Clause, the source of such a right is certainly a mystery to me.

^61. It is true that Griffin and Douglas also involved discrimination against indigents, that is, wealth discrimination. But, as the majority points out, ante, at 28—29, the Court has never deemed wealth discrimination alone to be sufficient to require strict judicial scrutiny; rather, such review of wealth classifications has been applied only where the discrimination affects an important individual interest, see, e.g., Harper v. Virginia Board of Elections, 383 U.S. 663, 86 S.Ct. 1079, 16 L.Ed.2d 169 (1966). Thus, I believe Griffin and Douglas can only be understood as premised on a recognition of the fundamental importance of the criminal appellate process.

^62. See, e.g., Duncan v. Louisiana, 391 U.S., 145, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968) (right to jury trial); Washington v. Texas, 388 U.S. 14, 87 S.Ct. 1920, 18 L.Ed.2d 1019 (1967) (right to compulsory process); Pointer v. Texas, 380 U.S. 400, 85 S.Ct. 1065, 13 L.Ed.2d 923 (1965) (right to confront one's accusers).

^63. See, e.g., McLaughlin v. Florida, 379 U.S. 184, 191—192, 85 S.Ct. 283, 287—289, 13 L.Ed.2d 222 (1964); Loving v. Virginia, 388 U.S. 1, 9, 87 S.Ct. 1817, 1822, 18 L.Ed.2d 1010 (1967).

^64. See Oyama v. California, 332 U.S. 633, 644—646, 68 S.Ct. 269, 274—275, 92 L.Ed. 249 (1948); Korematsu v. United States, 323 U.S. 214, 216, 65 S.Ct. 193, 194, 89 L.Ed. 194 (1944).

^65. See Graham v. Richardson, 403 U.S. 365, 372, 91 S.Ct. 1848, 1852, 29 L.Ed.2d 534 (1971).

^66. The Court noted that the challenged 'provision strips from indigent defendants the array of protective exemptions Kansas has erected for other civil judgment debtors, including restrictions on the amount of disposable earnings subject to garnishment, protection of the debtor from wage garnishment at times of severe personal or family sickness, and exemption from attachment and execution on a debtor's personal clothing, books and tools of trade.' 407 U.S., at 135, 92 S.Ct., at 2031.

^67. See generally Gunther, The Supreme Court, 1971 Term, Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection, 86 Harv.L.Rev. 1 (1972).

^68. See Brief of the National Education Association et al. as amici curiae App. A. All 48 of the 50 States which mandate public education also have compulsory-attendance laws which require school attendance for eight years or more. Id., at 20—21.

^69. Prior to this Court's decision in Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), every State had a constitutional provision directing the establishment of a system of public schools. But after Brown, South Carolina repealed its constitutional provision, and Mississippi made its constitutional provision discretionary with the state legislature.

^70. Developments in the Law—Equal Protection, 82 Harv.L.Rev. 1065, 1129 (1969).

^71. The President's Commission on School Finance, Schools, People, Money: The Need for Educational Reform 11 (1972), concluded that '(l)iterally, we cannot survive as a nation or as individuals without (education).' It further observed that:

'(I)n a democratic society, public understanding of public issues is necessary for public support. Schools generally include in their courses of instruction a wide variety of subjects related to the history, structure and principles of American government at all levels. In so doing, schools provide students with a background of knowledge which is deemed an absolute necessity for responsible citizenship.' Id., at 13—14.

^72. See J. Guthrie, G. Kleindorfer, H. Levin, & R. Stout, Schools and Inequality 103—105 (1971); R. Hess & J. Torney, The Development of Political Attitudes in Children 217—218 (1967); Campbell, The Passive Citizen, in 6 Acta Sociologica, Nos. 1—2, p. 9, at 20—21 (1962).

That education is the dominant factor in influencing political participation and awareness is sufficient, I believe, to dispose of the Court's suggestion that, in all events, there is no indication that Texas is not providing all of its children with a sufficient education to enjoy the right of free speech and to participate fully in the political process. Ante, at 36—37. There is, in short, no limit on the amount of free speech or political participation that the Constitution guarantees. Moreover, it should be obvious that the political process, like most other aspects of social intercourse, is to some degree competitive. It is thus of little benefit to an individual from a property-poor district to have 'enough' education if those around him have more than 'enough.' Cf. Sweatt v. Painter, 339 U.S. 629, 633—634, 70 S.Ct. 848, 849, 850, 94 L.Ed. 1114 (1950).

^73. See United States Department of Commerce, Bureau of the Census, Voting and Registration in the Election of November 1968, Current Population Reports, Series P—20, No. 192, Table 4, p. 17. See also Senate Select Committee on Equal Educational Opportunity, 92d Cong., 2d Sess., Levin, The Costs to the Nation of Inadequate Education 46—47 (Comm.Print 1972).

^74. I believe that the close nexus between education and our established constitutional values with respect to freedom of speech and participation in the political process makes this a different case from our prior decisions concerning discrimination affecting public welfare, see, e.g., Dandridge v. Williams, 397 U.S. 471, 90 S.Ct. 1153, 25 L.Ed.2d 491 (1970), or housing, see, e.g., Lindsey v. Normet, 405 U.S. 56, 92 S.Ct. 862, 31 L.Ed.2d 36 (1972). There can be no question that, as the majority suggests, constitutional rights may be less meaningful for someone without enough to eat or without decent housing. Ante, at 37. But the crucial difference lies in the closeness of the relationship. Whatever the severity of the impact of insufficient food or inadequate housing on a person's life, they have never been considered to bear the same direct and immediate relationship to constitutional concerns for free speech and for our political processes as education has long been recognized to bear. Perhaps, the best evidence of this fact is the unique status which has been accorded public education as the single public service nearly unanimously guaranteed in the constitutions of our States, see supra, at 111—112 and n. 68. Education, in terms of constitutional values, is much more analogous in my judgment, to the right to vote in state elections than to public welfare or public housing. Indeed, it is not without significance that we have long recognized education as an essential step in providing the disadvantaged with the tools necessary to achieve economic self-sufficiency.

^75. The majority's reliance on this Court's traditional deference to legislative bodies in matters of taxation falls wide of the mark in the context of this particular case. See ante, at 40—41. The decisions on which the Court relies were simply taxpayer suits challenging the constitutionality of a tax burden in the face of exemptions or differential taxation afforded to others. See, e.g., Allied Stores of Ohio, Inc. v. Bowers, 358 U.S. 522, 79 S.Ct. 437, 3 L.Ed.2d 480 (1959); Madden v. Kentucky, 309 U.S. 83, 60 S.Ct. 406, 84 L.Ed. 590 (1940); Carmichael v. Southern Coal & Coke Co., 301 U.S. 495, 57 S.Ct. 868, 81 L.Ed. 1245 (1937); Bell's Gap R. Co. v. Pennsylvania, 134 U.S. 232, 10 S.Ct. 533, 33 L.Ed. 892 (1890). There is no question that, from the perspective of the taxpayer, the Equal Protection Clause 'imposes no iron rule of equality, prohibiting the flexibility and variety that are appropriate to reasonable schemes of state taxation. The State may impose different specific taxes upon different trades and professions and may vary the rate of excise upon various products.' Allied Stores of Ohio, Inc. v. Bowers, supra, 358 U.S., at 526—527, 79 S.Ct., at 440—441. But in this case we are presented with a claim of discrimination of an entirely different nature—a claim that the revenue-producing mechanism directly discriminates against the interests of some of the intended beneficiaries; and, in contrast to the taxpayer suits, the interest adversely affected is of substantial constitutional and societal importance. Hence, a different standard of equal protection review than has been employed in the taxpayer suits is appropriate here. It is true that affirmance of the District Court decision would to some extent intrude upon the State's taxing power insofar as it would be necessary for the State to at least equalize taxable district wealth. But contrary to the suggestions of the majority, affirmance would not impose a strait jacket upon the revenue-raising powers of the State, and would certainly not spell the end of the local property tax. See infra, at 1347.

^76. This does not mean that the Court has demanded precise equality in the treatment of the indigent and the person of means in the criminal process. We have never suggested, for instance, that the Equal Protection Clause requires the best lawyer money can buy for the indigent. We are hardly equipped with the objective standards which such a judgment would require. But we have pursued the goal of substantial equality of treatment in the face of clear disparities in the nature of the appellate process afforded rich versus poor. See, e.g., Draper v. Washington, 372 U.S. 487, 495—496, 83 S.Ct. 774, 778—779, 9 L.Ed.2d 899 (1963); cf. Coppedge v. United States, 369 U.S. 438, 447, 82 S.Ct. 917, 922, 8 L.Ed.2d 21 (1962).

^77. Even if I put side the Court's misreading of Griffin and Douglas, the Court fails to offer any reasoned constitutional basis for restricting cases involving wealth discrimination to instances in which there is an absolute deprivation of the interest affected. As I have already discussed, see supra at 88 89, the Equal Protection Clause guarantees equality of treatment of those persons who are similarly situated; it does not merely bar some form of excessive discrimination between such persons. Outside the context of wealth discrimination, the Court's reapportionment decisions clearly indicate that relative discrimination is within the purview of the Equal Protection Clause. Thus, in Reynolds v. Sims, 377 U.S. 533, 562—563, 84 S.Ct. 1362, 1382, 12 L.Ed.2d 506 (1964), the Court recognized:

'It would appear extraordinary to suggest that a State could be constitutionally permitted to enact a law providing that certain of the State's voters could vote two, five, or 10 times for their legislative representatives, while voters living elsewhere could vote only once. . . . Of course, the effect of state legislative districting schemes which give the same number of representatives to unequal numbers of constituents is identical. Overweighting and overvaluation of the votes of those living here has the certain effect of dilution and undervaluation of the votes of those living there. . . . Their right to vote is simply not the same right to vote as that of those living in a favored part of the State. . . . One must be ever aware that the Constitution forbids 'sophisticated as well as simple-minded modes of discrimination." See also Gray v. Sanders, 372 U.S. 368, 380 381, 83 S.Ct. 801, 808—809, 9 L.Ed.2d 821 (1963). The Court gives no explanation why a case involving wealth discrimination should be treated any differently.

^78. But cf. Bullock v. Carter, 405 U.S. 134, 144, 92 S.Ct. 849, 856, 31 L.Ed.2d 92 (1972), where prospective candidates' threatended exclusion from a primary ballot because of their inability to pay a filing fee was seen as discrimination against both the impecunious candidates and the 'less affluent segment of the community' that supported such candidates but was also too poor as a group to contribute enough for the filing fees.

^79. But cf. M. Harrington, The Other America 13—17 (Penguin ed. 1963).

^80. See E. Banfield, The Unheavenly City 63, 75—76 (1970); cf. R. Lynd & H. Lynd, Middletown in Transition 450 (1937).

^81. Cf. City of New York v. Miln, 11 Pet. 102, 142, 9 L.Ed. 648 (1837).

^82. Theoretically, at least, it may provide a mechanism for implementing Texas' asserted interest in local educational control, see infra, at 126.

^83. True, a family may move to escape a property-poor school district, assuming it has the means to do so. But such a view would itself raise a serious constitutional question concerning an impermissible burdening of the right to travel, or, more precisely, the concomitant right to remain where one is. Cf. Shapiro v. Thompson, 394 U.S. 618, 629—631, 89 S.Ct. 1322, 1328 1330, 22 L.Ed.2d 600 (1969).

^84. Indeed, the political difficulties that seriously disadvantaged districts face in securing legislative redress are augmented by the fact that little support is likely to be secured from only mildly disadvantaged districts. Cf. Gray v. Sanders, 372 U.S. 368, 83 S.Ct. 801, 9 L.Ed.2d 821 (1963). See also n. 2, supra.

^85. See Tex. Cities, Towns and Villages Code, Civ.Stat.Ann. §§ 1011a—1011j (1963 and Supp.1972—1973). See also, e.g., Skinner v. Reed, 265 S.W.2d 850 (Tex.Civ.App.1954); City of Corpus Christi v. Jones, 144 S.W.2d 388 (Tex.Civ.App.1940).

^86. Serrano v. Priest, 5 Cal.3d, at 603, 96 Cal.Rptr., at 614, 487 P.2d, at 1254. See also Van Dusartz v. Hatfield, 334 F.Supp., at 875—876.

^87. Cf., e.g., Two Guys from Harrison-Allentown, Inc. v. McGinley, 366 U.S. 582, 81 S.Ct. 1135, 6 L.Ed.2d 551 (1961); McGowan v. Maryland, 366 U.S. 420, 81 S.Ct. 1101, 6 L.Ed.2d 393 (1961); Goesaert v. Cleary, 335 U.S. 464, 69 S.Ct. 198, 93 L.Ed. 163 (1948).

^88. Tex.Educ.Code Ann. §§ 21.101—21.117. Criminal penalties are provided for failure to teach certain required courses. Id., §§ 4.15—4.16.

^89. Id., §§ 12.11—12.35.

^90. Id., § 12.62.

^91. Id., §§ 13.031—13.046.

^92. Id., § 21.004.

^93. See Appendix II, infra.

^94. See Affidavit of Dr. Jose Cardenas, Superintendent of Schools, edgewood Independent School District, App. 234—238.

^95. See Appendix IV, infra.

^96. My Brother WHITE, in concluding that the Texas financing scheme runs afoul of the Equal Protection Clause, likewise finds on analysis that the means chosen by Texas—local property taxation dependent upon local taxable wealth—is completely unsuited in its present form to the achievement of the asserted goal of providing local fiscal control. Although my Brother WHITE purports to reach this result by application of that lenient standard of mere rationality traditionally applied in the context of commercial interest, it seems to me that the care with which he scrutinizes the practical effectiveness of the present local property tax as a device for affording local fiscal control reflects the application of a more stringent standard of review, a standard which at the least is influenced by the constitutional significance of the process of public education.

^97. See n. 98, infra.

^98. Centralized educational financing is, to be sure, one alternative. On analysis, though, it is clear that even centralized financing would

not deprive local school district of what has been considered to be the essence of local educational control. See Wright v. Council of the City of Emporia, 407 U.S. 451, 477—478, 92 S.Ct. 2196, 2210 2211, 33 L.Ed.2d 51 (Burger, C.J., dissenting). Central financing would leave in local hands the entire gamut of local educational policy-making—teachers, curriculum, school sites, the whole process of allocating resources among alternative educational objectives.

A second possibility is the much-discussed theory of district power equalization put forth by Professors Coons, Clune, and Sugarman in their seminal work, Private Wealth and Public Education 201—242 (1970). Such a scheme would truly reflect a dedication to local fiscal control. Under their system, each school district would receive a fixed amount of revenue per pupil for any particular level of tax effort regardless of the level of local property tax base. Appellants criticize this scheme on the rather extraordinary ground that it would encourage poorer districts to overtax themselves in order to obtain substantial revenues for education. But under the present discriminatory scheme, it is the poor districts that are already taxing themselves at the highest rates, yet are receiving the lowest returns.

District wealth reapportionment is yet another alternative which would accomplish directly essentially what district power equalization would seek to do artificially. Appellants claim that the calculations concerning state property required by such a scheme would be impossible as a practical matter. Yet Yexas is already making far more complex annual calculations—involving not only local property values but also local income and other economic factors—in conjunction with the Local Fund Assignment portion of the Minimum Foundation School Program. See 5 Governor's Committee Report 43—44.

A fourth possibility would be to remove commercial, industrial, and mineral property from local tax rolls, to tax this property on a statewide basis, and to return the resulting revenues to the local districts in a fashion that would compensate for remaining variations in the local tax bases.

None of these particular alternatives are necessarily constitutionally compelled; rather, they indicate the breadth of choice which would remain to the State if the present interdistrict disparities were eliminated.

^99. See n. 98, supra.

^100. Of course, nothing in the Court's decision today should inhibit further review of state educational funding schemes under state constitutional provisions. See Milliken v. Green, 389 Mich. 1, 203 N.W.2d 457 (1972), rehearing granted, Jan. 1973; Robinson v. Cahill, 118 N.J. Super. 223, 287 A.2d 187; 119 N.J.Super. 40, 289 A.2d 569 (1972); cf. Serrano v. Priest, 5 Cal.3d 584, 96 Cal.Rptr. 601, 487 P.2d 1241 (1971).

Based on Table V to affidavit of Joel S. Berke, App. 208, which was prepared on the basis of a sample of 110 selected Texas school districts from data for the 1967-1968 school year.

Based on Table II to affidavit of Joel S. Berke, App. 205, which was prepared on the basis of a sample of 110 selected Texas school districts from data for the 1967-1968 school year.

Based on Table XI to affidavit of Joel S. Berke, App. 220, which was prepared on the basis of a sample of six selected school districts located in Bexar County, Texas, from data for the 1967-1968 school year.

Based on Table IX to affidavit of Joel S. Berke, App. 218, which was prepared on the basis of the 12 school districts located in Bexar County, Texas, from data from the 1967-1968 school year.

Curtis Craig et al., Appellants, v. David Boren, etc., et al.

Argued Oct. 5, 1976.

Decided Dec. 20, 1976.

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

Syllabus

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

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

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

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

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

3. Oklahoma's gender-based differential constitutes an invidious discrimination against males 18-20 years of age in violation of the Equal Protection Clause. Appellees' statistics (the most relevant of which*191 show only that .18% of females and 2% of males in the 18-20-year-old age group were arrested for driving while under the influence of liquor) do not warrant the conclusion that sex represents an accurate proxy for the regulation of drinking and driving. Pp. 199-204.

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

399 F.Supp. 1304, reversed.

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

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

Mr. Justice BRENNAN delivered the opinion of the Court.

The interaction of two sections of an Oklahoma statute, Okla.Stat., Tit. 37, §§ 241 and 245 (1958 and Supp.1976),1*192 prohibits the sale of "nonintoxicating" 3.2% beer to males under the age of 21 and to females under the age of 18. The question to be decided is whether such a gender-based differential constitutes a denial to males 18-20 years of age of the equal protection of the laws in violation of the Fourteenth Amendment.

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

I

We first address a preliminary question of standing. Appellant Craig attained the age of 21 after we noted probable jurisdiction. Therefore, since only declaratory and injunctive relief against enforcement of the gender-based differential is sought, the controversy has been rendered moot as to Craig. See, e. g., DeFunis v. Odegaard, 416 U.S. 312, 94 S.Ct. 1704, 40 L.Ed.2d 164 (1974).2 The question thus arises whether appellant Whitener, the licensed vendor of 3.2% beer, who has a live controversy against enforcement of the statute, may rely upon the equal protection objections of males 18-20 years of age to establish her claim of*193 unconstitutionality of the age-sex differential. We conclude that she may.

Initially, it should be noted that, despite having had the opportunity to do so,3 appellees never raised before the District Court any objection to Whitener's reliance upon the claimed unequal treatment of 18-20-year-old males as the premise of her equal protection challenge to Oklahoma's 3.2% beer law. See 399 F.Supp., at 1306 n. 1. Indeed, at oral argument Oklahoma acknowledged that appellees always "presumed" that the vendor, subject to sanctions and loss of license for violation of the statute, was a proper party in interest to object to the enforcement of the sex-based regulatory provision. Tr. of Oral Arg. 41. While such a concession certainly would not be controlling upon the reach of this Court's constitutional authority to exercise jurisdiction under Art. III, see, e. g., Sierra Club v. Morton, 405 U.S. 727, 732 n. 3, 92 S.Ct. 1361, 1365, 31 L.Ed.2d 636 (1972); cf. Association of Data Processing Service Organizations, Inc. v. Camp, 397 U.S. 150, 151, 90 S.Ct. 827, 829, 25 L.Ed.2d 184 (1970), our decisions have settled that limitations on a litigant's assertion of jus tertii are not constitutionally mandated, but rather stem from a salutary "rule of self-restraint" designed to minimize unwarranted intervention into controversies where the applicable constitutional questions are ill-defined and speculative. See, e. g., Barrows v. Jackson, 346 U.S. 249, 255, 257, 73 S.Ct. 1031, 1034, 1035, 97 L.Ed. 1586 (1953); see also Singleton v. Wulff, 428 U.S. 106, 123-124, 96 S.Ct. 2868, 2877, 49 L.Ed.2d 826 (1976) (Powell, J., dissenting). These prudential objectives, thought to be enhanced by restrictions on third-party standing, cannot be furthered here, where the lower court already has entertained the relevant constitutional challenge and the parties have sought or at least have never resisted an authoritative constitutional determination. In such circumstances, a decision by us to forgo*194 consideration of the constitutional merits in order to await the initiation of a new challenge to the statute by injured third parties would be impermissibly to foster repetitive and time-consuming litigation under the guise of caution and prudence. Moreover, insofar as the applicable constitutional questions have been and continue to be presented vigorously and "cogently," Holden v. Hardy, 169 U.S. 366, 397, 18 S.Ct. 383, 390, 42 L.Ed. 780 (1898), the denial of jus tertii standing in deference to a direct class suit can serve no functional purpose. Our Brother Blackmun's comment is pertinent: "(I)t may be that a class could be assembled, whose fluid membership always included some (males) with live claims. But if the assertion of the right is to be 'representative' to such an extent anyway, there seems little loss in terms of effective advocacy from allowing its assertion by" the present jus tertii champion. Singleton v. Wulff, supra, at 117-118, 96 S.Ct. at 2875.

In any event, we conclude that appellant Whitener has established independently her claim to assert jus tertii standing. The operation of §§ 241 and 245 plainly has inflicted "injury in fact" upon appellant sufficient to guarantee her "concrete adverseness," Baker v. Carr, 369 U.S. 186, 204, 82 S.Ct. 691, 703, 7 L.Ed.2d 663 (1962), and to satisfy the constitutionally based standing requirements imposed by Art. III. The legal duties created by the statutory sections under challenge are addressed directly to vendors such as appellant. She is obliged either to heed the statutory discrimination, thereby incurring a direct economic injury through the constriction of her buyers' market, or to disobey the statutory command and suffer, in the words of Oklahoma's Assistant Attorney General, "sanctions and perhaps loss of license." Tr. of Oral Arg. 41. This Court repeatedly has recognized that such injuries establish the threshold requirements of a "case or controversy" mandated by Art. III. See, e. g., Singleton v. Wulff, supra, at 113, 96 S.Ct., at 2873 (doctors who receive payments for their abortion services are "classically adverse" to government as payer); Sullivan v. Little Hunting*195 Park, 396 U.S. 229, 237, 90 S.Ct. 400, 404, 24 L.Ed.2d 386 (1969); Barrows v. Jackson, supra, 346 U.S., at 255-256, 73 S.Ct., at 1034-1035.

As a vendor with standing to challenge the lawfulness of §§ 241 and 245, appellant Whitener is entitled to assert those concomitant rights of third parties that would be "diluted or adversely affected" should her constitutional challenge fail and the statutes remain in force. Griswold v. Connecticut, 381 U.S. 479, 481, 85 S.Ct. 1678, 1679, 14 L.Ed.2d 510 (1965); see Note, Standing to Assert Constitutional Jus Tertii, 88 Harv.L.Rev. 423, 432 (1974). Otherwise, the threatened imposition of governmental sanctions might deter appellant Whitener and other similarly situated vendors from selling 3.2% beer to young males, thereby ensuring that "enforcement of the challenged restriction against the (vendor) would result indirectly in the violation of third parties' rights." Warth v. Seldin, 422 U.S. 490, 510, 95 S.Ct. 2197, 2211, 45 L.Ed.2d 343 (1975). Accordingly, vendors and those in like positions have been uniformly permitted to resist efforts at restricting their operations by acting as advocates of the rights of third parties who seek access to their market or function. See, e. g., Eisenstadt v. Baird, 405 U.S. 438, 92 S.Ct. 1029, 31 L.Ed.2d 349 (1972); Sullivan v. Little Hunting Park, supra ; Barrows v. Jackson, supra.4*196 Indeed, the jus tertii question raised here is answered by our disposition of a like argument in Eisenstadt v. Baird, supra. There, as here, a state statute imposed legal duties and disabilities upon the claimant, who was convicted of distributing a package of contraceptive foam to a third party.5 Since the statute was directed at Baird and penalized his conduct, the Court did not hesitate again as here to conclude that the "case or controversy" requirement of Art. III was satisfied. 405 U.S., at 443, 92 S.Ct., at 1033. In considering Baird's constitutional objections, the Court fully recognized his standing to defend the privacy interests of third parties. Deemed crucial to the decision to permit jus tertii standing was the recognition of "the impact of the litigation on the third-party interests." Id., at 445, 92 S.Ct., at 1034. Just as the defeat of Baird's suit and the "(e)nforcement of the Massachusetts statute will materially impair the ability of single persons to obtain contraceptives," id., at 446, 92 S.Ct., at 1034, so too the failure of Whitener to prevail in this suit and the continued enforcement of §§ 241 and 245 will "materially impair the ability of" males 18-20 years of age to purchase 3.2% beer despite their classification by an overt gender-based criterion. Similarly, just as the Massachusetts law in Eisenstadt "prohibit(ed),*197 not use, but distribution," 405 U.S., at 446, 92 S.Ct., at 1034, and consequently the least awkward challenger was one in Baird's position who was subject to that proscription, the law challenged here explicitly regulates the sale rather than use of 3.2% beer, thus leaving a vendor as the obvious claimant.

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

II

A.

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

Analysis may appropriately begin with the reminder that Reed emphasized that statutory classifications that distinguish between males and females are "subject to scrutiny under the Equal Protection Clause." 404 U.S., at 75, 92 S.Ct., at 253. To withstand constitutional challenge, previous cases establish that classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives. Thus, in Reed, the ob*198jectives of "reducing the workload on probate courts," id., at 76, 92 S.Ct., at 254, and "avoiding intrafamily controversy," id., at 77, 92 S.Ct., at 254, were deemed of insufficient importance to sustain use of an overt gender criterion in the appointment of administrators of intestate decedents' estates. Decisions following Reed similarly have rejected administrative ease and convenience as sufficiently important objectives to justify gender-based classifications. See, e. g., Stanley v. Illinois, 405 U.S. 645, 656, 92 S.Ct. 1208, 1215, 31 L.Ed.2d 551 (1972); Frontiero v. Richardson, 411 U.S. 677, 690, 93 S.Ct. 1764, 1772, 36 L.Ed.2d 583 (1973); cf. Schlesinger v. Ballard, 419 U.S. 498, 506-507, 95 S.Ct. 572, 576-577, 42 L.Ed.2d 610 (1975). And only two Terms ago, Stanton v. Stanton, 421 U.S. 7, 95 S.Ct. 1373, 43 L.Ed.2d 688 (1975), expressly stating that Reed v. Reed was "controlling," 421 U.S., at 13, 95 S.Ct., at 1377, held that Reed required invalidation of a Utah differential age-of-majority statute, notwithstanding the statute's coincidence with and furtherance of the State's purpose of fostering "old notions" of role typing and preparing boys for their expected performance in the economic and political worlds. 421 U.S., at 14-15, 95 S.Ct., at 1378.6

Reed v. Reed has also provided the underpinning for decisions that have invalidated statutes employing gender as an inaccurate proxy for other, more germane bases of classification. Hence, "archaic and overbroad" generalizations, Schlesinger v. Ballard, supra, 419 U.S., at 508, 95 S.Ct., at 577, concerning the financial position of servicewomen, Frontiero v. Richardson, supra, 411 U.S., at 689 n. 23, 93 S.Ct., at 1772, and working women, Weinberger v. Wiesenfeld, 420 U.S. 636, 643, 95 S.Ct. 1225, 1230, 43 L.Ed.2d 514 (1975), could not justify use of a gender line in determining eligibility for certain governmental entitlements. Similarly, increasingly outdated*199 misconceptions concerning the role of females in the home rather than in the "marketplace and world of ideas" were rejected as loose-fitting characterizations incapable of supporting state statutory schemes that were premised upon their accuracy. Stanton v. Stanton, supra ; Taylor v. Louisiana, 419 U.S. 522, 535 n. 17, 95 S.Ct. 692, 700, 42 L.Ed.2d 690 (1975). In light of the weak congruence between gender and the characteristic or trait that gender purported to represent, it was necessary that the legislatures choose either to realign their substantive laws in a gender-neutral fashion, or to adopt procedures for identifying those instances where the sex-centered generalization actually comported with fact. See, e. g., Stanley v. Illinois, supra, 405 U.S., at 658, 92 S.Ct., at 1216; cf. Cleveland Board of Education v. LaFleur, 414 U.S. 632, 650, 94 S.Ct. 791, 801, 39 L.Ed.2d 52 (1974).

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

B

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

C

We accept for purposes of discussion the District Court's identification of the objective underlying §§ 241 and 245 as the enhancement of traffic safety. 7 Clearly, the protection*200 of public health and safety represents an important function of state and local governments. However, appellees' statistics in our view cannot support the conclusion that the gender-based distinction closely serves to achieve that objective and therefore the distinction cannot under Reed withstand equal protection challenge.

The appellees introduced a variety of statistical surveys. First, an analysis of arrest statistics for 1973 demonstrated that 18-20-year-old male arrests for "driving under the influence" and "drunkenness" substantially exceeded female arrests for that same age period.8 Similarly, youths aged 17-21 were found to be overrepresented among those killed*201 or injured in traffic accidents, with males again numerically exceeding females in this regard.9 Third, a random roadside survey in Oklahoma City revealed that young males were more inclined to drive and drink beer than were their female counterparts. 10Fourth, Federal Bureau of Investigation nationwide statistics exhibited a notable increase in arrests for "driving under the influence." 11 Finally, statistical evidence gathered in other jurisdictions, particularly Minnesota and Michigan, was offered to corroborate Oklahoma's experience by indicating the pervasiveness of youthful participation in motor vehicle accidents following the imbibing of alcohol. Conceding that "the case is not free from doubt," 399 F.Supp., at 1314, the District Court nonetheless concluded that this statistical showing substantiated "a rational basis for the legislative judgment underlying the challenged classification." Id., at 1307.

Even were this statistical evidence accepted as accurate, it nevertheless offers only a weak answer to the equal protection question presented here. The most focused and relevant of the statistical surveys, arrests of 18-20-year-olds for alcohol-related driving offenses, exemplifies the ultimate unpersuasiveness of this evidentiary record. Viewed in terms of the correlation between sex and the actual activity that Oklahoma seeks to regulate driving while under the influence of alcohol the statistics broadly establish that .18% of females and 2% of males in that age group were arrested for that offense. While such a disparity is not trivial in a statistical sense, it hardly can form the basis for employment of a gender line as a classifying device. Certainly if male*202ness is to serve as a proxy for drinking and driving, a correlation of 2% must be considered an unduly tenuous "fit." 12 Indeed, prior cases have consistently rejected the use of sex as a decisionmaking factor even though the statutes in question certainly rested on far more predictive empirical relationships than this.13

Moreover, the statistics exhibit a variety of other shortcomings that seriously impugn their value to equal protection analysis. Setting aside the obvious methodological problems,14 the surveys do not adequately justify the salient*203 features of Oklahoma's gender-based traffic-safety law. None purports to measure the use and dangerousness of 3.2% beer as opposed to alcohol generally, a detail that is of particular importance since, in light of its low alcohol level, Oklahoma apparently considers the 3.2% beverage to be "nonintoxicating." Okla.Stat., Tit. 37, § 163.1 (1958); see State ex rel. Springer v. Bliss, 199 Okl. 198, 185 P.2d 220 (1947). Moreover, many of the studies, while graphically documenting the unfortunate increase in driving while under the influence of alcohol, make no effort to relate their findings to age-sex differentials as involved here.15 Indeed, the only survey that explicitly centered its attention upon young drivers and their use of beer albeit apparently not of the diluted 3.2% variety reached results that hardly can be viewed as impressive in justifying either a gender or age classification.16*204 There is no reason to belabor this line of analysis. It is unrealistic to expect either members of the judiciary or state officials to be well versed in the rigors of experimental or statistical technique. But this merely illustrates that proving broad sociological propositions by statistics is a dubious business, and one that inevitably is in tension with the normative philosophy that underlies the Equal Protection Clause.17 Suffice to say that the showing offered by the appellees does not satisfy us that sex represents a legitimate, accurate proxy for the regulation of drinking and driving. In fact, when it is further recognized that Oklahoma's statute prohibits only the selling of 3.2% beer to young males and not their drinking the beverage once acquired (even after purchase by their 18-20-year-old female companions), the relationship between gender and traffic safety becomes far too tenuous to satisfy Reed's requirement that the gender-based difference be substantially related to achievement of the statutory objective.

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

D

Appellees argue, however, that §§ 241 and 245 enforce state policies concerning the sale and distribution of alcohol and by force of the Twenty-first Amendment should therefore be held to withstand the equal protection challenge. The District Court's response to this contention is unclear. The court assumed that the Twenty-first Amendment "strengthened" the State's police powers with respect to alcohol regulation, 399 F.Supp. at 1307, but then said that "the standards of review that (the Equal Protection Clause) mandates are not relaxed." Id., at 1308. Our view is, and we hold, that the Twenty-first Amendment does not save the*205 invidious gender-based discrimination from invalidation as a denial of equal protection of the laws in violation of the Fourteenth Amendment.

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

The Twenty-first Amendment repealed the Eighteenth Amendment in 1933. The wording of § 2 of the Twenty-first Amendment 20 closely follows the Webb-Kenyon and Wil*206son Acts, expressing the framers' clear intention of constitutionalizing the Commerce Clause framework established under those statutes. This Court's decisions since have confirmed that the Amendment primarily created an exception to the normal operation of the Commerce Clause. See, e. g., Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 330, 84 S.Ct. 1293, 1297, 12 L.Ed.2d 350 (1964); Carter v. Virginia, 321 U.S. 131, 139-140, 64 S.Ct. 464, 469-470, 88 L.Ed. 605 (1944) (Frankfurter, J., concurring); Finch & Co. v. McKittrick, 305 U.S. 395, 398, 59 S.Ct. 256, 257, 83 L.Ed. 246 (1939). Even here, however, the Twenty-first Amendment does not pro tanto repeal the Commerce Clause, but merely requires that each provision "be considered in the light of the other, and in the context of the issues and interests at stake in any concrete case." Hostetter v. Idlewild Bon Voyage Liquor Corp., supra, 377 U.S., at 332, 84 S.Ct., at 1298; cf. Department of Revenue v. James Beam Distilling Co., 377 U.S. 341, 84 S.Ct. 1247, 12 L.Ed.2d 362 (1964); Collins v. Yosemite Park & Curry Co., 304 U.S. 518, 58 S.Ct. 1009, 82 L.Ed. 1502 (1938).

Once passing beyond consideration of the Commerce Clause, the relevance of the Twenty-first Amendment to other constitutional provisions becomes increasingly doubtful. As one commentator has remarked: "Neither the text nor the history of the Twenty-first Amendment suggests that it qualifies individual rights protected by the Bill of Rights and the Fourteenth Amendment where the sale or use of liquor is concerned." P. Brest, Processes of Constitutional Decisionmaking, Cases and Materials 258 (1975). Any departures from this historical view have been limited and sporadic. Two States successfully relied upon the Twenty-first Amendment to respond to challenges of major liquor importers to state authority to regulate the importation and manufacture of alcoholic beverages on Commerce Clause and Fourteenth Amendment grounds. See Mahoney v. Joseph Triner Corp., 304 U.S. 401, 58 S.Ct. 952, 82 L.Ed. 1424 (1938); State Board of Equalization of California v. Young's Market Co.,*207 299 U.S. 59, 64, 57 S.Ct. 77, 79, 81 L.Ed. 38 (1936). In fact, however, the arguments in both cases centered upon importation of intoxicants, a regulatory area where the State's authority under the Twenty-first Amendment is transparently clear, Hostetter v. Idlewild Bon Voyage Liquor Corp., supra, 377 U.S., at 330, and n.9, 84 S.Ct., at 1297, and touched upon purely economic matters that traditionally merit only the mildest review under the Fourteenth Amendment, see, e. g., Joseph E. Seagram & Sons v. Hostetter, 384 U.S. 35, 47-48, 50-51, 86 S.Ct. 1254, 1262-1263, 1263-1264, 16 L.Ed.2d 336 (1966) (rejecting Fourteenth Amendment objections to state liquor laws on the strength of Ferguson v. Skrupa, 372 U.S. 726, 729-730, 83 S.Ct. 1028, 1030-1031, 10 L.Ed.2d 93 (1963) and Williamson v. Lee Optical Co., 348 U.S. 483, 75 S.Ct. 461, 99 L.Ed. 563 (1955) ).21 Cases involving individual rights protected by the Due Process Clause have been treated in sharp contrast. For example, when an individual objected to the mandatory "posting" of her name in retail liquor establishments and her characterization as an "excessive drink(er)," the Twenty-first Amendment was held not to qualify the scope of her due process rights. Wisconsin v. Constantineau, 400 U.S. 433, 436, 91 S.Ct. 507, 509, 27 L.Ed.2d 515 (1971).

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

*208Rather, Moose Lodge No. 107 v. Irvis, 407 U.S. 163, 178-179, 92 S.Ct. 1965, 1974, 32 L.Ed.2d 627 (1972), establishes that state liquor regulatory schemes cannot work invidious discriminations that violate the Equal Protection Clause.

Following this approach, both federal and state courts uniformly have declared the unconstitutionality of gender lines that restrain the activities of customers of state-regulated liquor establishments irrespective of the operation of the Twenty-first Amendment. See, e. g., White v. Fleming, 522 F.2d 730 (CA7 1975); Women's Liberation Union of R.I. v. Israel, 512 F.2d 106 (CA1 1975); Daugherty v. Daley, 370 F.Supp. 338 (ND Ill.1974) (three-judge court); Seidenberg v. McSorleys' Old Ale House, Inc., 317 F.Supp. 593 (SDNY 1970); Commonwealth Alcoholic Beverage Control Bd. v. Burke, 481 S.W.2d 52 (Ky.1972); cf. Sail'er Inn, Inc. v. Kirby, 5 Cal.3d 1, 95 Cal.Rptr. 329, 485 P.2d 529 (1971); Paterson Tavern & G. O. A. v. Hawthorne, 57 N.J. 180, 270 A.2d 628 (1970). Even when state officials have posited sociological or empirical justifications for these gender-based differentiations, the courts have struck down discriminations aimed at an entire class under the guise of alcohol regulation. In fact, social science studies that have uncovered quantifiable differences in drinking tendencies dividing along both racial and ethnic lines strongly suggest the need for application of the Equal Protection Clause in preventing discriminatory treatment that almost certainly would be perceived as invidious.22 In sum, the principles embodied in the Equal*209 Protection Clause are not to be rendered inapplicable by statistically measured but loose-fitting generalities concerning the drinking tendencies of aggregate groups. We thus hold that the operation of the Twenty-first Amendment does not alter the application of equal protection standards that otherwise govern this case.

*210We conclude that the gender-based differential contained in Okla.Stat., Tit. 37, § 245 (1976 Supp.) constitutes a denial of the equal protection of the laws to males aged 18-20 23 and reverse the judgment of the District Court. 24

It is so ordered.

1

1. Sections 241 and 245 provide in pertinent part:

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

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

2

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

3

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

4

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

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

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

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

23

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

24

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

Mr. Justice POWELL, concurring.

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

With respect to the equal protection standard, I agree that Reed v. Reed, 404 U.S. 71, 92 S.Ct. 251, 30 L.Ed.2d 225 (1971), is the most relevant precedent. But I find it unnecessary, in deciding this case, to read that decision as broadly as some of the Court's language may imply. Reed and subsequent cases involving gender-based classifications make clear that the Court subjects such classifications to a more critical examination than is normally applied when "fundamental" constitutional rights and "suspect classes" are not present. **211 I view this as a relatively easy case. No one questions the legitimacy or importance of the asserted governmental objective: the promotion of highway safety. The decision of the case turns on whether the state legislature, by the classification it has chosen, had adopted a means that bears a " 'fair and substantial relation' " to this objective. Id., at 76, 92 S.Ct., at 254, quoting Royster Guano Co. v. Virginia, 253 U.S. 412, 415, 40 S.Ct. 560, 561, 64 L.Ed. 989 (1920).

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

*

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

Mr. Justice STEVENS, concurring.

There is only one Equal Protection Clause. It requires every State to govern impartially. It does not direct the*212 courts to apply one standard of review in some cases and a different standard in other cases. Whatever criticism may be leveled at a judicial opinion implying that there are at least three such standards applies with the same force to a double standard.

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

In this case, the classification is not as obnoxious as some the Court has condemned,1 nor as inoffensive as some the Court has accepted. It is objectionable because it is based on an accident of birth,2 because it is a mere remnant of the now almost universally rejected tradition of discriminating against males in this age bracket,3 and because, to the extent it reflects any physical difference between males and*213 females, it is actually perverse. 4 The question then is whether the traffic safety justification put forward by the State is sufficient to make an otherwise offensive classification acceptable.

The classification is not totally irrational. For the evidence does indicate that there are more males than females in this age bracket who drive and also more who drink. Nevertheless, there are several reasons why I regard the justification as unacceptable. It is difficult to believe that the statute was actually intended to cope with the problem of traffic safety,5 since it has only a minimal effect on access to a not very intoxicating beverage and does not prohibit its consumption.6 Moreover, the empirical data submitted by*214 the State accentuate the unfairness of treating all 18-21-year-old males as inferior to their female counterparts. The legislation imposes a restraint on 100% of the males in the class allegedly because about 2% of them have probably violated one or more laws relating to the consumption of alcoholic beverages. 7 It is unlikely that this law will have a significant deterrent effect either on that 2% or on the law-abiding 98%. But even assuming some such slight benefit, it does not seem to me that an insult to all of the young men of the State can be justified by visiting the sins of the 2% on the 98%.

1

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

2

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

3

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

4

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

5

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

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

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

6

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

7

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

Mr. Justice BLACKMUN, concurring in part.

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

Mr. Justice STEWART, concurring in the judgment.

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

*215Every State has broad power under the Twenty-first Amendment to control the dispensation of alcoholic beverages within its borders. E. g., California v. LaRue, 409 U.S. 109, 93 S.Ct. 390, 34 L.Ed.2d 342; Joseph E. Seagram & Sons v. Hostetter, 384 U.S. 35, 86 S.Ct. 1254, 16 L.Ed.2d 336; Hostetter v. Idlewild Bon Voyage Liquor Corp., 377 U.S. 324, 330, 84 S.Ct. 1293, 1296, 12 L.Ed.2d 350; Mahoney v. Joseph Triner Corp., 304 U.S. 401, 58 S.Ct. 952, 82 L.Ed. 1424; State Board of Equalization v. Young's Market Co., 299 U.S. 59, 57 S.Ct. 77, 81 L.Ed. 38. But "(t)his is not to say that the Twenty-first Amendment empowers a State to act with total irrationality or invidious discrimination in controlling the distribution and dispensation of liquor . . . ." California v. LaRue, supra, 409 U.S., at 120, 93 S.Ct., at 398 n. (concurring opinion).

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

Mr. Chief Justice BURGER, dissenting.

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

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

*216This is not Sullivan v. Little Hunting Park, 396 U.S. 229, 90 S.Ct. 400, 24 L.Ed.2d 386 (1969), or Barrows v. Jackson, 346 U.S. 249, 73 S.Ct. 1031, 97 L.Ed. 1586 (1953), for these is here no barrier whatever to Oklahoma males 18-20 years of age asserting, in an appropriate forum, any constitutional rights they may claim to purchase 3.2% beer. Craig's successful litigation of this very issue was prevented only by the advent of his 21st birthday. There is thus no danger of interminable dilution of those rights if appellant Whitener is not permitted to litigate them here. Cf. Eisenstadt v. Baird, 405 U.S. 438, 445-446, 92 S.Ct. 1029, 1034-1035, 31 L.Ed.2d 349 (1972).

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

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

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

On the merits, we have only recently recognized that our duty is not "to create substantive constitutional rights in the name of guaranteeing equal protection of the laws." San Antonio School Dist. v. Rodriguez, 411 U.S. 1, 33, 93 S.Ct. 1278, 1297, 36 L.Ed.2d 16 (1973). Thus, even interests of such importance in our society as public education and housing do not qualify as "fundamental rights" for equal protection purposes because they have no*217 textually independent constitutional status. See id., at 29-39, 93 S.Ct., at 1294-1300 (education); Lindsey v. Normet, 405 U.S. 56, 92 S.Ct. 862, 31 L.Ed.2d 36 (1972) (housing). Though today's decision does not go so far as to make gender-based classifications "suspect," it makes gender a disfavored classification. Without an independent constitutional basis supporting the right asserted or disfavoring the classification adopted, I can justify no substantive constitutional protection other than the normal McGowan v. Maryland, supra, at 425-426, 81 S.Ct., at 1104-1105, protection afforded by the Equal Protection Clause.

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

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

Mr. Justice REHNQUIST, dissenting.

The Court's disposition of this case is objectionable on two grounds. First is its conclusion that men challenging a gender-based statute which treats them less favorably than women may invoke a more stringent standard of judicial review than pertains to most other types of classifications. Second is the Court's enunciation of this standard, without citation to any source, as being that "classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives." Ante at 197 (emphasis added). The only redeeming feature of the Court's opinion, to my mind, is that it apparently signals a retreat by those who joined the plurality opinion in Frontiero v. Richardson, 411 U.S. 677, 93 S.Ct. 1764, 36 L.Ed.2d 583 (1973), from their view that sex is a "suspect" classification for purposes of equal protection analysis. I think the Oklahoma statute challenged here need pass only the "rational basis" equal*218 protection analysis expounded in cases such as McGowan v. Maryland, 366 U.S. 420, 81 S.Ct. 1101, 6 L.Ed.2d 393 (1961), and Williamson v. Lee Optical Co., 348 U.S. 483, 75 S.Ct. 461, 99 L.Ed. 563 (1955), and I believe that it is constitutional under that analysis.

I

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

Subsequent to Frontiero, the Court has declined to hold that sex is a suspect class, Stanton v. Stanton, supra, at 13, 95 S.Ct., at 1377, and no such holding is imported by the Court's resolution of this case. However, the Court's application here of an elevated or "intermediate" level scrutiny, like that invoked in cases dealing with discrimination against females, raises the question of why the statute here should be treated any differently from countless legislative classifications unrelated to sex which have been upheld under a minimum rationality standard. Jefferson v. Hackney, 406 U.S. 535, 546-547, 92 S.Ct. 1724, 1731-1732, 32 L.Ed.2d 285 (1972); Richardson v. Belcher, 404 U.S. 78, 81-84, 92 S.Ct. 254, 257-259, 30 L.Ed.2d 231 (1971); Dandridge v. Williams, 397 U.S. 471, 484-485, 90 S.Ct. 1153, 1161-1162, 25 L.Ed.2d 491 (1970);*219 McGowan v. Maryland, supra, 366 U.S., at 425-426, 81 S.Ct., at 1104-1105; Flemming v. Nestor, 363 U.S. 603, 611, 80 S.Ct. 1367, 1373, 4 L.Ed.2d 1435 (1960); Williamson v. Lee Optical Co., supra, 348 U.S., at 488-489, 75 S.Ct., at 464-465.

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

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

It is true that a number of our opinions contain broadly phrased dicta implying that the same test should be applied to all classifications based on sex, whether affecting females or males. E. g., Frontiero v. Richardson, supra, 411 U.S., at 688, 93 S.Ct., at 1771; Reed v. Reed, 404 U.S. 71, 76, 92 S.Ct. 251, 254, 30 L.Ed.2d 225 (1971). However, before today, no decision of this Court has applied an elevated level of scrutiny to invalidate a statutory discrimination harmful to males, except where the statute impaired an important personal interest protected by the Constitution.1 There being no such interest*220 here, and there being no plausible argument that this is a discrimination against females,2 the Court's reliance on our previous sex-discrimination cases is ill-founded. It treats gender classification as a talisman which without regard to the rights involved or the persons affected calls into effect a heavier burden of judicial review.

The Court's conclusion that a law which treats males less favorably than females "must serve important governmental objectives and must be substantially related to achievement of those objectives" apparently comes out of thin air. The Equal Protection Clause contains no such language, and none of our previous cases adopt that standard. I would think we have had enough difficulty with the two standards of review which our cases have recognized the*221 norm of "rational basis," and the "compelling state interest" required where a "suspect classification" is involved so as to counsel weightily against the insertion of still another "standard" between those two. How is this Court to divine what objectives are important? How is it to determine whether a particular law is "substantially" related to the achievement of such objective, rather than related in some other way to its achievement? Both of the phrases used are so diaphanous and elastic as to invite subjective judicial preferences or prejudices relating to particular types of legislation, masquerading as judgments whether such legislation is directed at "important" objectives or, whether the relationship to those objectives is "substantial" enough.

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

II

The applicable rational-basis test is one which

"permits the States a wide scope of discretion in enacting laws which affect some groups of citizens differently than*222 others. The constitutional safeguard is offended only if the classification rests on grounds wholly irrelevant to the achievement of the State's objective. State legislatures are presumed to have acted within their constitutional power despite the fact that, in practice, their laws result in some inequality. A statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it." McGowan v. Maryland, supra, 366 U.S., at 425-426, 81 S.Ct., at 1105 (citations omitted).

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

The Court "accept(s) for purposes of discussion" the District Court's finding that the purpose of the provisions in question was traffic safety, and proceeds to examine the statistical evidence in the record in order to decide if "the gender-based distinction closely serves to achieve that objective." Ante, at 199, 200 (emphasis added). (Whether there is a difference between laws which "closely serv(e)" objectives and those which are only "substantially related" to their achievement, ante, at 197, we are not told.) I believe that a more traditional type of scrutiny is appropriate in this case, and I think that the Court would have done well here to heed its own warning that "(i)t is unrealistic to expect . . . members of the judiciary . . . to be well versed in the rigors of experimental or statistical technique." Ante, at 204. One*223 need not immerse oneself in the fine points of statistical analysis, however, in order to see the weaknesses in the Court's attempted denigration of the evidence at hand.

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

Other surveys indicated (1) that over the five-year period from 1967 to 1972, nationwide arrests among those under 18 for drunken driving increased 138%, and that 93% of all persons arrested for drunken driving were male; 4 (2) that youths in the 17-21 age group were overrepresented among those killed or injured in Oklahoma traffic accidents, that male casualties substantially exceeded female, and that deaths in this age group continued to rise while overall traffic deaths declined; 5 (3) that over three-fourths of the drivers under 20 in the Oklahoma City area are males, and that each of them, on average, drives half again as many miles per year as their female counterparts; 6 (4) that four-fifths of male drivers*224 under 20 in the Oklahoma City area state a drink preference for beer, while about three-fifths of female drivers of that age state the same preference; 7 and (5) that the percentage of male drivers under 20 admitting to drinking within two hours of driving was half again larger than the percentage for females, and that the percentage of male drivers of that age group with a blood alcohol content greater than .01% was almost half again larger than for female drivers.8

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

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

The Court also complains of insufficient integration of the various surveys on several counts that the injury and death figures are in no way directly correlated with intoxication, ante, at 201 n. 9; that the national arrest figures for drunk driving contain no breakdown for the 18 to 21-year-old group,*225 ante, at 459, n. 11; and that the arrest records for intoxication are not tied to the consumption of 3.2% beer, ante, at 201-202, nn. 11 and 12. But the State of Oklahoma and certainly this Court for purposes of equal protection review can surely take notice of the fact that drunkenness is a significant cause of traffic casualties, and that youthful offenders have participated in the increase of the drunk-driving problem. On this latter point, the survey data indicating increased driving casualties among 18-21-year-olds, while overall casualties dropped, are not irrelevant.

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

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

The rationality of a statutory classification for equal protection purposes does not depend upon the statistical "fit" between the class and the trait sought to be singled out. It turns on whether there may be a sufficiently higher in*226cidence of the trait within the included class than in the excluded class to justify different treatment. Therefore the present equal protection challenge to this gender-based discrimination poses only the question whether the incidence of drunk driving among young men is sufficiently greater than among young women to justify differential treatment. Notwithstanding the Court's critique of the statistical evidence, that evidence suggests clear differences between the drinking and driving habits of young men and women. Those differences are grounds enough for the State reasonably to conclude that young males pose by far the greater drunk-driving hazard, both in terms of sheer numbers and in terms of hazard on a per-driver basis. The gender-based difference in treatment in this case is therefore not irrational.

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

Cast in those terms, the argument carries little weight, in light of our decisions indicating that such questions call for a balance of the State's interest against the harm resulting from any overinclusiveness or underinclusiveness. Vlandis v. Kline, 412 U.S. 441, 448-452, 93 S.Ct. 2230, 2234-2236, 37 L.Ed.2d 63 (1973). The personal interest harmed*227 here is very minor the present legislation implicates only the right to purchase 3.2% beer, certainly a far cry from the important personal interests which have on occasion supported this Court's invalidation of statutes on similar reasoning. Cleveland Board of Education v. LaFleur, 414 U.S. 632, 640, 94 S.Ct. 791, 796, 39 L.Ed.2d 52 (1974); Stanley v. Illinois, 405 U.S. 645, 651, 92 S.Ct. 1208, 1212, 31 L.Ed.2d 551 (1972). And the state interest involved is significant the prevention of injury and death on the highways.

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

The Oklahoma Legislature could have believed that 18-20-year-old males drive substantially more, and tend more often to be intoxicated than their female counterparts; that they prefer beer and admit to drinking and driving at a higher rate than females; and that they suffer traffic injuries out of proportion to the part they make up of the population. Under the appropriate rational-basis test for equal protection, it is neither irrational nor arbitrary to bar them from making purchases of 3.2% beer, which purchases might in many cases be made by a young man who immediately returns to his vehicle with the beverage in his possession. The record does not give any good indication of the true proportion of males in the age group who drink and drive (ex*228cept that it is no doubt greater than the 2% who are arrested), but whatever it may be I cannot see that the mere purchase right involved could conceivably raise a due process question. There being no violation of either equal protection or due process, the statute should accordingly be upheld.

1

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

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

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

2

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

3

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

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

5

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

6

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

7

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

8

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

MICHAEL M. v. SUPERIOR COURT OF SONOMA COUNTY (CALIFORNIA, REAL PARTY IN INTEREST)

No. 79-1344.

Argued November 4, 1980

Decided March 23, 1981

Rehnquist, J., announced the judgment of the Court and delivered an opinion, in which Burger, C. J., and Stewart and Powell, JJ., joined. Stewart, J., filed a concurring opinion, post, p. 476. Blackmun, J., filed an opinion concurring in the judgment, post, p. 481. Brennan, J., filed a dissenting opinion, in which White and Marshall, JJ., joined, post, p. 488. Stevens, J., filed a dissenting opinion; post, p. 496.

Gregory F. Jilka argued the cause and filed a brief for petitioner.

Sandy B. Kriegler, Deputy Attorney General of California, argued the cause for respondent. With him on the brief were George Deukmejian, Attorney General, Robert H. Phili-bosian, Chief Assistant Attorney General, S. Clark Moore, Assistant Attorney General, and William R. Pounders, Deputy Attorney General.*

*

Briefs of amici curiae urging reversal were filed by Bruce J. Ennis, Jr., for the American Civil Liberties Union et al; and by John W. Karr for the Women’s Legal Defense Fund.

Solicitor General McCree, Assistant Attorney General Heymann, and Sara Criscitelli filed a brief for the United States as amicus curiae urging affirmance.

Justice Rehnquist

announced the judgment of the Court and delivered an opinion, in which The Chief Justice, Justice Stewart, and Justice Powell joined.

The question presented in this case is whether California’s “statutory rape” law, § 261.5 of the Cal. Penal Code Ann. (West Supp. 1981), violates the Equal Protection Clause of the Fourteenth Amendment. Section 261.5 defines unlawful sexual intercourse as “an act of sexual intercourse accomplished with a female not the wife of the perpetrator, where the female is under the age of 18 years.” The statute thus makes men alone criminally liable for the act of sexual intercourse.

In July 1978, a complaint was filed in the Municipal Court of Sonoma County, Cal., alleging that petitioner, then a 17%-year-old male, had had unlawful sexual intercourse with a female under the age of 18, in violation of § 261.5. The evidence adduced at a preliminary hearing showed that at approximately midnight on June 3, 1978, petitioner and two friends approached Sharon, a 16%-year-old female, and her sister as they waited at a bus stop. Petitioner and Sharon, who had already been drinking, moved away from the others and began to kiss. After being struck in the face for rebuffing petitioner’s initial advances, Sharon submitted to sexual intercourse with petitioner. Prior to trial, petitioner sought to set aside the information on both state and federal constitutional grounds, asserting that § 261.5 unlawfully discriminated on the basis of gender. The trial court and the California Court of Appeal denied petitioner’s request for relief and petitioner sought review in the Supreme Court of California.

The Supreme Court held that “section 261.5 discriminates on the basis of sex because only females may be victims, and only males may violate the section.” 25 Cal. 3d 608, 611, 601 P. 2d 572, 574. The court then subjected the classification to “strict scrutiny,” stating that it must be justified by a compelling state interest. It found that the classification was “supported not by mere social convention but by the immutable physiological fact that it is the female exclusively who can become pregnant.” Ibid. Canvassing “the tragic human costs of illegitimate teenage pregnancies,” including the large number of teenage abortions, the increased medical risk associated with teenage .pregnancies, and the social consequences of teenage childbearing, the court concluded that the State has a compelling interest in preventing such pregnancies. Because males alone can “physiologically cause the result which the law properly seeks to avoid,” the court further held that the gender classification was readily justified as a means of identifying offender and victim. For the reasons stated below, we affirm the judgment of the California Supreme Court.1

As is evident from our opinions, the Court has had some difficulty in agreeing upon the proper approach and analysis in cases involving challenges to gender-based classifications. The issues posed by such challenges range from issues of standing, see Orr v. Orr, 440 U. S. 268 (1979), to the appropriate standard of judicial review for the substantive classification. Unlike the California Supreme Court, we have not held that gender-based classifications are “inherently suspect” and thus we do not apply so-called “strict scrutiny” to those classifications. See Stanton v. Stanton, 421 U. S. 7 (1975). Our cases have held, however, that the traditional minimum rationality test takes on a somewhat “sharper focus” when gender-based classifications are challenged. See Craig v. Boren, 429 U. S. 190, 210 n." (1976) (Powell, J., concurring). In Reed v. Reed, 404 U. S. 71 (1971), for example, the Court stated that a gender-based classification will be upheld if it bears a “fair and substantial relationship” to legitimate state ends, while in Craig v. Boren, supra, at 197, the Court restated the test to require the classification to bear a “substantial relationship” to “important governmental objectives.”

Underlying these decisions is the principle that a legislature may not “make overbroad generalizations based on sex which are entirely unrelated to any differences between men and women or which demean the ability or social status of the affected class.” Parham v. Hughes, 441 U. S. 347, 354 (1979) (plurality opinion of Stewart, J.). But because the Equal Protection Clause does not “demand that a statute necessarily apply equally to all persons” or require “ 'things which are different in fact... to be treated in law as though they were the same/ ” Rinaldi v. Yeager, 384 U. S. 305, 309 (1966), quoting Tigner v. Texas, 310 U. S. 141, 147 (1940), this Court has consistently upheld statutes where the gender classification is not invidious, but rather realistically reflects the fact that the sexes are not similarly situated in certain circumstances. Parham v. Hughes, supra; Califano v. Webster, 430 U. S. 313 (1977); Schlesinger v. Ballard, 419 U. S. 498 (1975); Kahn v. Shevin, 416 U. S. 351 (1974). As the Court has stated, a legislature may “provide for the special problems of women.” Weinberger v. Wiesenfeld, 420 U. S. 636, 653 (1975).

Applying those principles to this case, the fact that the California Legislature criminalized the act of illicit sexual intercourse with a minor female is a sure indication of its intent or purpose to discourage that conduct.2 Precisely why the legislature desired that result is of course somewhat less clear. This Court has long recognized that “[¿Inquiries into congressional motives or purposes are a hazardous matter,” United States v. O’Brien, 391 U. S. 367, 383-384 (1968); Palmer v. Thompson, 403 U. S. 217, 224 (1971), and the search for the “actual” or “primary” purpose of a statute is likely to be elusive. Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 265 (1977); McGinnis v. Royster, 410 U. S. 263, 276-277 (1973). Here, for example, the individual legislators may have voted for the statute for a variety of reasons. Some legislators may have been concerned about preventing teenage pregnancies, others about protecting young females from physical injury or from the loss of “chastity,” and still others about promoting various religious and moral attitudes towards premarital sex.

The justification for the statute offered by the State, and accepted by the Supreme Court of California, is that the legislature sought to prevent illegitimate teenage pregnancies. That finding, of course, is entitled to great deference. Reitman v. Mulkey, 387 U. S. 369, 373-374 (1967). And although our cases establish that the State’s asserted reason for the enactment of a statute may be rejected, if it “could not have been a goal of the legislation,” Weinberger v. Wiesenfeld, supra, at 648, n. 16, this is not such a case.

We are satisfied not only that the prevention of illegitimate pregnancy is at least one of the “purposes” of the statute, but also that the State has a strong interest in preventing such pregnancy. At the risk of stating the obvious, teenage pregnancies, which have increased dramatically over the last two decades,3 have significant social, medical, and economic consequences for both the mother and her child, and the State.4 Of particular concern to the State is that approximately half of all teenage pregnancies end in abortion.5 And of those children who are born, their illegitimacy makes them likely candidates to become wards of the State.6

We need not be medical doctors to discern that young men and young women are not similarly situated with respect to the problems and the risks of sexual intercourse. Only women may become pregnant, and they suffer disproportionately the profound physical, emotional, and psychological consequences of sexual activity. The statute at issue here protects women from sexual intercourse at an age when those consequences are particularly severe.7

The question thus boils down to whether a State may attack the problem of sexual intercourse and teenage pregnancy directly by prohibiting a male from having sexual intercourse with a minor female.8 We hold that such a statute is sufficiently related to the State’s objectives to pass constitutional muster.

Because virtually all of the significant harmful and inescapably identifiable consequences of teenage pregnancy fall on the young female, a legislature acts well within its authority when it elects to punish only the participant who, by nature, suffers few of the consequences of his conduct. It is hardly unreasonable for a legislature acting to protect minor females to exclude them from punishment. Moreover, the risk of pregnancy itself constitutes a substantial deterrence to young females. No similar natural sanctions deter males. A criminal sanction imposed solely on males thus serves to roughly “equalize” the deterrents on the sexes.

We are unable to accept petitioner’s contention that the statute is impermissibly underinclusive and must, in order to pass judicial scrutiny, be broadened so as to hold the female as criminally liable as the male. It is argued that this statute is not necessary to deter teenage pregnancy because a gender-neutral statute, where both male and female would be subject to prosecution, would serve that goal equally well. The relevant inquiry, however, is not whether the statute is drawn as precisely as it might have been, but whether the line chosen by the California Legislature is within constitutional limitations. Kahn v. Shevin, 416 U. S., at 356, n. 10.

In any event, we cannot say that a gender-neutral statute would be as effective as the statute California has chosen to enact. The State persuasively contends that a gender-neutral statute would frustrate its interest in effective enforcement. Its view is that a female is surely less likely to report violations of the statute if she herself would be subject to criminal prosecution.9 In an area already fraught with prose-cutorial difficulties, we decline to hold that the Equal Protection Clause requires a legislature to enact a statute so broad that it may well be incapable of enforcement.10

We similarly reject petitioner’s argument that § 261.5 is impermissibly overbroad because it makes unlawful sexual intercourse with prepubescent females, who are, by definition, incapable of becoming pregnant. Quite apart from the fact that the statute could well be justified on the grounds that very young females are particularly susceptible to physical injury from sexual intercourse, see Rundlett v. Oliver, 607 F. 2d 495 (CA1 1979), it is ludicrous to suggest that the Constitution requires the California Legislature to limit the scope of its rape statute to older teenagers and exclude young girls.

There remains only petitioner’s contention that the statute is unconstitutional as it is applied to him because he, like Sharon, was under 18 at the time of sexual intercourse. Petitioner argues that the statute is flawed because it presumes that as between two persons under 18, the male is the culpable aggressor We find petitioner’s contentions unpersuasive. Contrary to his assertions, the statute does not rest on the assumption that males are generally the aggressors. It is instead an attempt by a legislature to prevent illegitimate teenage pregnancy by providing an additional deterrent for men. The age of the man is irrelevant since young men are as capable as older men of inflicting the harm sought to be. prevented.

In upholding the California statute we also recognize that this is not a case where a statute is being challenged on the grounds that it “invidiously discriminates” against females. To the contrary, the statute places a burden on males which is not shared by females. But we find nothing to suggest that men, because of past discrimination or peculiar disadvantages, are in need of the special solicitude of the courts. Nor is this a case where the gender classification is made “solely for . . . administrative convenience,” as in Frontiero v. Richardson, 411 U. S. 677, 690 (1973) (emphasis omitted), or rests on “the baggage of sexual stereotypes” as in Orr v. Orr, 440 U. S., at 283. As we have held, the statute instead reasonably reflects the fact that the consequences of sexual intercourse and pregnancy fall more heavily on the female than on the male.

Accordingly the judgment of the California Supreme Court is

Affirmed.

1

The lower federal courts and state courts have almost uniformly concluded that statutory rape laws are constitutional. See, e. g., Rundlett v. Oliver, 607 F. 2d 495 (CA1 1979); Hall v. McKenzie, 537 F. 2d 1232 (CA4 1976); Hall v. State, 365 So. 2d 1249, 1252-1253 (Ala. App. 1978), cert. denied, 365 So. 2d 1253 (Ala. 1979); State v. Gray, 122 Ariz. 445, 446-477, 595 P. 2d 990, 991-992 (1979); People v. Mackey, 46 Cal. App. 3d 755, 760-761, 120 Cal. Rptr. 157, 160, cert. denied, 423 U. S. 951 (1975); People v. Salinas, 191 Colo. 171, 551 P. 2d 703 (1976); State v. Brothers, 384 A. 2d 402 (Del. Super. 1978); In re W. E. P., 318 A. 2d 286, 289-290 (DC 1974); Barnes v. State, 244 Ga. 302, 303-304, 260 S. E. 2d 40, 41-42 (1979); State v. Drake, 219 N. W. 2d 492, 495-496 (Iowa 1974); State v. Bell, 377 So. 2d 303 (La. 1979); State v. Rundlett, 391 A. 2d 815 (Me. 1978); Green v. State, 270 So. 2d 695 (Miss. 1972); In re J. D. G., 498 S. W. 2d 786, 792-793 (Mo. 1973); State v. Meloon, 116 N. H. 669, 366 A. 2d 1176 (1976); State v. Thompson, 162 N. J. Super. 302, 392 A. 2d 678 (1978); People v. Whidden, 51 N. Y. 2d 457, 415 N. E. 2d 927 (1980); State v. Wilson, 296 N. C. 298, 311-313, 250 S. E. 2d 621, 629-630 (1979); Olson v. State, 588 P. 2d 1018 (Nev. 1979); State v. Elmore, 24 Ore. App. 651, 546 P. 2d 1117 (1976); State v. Ware, - R. I. -, 418 A. 2d 1 (1980); Roe v. State, 584 S. W. 2d 257, 259 (Tenn. Crim. App. 1979); Ex parte Groves, 571 S. W. 2d 888, 892-893 (Tex. Crim. App. 1978); Moore v. McKenzie, 236 S. E. 2d 342, 342-343 (W. Ya. 1977) ; Flores v. State, 69 Wis. 2d 509, 510-511, 230 N. W. 2d 637, 638 (1975). Contra, Navedo v. Preisser, 630 F. 2d 636 (CA8 1980); United States v. Hicks, 625 F. 2d 216 (CA9 1980); Meloon v. Helgemoe, 564 F. 2d 602 (CA1 1977) (limited in Rundlett v. Oliver, supra), cert. denied, 436 U. S. 950 (1978).

2

The statute was enacted as part of California’s' first penal code in 1850, 1850 Cal. Stats., ch. 99, § 47, p. 234, and recodified and amended in 1970.

3

In 1976 approximately one million 15-to-19-year-olds became pregnant, one-tenth of all women in that age group. Two-thirds of the pregnancies were illegitimate. Illegitimacy rates for teenagers (births per 1,000 unmarried females ages 14 to 19) increased 75% for 14-to-17-year-olds between 1961 and 1974 and 33% for 18-to-19-year-olds. Alan Guttmacher Institute, 11 Million Teenagers 10, 13 (1976); C. Chilman, Adolescent Sexuality In a Changing American Society 195 (NIH Pub. No. 80-1426, 1980).

4

The risk of maternal death is 60% higher for a teenager under the age of 15 than for a women in her early twenties. The risk is 13% higher for 15-to-19-year-olds. The statistics further show that most teenage mothers drop out of school and face a bleak economic future. See, e. g., 11 Million Teenagers, supra, at 23, 25; Bennett & Bardon, The Effects of a School Program On Teenager Mothers and Their Children, 47 Am. J. Orthopsychiatry 671 (1977); Phipps-Yonas, Teenage Pregnancy and Motherhood, 50 Am. J. Orthopsychiatry 403, 414 (1980).

5

This is because teenagers are disproportionately likely to seek abortions. Center for Disease Control, Abortion Surveillance 1976, pp. 22-24 (1978). In 1978, for example, teenagers in California had approximately 54,000 abortions and 53,800 live births. California Center for Health Statistics, Reproductive Health Status of California Teenage Women 1, 23 (Mar. 1980).

6

The policy and intent of the California Legislature evinced in other legislation buttresses our view that the prevention of teenage pregnancy is a purpose of the statute. The preamble to the Pregnancy Freedom of Choice Act, for example, states: “The legislature finds that pregnancy among unmarried persons under 21 years of age constitutes an increasing social problem in the State of California.” Cal. Welf. & Inst. Code Ann. §16145 (West 1980).

Subsequent to the decision below, the California Legislature considered and rejected proposals to render § 261.5 gender neutral, thereby ratifying the judgment of the California Supreme Court-. That is enough to answer petitioner’s contention that the statute was the “ 'accidental byproduct of a traditional way of thinking about females.’” Califano v. Webster, 430 U. S. 313, 320 (1977) (quoting Califano v. Goldfarb, 430 U. S. 199, 223 (1977) (Stevens, J., concurring in judgment)). Certainly this decision of the California Legislature is as good a source as is this Court in deciding what is “current” and what is “outmoded” in the perception of women.

7

Although petitioner concedes that the State has a “compelling” interest in preventing teenage pregnancy, he contends that the “true” purpose of § 261.5 is to protect the virtue and chastity of young .women. As such, the statute is unjustifiable because it rests on archaic stereotypes. What we have said above is enough to dispose of that- contention. The question for us — and the only question under the Federal/Constitution — is whether the legislation violates the Equal Protection Clause of the Fourteenth Amendment, not whether its supporters may have^endorsed it for reasons no longer generally accepted. Even if the preservation of female chastity were one of the motives of the statute, and even if that motive be impermissible, petitioner’s argument must fail because “[i]t is a familiar practice of constitutional law that this court will not strike down an otherwise constitutional statute on the basis of an alleged illicit legislative motive.” United States v. O’Brien, 391 U. S. 367, 383 (1968). In Orr v. Orr, 440 U. S. 268 (1979), for example, the Court rejected one asserted purpose as impermissible, but then considered other purposes to determine if they could justify the statute. Similarly, in Washington v. Davis, 426 U. S. 229, 243 (1976), the Court distinguished Palmer v. Thompson, 403 U. S. 217 (1971), on the grounds that the purposes of the ordinance there were not open to impeachment by evidence that the legislature was actually motivated by an impermissible purpose. See also Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 270, n. 21 (1977); Mobile v. Bolden, 446 U. S. 55, 91 (1980) (Stevens, J., concurring in judgment).

8

We do not understand petitioner to question a State’s authority to make sexual intercourse among teenagers a criminal act, at least on a gender-neutral basis. In Carey v. Population Services International, 431 U. S. 678, 694, n. 17 (1977) (plurality opinion of Brennan, J.), four Members of the Court assumed for the purposes of that case that a State may regulate the sexual behavior of minors, while four other Members of the Court more emphatically stated that such regulation would be permissible. Id., at 702, 703 (White, J., concurring in part and concurring in result); id., at 705-707, 709 (Powell, J., concurring in part and concurring in judgment); id., at 713 (Stevens, J., concurring in part and concurring in judgment); id., at 718 (Rehnquist, J., dissenting). The Court has long recognized that a State has even broader authority to protect the physical, mental, and moral well-being of its youth, than of its adults. See, e. g., Planned Parenthood of Central Mo. v. Danforth, 428 U. S. 52, 72-74 (1976); Ginsberg v. New York, 390 U. S. 629, 639-640 (1968); Prince v. Massachusetts, 321 U. S. 158, 170 (1944).

9

Petitioner contends that a gender-neutral statute would not hinder prosecutions because the prosecutor could take into account the relative burdens on females and males and generally only prosecute males. But to concede this is to concede all. If the prosecutor, in exercising discretion, will virtually always prosecute just the man and not the woman, we do not see why it is impermissible for the legislature to enact a statute to the same effect.

10

The question whether a statute is substantially related to its asserted goals is at best an opaque one. It can be plausibly argued that a gender-neutral statute would produce fewer prosecutions than the statute at issue here. See Stewart, J., concurring, post, at 481, n. 13. Justice Brennan’s dissent argues, on the other hand, that

“even assuming that a gender-neutral statute would be more difficult to enforce, . . . [cjommon sense . . . suggests that a gender-neutral statutory rape law is potentially a greater deterrent of sexual activity than a gender-based law, for the simple reason that a gender-neutral law subjects both men and women to criminal sanctions and thus arguably has a deterrent effect on twice as many potential violators.” Post, at 493-494 (emphasis deleted).

Where such differing speculations as to the effect of a statute are plausible, we think it appropriate to defer to the decision of the California Supreme Court, “armed as it was with the knowledge of the facts and circumstances concerning the passage and potential impact of [the statute], and familiar with the milieu in which that provision would operate.” Reitman v. Mulkey, 387 ü. S. 369, 378-379 (1967).

It should be noted that two of the three cases relied upon by Justice Brennan’s dissent are readily distinguishable from the instant one. See post, at 490, n. 3. In both Navedo v. Preisser, 630 F. 2d 636 (CA8 1980), and Meloon v. Helgemoe, 564 F. 2d 602 (CA1 1977), cert. denied, 436 U. S. 950 (1978), the respective governments asserted that the purpose of the statute was to protect young women from physical injury. Both courts rejected the justification on the grounds that there had been no showing that young females are more likely than males to suffer physical injury from sexual intercourse. They further held, contrary to our decision, that pregnancy prevention was not a “plausible” purpose of the legislation. Thus neither court reached the issue presented here, whether the statute is substantially related to the prevention of teenage pregnancy. Significantly, Meloon has been severely limited by Rundlett v. Oliver, 607 F. 2d 495 (CA1 1979), where the court upheld a statutory rape law on the ground that the State had shown that sexual intercourse physically injures young women more than males. Here, of course, even Justice Brennan’s dissent does not dispute that young women suffer disproportionately the deleterious consequences of illegitimate pregnancy.

Justice Stewart,

concurring.

Section 261.5, on its face, classifies on the basis of sex. A male who engages in sexual intercourse with an underage female who is not his wife violates the statute; a female who engages in sexual intercourse with an underage male who is not her husband does not.1 The petitioner contends that this state law, which punishes only males for the conduct in question, violates his Fourteenth Amendment right to the equal protection of the law. The Court today correctly rejects that contention.

A

At the outset, it should be noted that the statutory discrimination, when viewed as part of the wider scheme of California law, is not as clearcut as might at first appear. Females are not freed from criminal liability in California for engaging in sexual activity that may be harmful. It is unlawful, for example, for any person, of either sex, to molest, annoy, or contribute to the delinquency of anyone under 18 years of age.2 All persons are prohibited from committing “any lewd or lascivious act,” including consensual intercourse, with a child under 14.3 And members of both sexes may be convicted for engaging in deviant sexual acts with anyone under 18.4 Finally, females may be brought within the proscription of § 261.5 itself, since a female may be charged with aiding and abetting its violation.5

Section 261.5 is thus but one part of a broad statutory scheme that protects all minors from the problems and risks attendant upon adolescent sexual activity. To be sure, § 261.5 creates an additional measure of punishment for males who engage in sexual intercourse with females between the ages of 14 and 17.6 The question then is whether the Constitution prohibits a state legislature from imposing this additional sanction on a gender-specific basis.

B

The Constitution is violated when government, state or federal, invidiously classifies similarly situated people on the basis of the immutable characteristics with which they were

born. • Thus, detrimental racial classifications by government always violate the Constitution, for the simple reason that, so far as the Constitution is concerned, people of different races are always similarly situated. See Fullilove v. Klutznick, 448 U. S. 448, 522 (dissenting opinion); McLaughlin v. Florida, 379 U. S. 184, 198 (concurring opinion); Brown v. Board of Ed., 347 U. S. 483; Plessy v. Ferguson, 163 U. S. 537, 552 (dissenting opinion). By contrast, while detrimental gender classifications by government often violate the Constitution, they do not always do so, for the reason that there are differences between males and females that the Constitution necessarily recognizes. In this case we deal with the most basic of these differences: females can become pregnant as the result of sexual intercourse; males cannot.

As was recognized in Parham v. Hughes, 441 U. S. 347, 354, “a State is not free to make overbroad generalizations based on sex which are entirely unrelated to any differences between men and women or which demean the ability or social status of the affected class.” Gender-based classifications may not be based upon administrative convenience, or upon archaic assumptions about the proper roles of the sexes. Craig v. Boren, 429 U. S. 190; Frontiero v. Richardson, 411 U. S. 677; Reed v. Reed, 404 U. S. 71. But we have recognized that in certain narrow circumstances men and women are not similarly situated; in these circumstances a gender classification based on clear differences between the sexes is not invidious, and a legislative classification realistically based upon those differences is not unconstitutional. See Parham v. Hughes, supra; Califano v. Webster, 430 U. S. 313, 316-317; Schlesinger v. Ballard, 419 U. S. 498; cf. San Antonio Independent School Dist. v. Rodriguez, 411 U. S. 1, 59 (concurring opinion). “ [G] ender-based classifications are not invariably invalid. When men and women are not in fact similarly situated in the area covered by the legislation in question, the Equal Protection Clause is not violated.” Caban v. Mohammed, 441 U. S. 380, 398 (dissenting opinion).

Applying these principles to the classification enacted by the California Legislature, it is readily apparent that § 261.5 does not violate the Equal Protection Clause. Young women and men are not similarly situated with respect to the problems and risks associated with intercourse and pregnancy, and the statute is realistically related to the legitimate state purpose of reducing those problems and risks.

C

As the California Supreme Court’s catalog shows, the pregnant unmarried female confronts problems more numerous and more severe than any faced by her male partner.7 She alone endures the medical risks of pregnancy or abortion.8 She suffers disproportionately the social, educational, and emotional consequences of pregnancy.9 Recognizing this disproportion, California has attempted to protect teenage females by prohibiting males from participating in the act necessary for conception.10

The fact that males and females are not-similarly situated with respect to the risks of sexual intercourse applies with the same force to males under 18 as it does to older males. The risk of pregnancy is a significant deterrent for unwed young females that is not shared by unmarried males, regardless of their age. Experienced observation confirms the commonsense notion that adolescent males disregard the possibility of pregnancy far more than do adolescent females.11 And to the extent that § 261.5 may punish males for intercourse with prepubescent females, that punishment is justifiable because of the substantial physical risks for prepubescent females that are not shared by their male counterparts.12

D

The petitioner argues that the California Legislature could have drafted the statute differently, so that its purpose would be accomplished more precisely. “But the issue, of course, is not whether the statute could have been drafted more wisely, but whether the lines chosen by the . . . [legislature are within constitutional limitations.” Kahn v. Shevin, 416 U. S. 351, 356, n. 10. That other States may have decided to attack the sanie problems more broadly, with gender-neutral statutes, does not mean that every State is constitutionally compelled to do so.13

E

In short, the Equal Protection Clause does not mean that the physiological differences between men and women must be disregarded. While those differences must never be permitted to become a pretext for invidious discrimination, no such discrimination is presented by this case. The Constitution surely does not require a State to pretend that demonstrable differences between men and women do not really exist.

1

But see n. 5 and accompanying text, infra.

2

See Cal. Penal Code Ann. §§272, 647a (West Supp. 1981).

3

Cal. Penal Code Ann. §288 (West Supp. 1981). See People v. Dontanville, 10 Cal. App. 3d 783, 796, 89 Cal. Rptr. 172, 180 (2d Dist,).

4

See Cal. Penal Code Ann. §§286 (b)(1), 288a (b)(1) (West Supp. 1981).

5

See Cal. Penal Code Ann. §31 (West 1970); People v. Haywood, 131 Cal. App. 2d 259, 280 P. 2d 180 (2d Dist.); People v. Lewis, 113 Cal. App. 2d 468, 248 P. 2d 461 (1st Dist.). According to statistics maintained by the California Department of Justice Bureau of Criminal Statistics, approximately 14% of the juveniles arrested for participation in acts made unlawful by § 261.5 between 1975 and 1979 were females. Moreover, an underage female who is as culpable as her male partner, or more culpable, may be prosecuted as a juvenile delinquent. Cal. Welf. & Inst. Code Ann. § 602 (West Supp. 1981); In re Gladys R., 1 Cal. 3d 855, 867-869, 464 P. 2d 127, 136-138.

6

Males and females are equally prohibited by § 288 from sexual intercourse with minors under 14. Compare Cal. Penal Code Ann. § 288 (West Supp. 1981) with Cal. Penal Code Ann. §§18, 264 (West Supp. 1981).

7

The court noted that from 1971 through 1976, 83.6% of the 4,860 children born to girls under 15 in California were illegitimate, as were 51% of those bom to girls 15 to 17. The court also observed that while accounting for only 21% of California pregnancies in 1976, teenagers accounted for 34.7% of legal abortions. See ante, at 470, n. 3.

8

There is also empirical evidence that sexual abuse of young females is a more serious problem than sexual abuse of young males. For example, a review of five studies found that 88% of sexually abused minors were female. Jaffe, Dynneson, & ten Bensel, Sexual Abuse of Children 129 Am. J. of Diseases of Children 689, 690 (1975). Another study, involving admissions to a hospital emergency room over a 3-year period, reported that 86 of 100 children examined for sexual abuse were girls. Orr & Prietto, Emergency Management of Sexually Abused Children, 133 Am. J. of Diseased Children 630 (1979). See also State v. Craig, 169 Mont. 150, 156-157, 545 P. 2d 649, 653; Sarafino, An Estimate of Nationwide Incidence of Sexual Offenses Against Children, 58 Child Welfare 127, 131 (1979).

9

Most teenage mothers do not finish high school and are disadvantaged economically thereafter. See Moore, Teenage Childbirth and Welfare Dependency, 10 Family Planning Perspectives 233-235 (1978). The suicide rate for teenage mothers is seven times greater than that for teenage girls without children. F. Nye, School-Age Parenthood (Wash. State U. Ext. Bull. No. 667) 8 (1976). And 60% of adolescent mothers aged 15 to 17 are on welfare within two to five years of the birth of their children. Teenage Pregnancy, Everybody’s Problem 3-4 (DHEW Publication (HSA) No. 77-5619).

10

Despite the increased availability of contraceptives and sex education, the pregnancy rates for young women are increasing. See Alan Gutt-macher Institute, 11 Million Teenagers 12 (1976). See generally C. Chil-man, Adolescent Sexuality in a Changing American Society (NIH Pub. No. 80-1426, 1980).

The petitioner contends that the statute is overinclusive because it does not allow a defense that contraceptives were used, or that procreation was for some other reason impossible. The petitioner does not allege, however, that he used a contraceptive, or that pregnancy could not have resulted from the conduct with which he -was charged. But even assuming the petitioner’s standing to raise the claim of overbreadth, it is clear that a statute recognizing the defenses he suggests would encounter difficult if not impossible problems of proof.

11

See, e. g., Phipps-Yonas, Teenage Pregnancy and Motherhood, 50 Am. J. Orthopsychiatry 403, 412 (1980). See also State v. Bundlett, 391 A. 2d 815, 819, n. 13, 822 (Me.); Bundlett v. Oliver, 607 F. 2d 495, 502 (CA1).

12

See Barnes v. State, 244 Ga. 302, 260 S. E. 2d 40; see generally Orr & Prietto, supra; Jaffee, Dynneson, & ten Bensel, supra; Chilman, supra.

13

The fact is that a gender-neutral statute would not necessarily lead to a closer fit with the aim of reducing the problems associated with teenage pregnancy. If both parties were equally liable to prosecution, a female would be far less likely to complain; the very complaint would be self-incriminating. Accordingly, it is possible that a gender-neutral statute would result in fewer prosecutions than the one before us.

In any event, a state legislature is free to address itself to what it believes to be the most serious aspect of a broader problem. “[T]he Equal Protection Clause does not require that a State must choose between attacking every aspect of a problem or not attacking the problem at all.” Dandridge v. Williams, 397 U. S. 471, 486-487; see also Williamson v. Lee Optical Co., 348 U. S. 483.

Justice Blackmun,

concurring in the judgment.

It is gratifying that the plurality recognizes that “[a]t the risk of stating the obvious, teenage pregnancies . . . have increased dramatically over the last two decades” and “have significant social, medical, and economic consequences for both the mother and her child, and the State.” Ante, at 470 (footnotes omitted). There have been times when I have wondered whether the Court was capable of this perception, particularly when it has struggled with the different but not unrelated problems that attend abortion issues. See, for example, the opinions (and the dissenting opinions) in Beal v. Doe, 432 U. S. 438 (1977); Maher v. Roe, 432 U. S. 464 (1977) ; Poelker v. Doe, 432 U. S. 519 (1977); Harris v. McRae, 448 U. S. 297 (1980); Williams v. Zbaraz, 448 U. S. 358 (1980); and today’s opinion in H. L. v. Matheson, ante, p. 389.

Some might conclude that the two uses of the criminal sanction — here flatly to forbid intercourse in order to forestall teenage pregnancies, and in Matheson to prohibit a physician’s abortion procedure except upon notice to the parents of the pregnant minor — are vastly different proscriptions. But the basic social and privacy problems are much the same. Both Utah’s statute in Matheson and California’s statute in this case are legislatively created tools intended to achieve similar ends and addressed to the same societal concerns: the control and direction of young people’s sexual activities. The plurality opinion impliedly concedes as much when it notes that “approximately half of all teenage pregnancies end in abortion,” and that “those children who are born” are “likely candidates to become wards of the State,” Ante, at 471, and n. 6.

I, however, cannot vote to strike down the California statutory rape law, for I think it is a sufficiently reasoned and constitutional effort to control the problem at its inception. For me, there is an important difference between this state action and a State’s adamant and rigid refusal to face, or even to recognize, the “significant . . . consequences” — to the woman — of a forced or unwanted conception. I have found it difficult to rule constitutional, for example, state efforts to block, at that later point, a woman’s attempt to deal with the enormity of the problem confronting her, just as I have rejected state efforts to prevent women from rationally taking steps to prevent that problem from arising. See, e. g., Carey v. Population Services International, 431 U. S. 678 (1977). See also Griswold v. Connecticut, 381 U. S. 479 (1966). In contrast, I am persuaded that, although a minor has substantial privacy rights in intimate affairs connected with procreation, California’s efforts to prevent teenage pregnancy are to be viewed differently from Utah’s efforts to inhibit a woman from dealing with pregnancy once it has become an inevitability.

Craig v. Boren, 429 U. S. 190 (1976), was an opinion which, in large part, I joined, id., at 214. The plurality opinion in the present case points out, ante, at 468-469, the Court’s respective phrasings of the applicable test in Reed v. Reed, 404 U. S. 71, 76 (1971), and in Craig v. Boren, 429 U. S., at 197. I vote to affirm the judgment of the Supreme Court of California and to uphold the State’s gender-based classification on that test and as exemplified by those two cases and, by Schlesinger v. Ballard, 419 U. S. 498 (1975); Weinberger v. Wiesenfeld, 420 U. S. 636 (1975); and Kahn v. Shevin, 416 U. S. 351 (1974).

I note, also, that § 261.5 of the California Penal Code is just one of several California statutes intended to protect the juvenile. Justice Stewart, in his concurring opinion, appropriately observes that § 261.5 is “but one part of a broad statutory scheme that protects all minors from the problems and risks attendant upon adolescent sexual activity.” Ante, at 477.

I think, too, that it is only fair, with respect to this particular petitioner, to point out that his partner, Sharon, appears not to have been an unwilling participant in at least the initial stages of the intimacies that took place the night of June 3, 1978.* Petitioner’s and Sharon’s nonacquaintance with each other before the incident; their drinking; their withdrawal from the others of the group; their foreplay, in which she willingly participated and seems to have encouraged; and the closeness of their ages (a difference of only one year and 18 days) are factors that should make this case an unattractive one to prosecute at all, and especially to prosecute as a felony, rather than as a misdemeanor chargeable under § 261.5. But the State has chosen to prosecute in that manner, and the facts, I reluctantly conclude, may fit the crime.

*

Sharon at the preliminary hearing testified as follows:

“Q [by the Deputy District Attorney]. On June the 4th, at approximately midnight' — midnight of June the 3rd, were you in Rohnert Park?

“A [by Sharon], Yes.

“Q. Is that in Sonoma County?

“A. Yes.

“Q. Did anything unusual happen to you that night in Rohnert Park?

“A. Yes.

"Q. Would you briefly describe what happened that night? Did you see the defendant that night in Rohnert Park?

“A. Yes.

“Q. Where did you first meet him?

“A. At a bus stop.

“Q. Was anyone with you?

“A. My sister.

“Q. Was anyone with the defendant?

“A. Yes.

“Q. How many people were with the defendant?

“A. Two.

“Q. Now, after you met the defendant, what happened?

“A. We walked down to the railroad tracks.

“Q. What happened at the railroad tracks?

“A. We were drinking at the railroad tracks and we walked over to this bush and he started kissing me and stuff, and I was kissing him back, too, at first. Then, I was telling him to stop—

“Q. Yes.

“A. —and I was telling him to slow down and stop. He said, ‘Okay, okay.’ But then he just kept doing it. He just kept doing it and then my sister and two other guys came over to where we were and my sister said — told me to get up and come home. And then I didn’t—

“Q. Yes.

“A. —and then my sister and—

“Q. All right.

“A. —David, one of the boys that were there, started walking home and we stayed there and then later—

“Q. All right.

“A. —Bruce left Michael, you know.

"The Court: Michael being the defendant?

"The Witness: Yeah. We was laying there and we were kissing each other, and then he asked me if I wanted to walk him over to the park; so we walked over to the park and we sat down on a bench and then he started kissing me again and we were laying on the bench. And he told me to take my pants off.

“I said, ‘No,’ and I was trying to get up and he hit me back down on the bench and then I just said to myself, ‘Forget it,’ and I let him do what he wanted to do and he took my pants off and he was telling me to put my legs around him and stuff—

“Q. Did you have sexual intercourse with the defendant?

“A. Yeah.

“Q. He did put his penis into your vagina?

“A. Yes. '

“Q. You said that he hit you?

“A. Yeah.

“Q. How did he hit you?

“A. He slugged me in the face.

“Q. With what did he slug you?

“A. His fist.

“Q. Where abouts in the face?

“A. On my chin.

“Q. As a result of that, did you have any bruises or any kind of an injury?

“A. Yeah.

“Q. What happened?

“A. I had bruises.

“The Court: Did he hit you one time or did he hit you more than once?

“The Witness: He hit me about two or three times.

“Q. Now, during the course of that evening, did the defendant ask you your age?

“A. Yeah.

“Q. And what did you tell him?

“A. Sixteen.

“Q. Did you tell him you were sixteen?

“A. Yes.

“Q. Now, you said you had been drinking, is that correct?

“A. Yes.

“Q. Would you describe your condition as a result of the drinking?

“A. I was a little drunk.” App. 20-23.

CROSS-EXAMINATION

“Q. Did you go off with Mr. M. away from the others?

“A. Yeah.

“Q. Why did you do that?

“A. I don’t know. I guess I wanted to.

“Q. Did you have any need to go to the bathroom when you were there.

“A. Yes.

“Q. And what did you do?

“A. Me and my sister walked down the railroad tracks to some bushes and went to the bathroom.

“Q. Now, you and Mr. M., as I understand it, went off into the bushes, is that correct?

“A. Yes.

“Q. Okay. And what did you do when you and Mr. M. were there in the bushes?

“A. We were kissing and hugging.

“Q. Were you sitting up?

“A. We were laying down.

“Q. You were lying down. This was in the bushes?

“A. Yes.

“Q. How far away from the rest of them were you?

“A. They were just bushes right next to the railroad tracks. We just walked off into the bushes; not very far.

“Q. So your sister and the other two boys came over to where you were, you and Michael were, is that right?

“A. Yeah.

“Q. What did they say to you, if you remember?

“A. My sister didn’t say anything. She said, 'Come on, Sharon, let’s go home.’

“Q. She asked you to go home with her?

“A. (Affirmative nod.)

“Q. Did you go home with her?

“A. No.

“Q. You wanted to stay with Mr. M.?

“A. I don’t know.

"Q.'Was this before or after he hit you?

“A. Before.

“Q. What happened in the five minutes that Bruce stayed there with you and Michael?

“A. I don’t remember.

“Q. You don’t remember at all?

“A. (Negative head shake.)

“Q. Did you have occasion at that time to kiss Bruce?

“A. Yeah.

“Q. You did? You were kissing Bruce at that time?

“A. (Affirmative nod.)

“Q. Was Bruce kissing you?

“A. Yes.

“Q. And were you standing up at this time?

“A. No, we were sitting down.

“Q. Okay. So at this point in time you had left Mr. M. and you were hugging and kissing with Bruce, is that right?

“A. Yeah.

“Q. And you were sitting up.

"A. Yes.

“Q. Was your sister still there then?

“A. No. Yeah, she was at first.

“Q. What was she doing?

“A. She was standing up with Michael and David.

“Q. Yes. Was she doing anything with Michael and David?

“A. No, I don’t think so.

“Q. Whose idea was it for you and Bruce to kiss? Did you initiate that?

“A. Yes.

“Q. What happened after Bruce left?

“A. Michael asked me if I wanted to go walk to the park.

“Q. And what did you say?

“A. I said, ‘Yés.’

“Q. And then what happened?

“A. We walked to the park.

“Q. How long did it take you to get to the park?

“A. About ten or fifteen minutes.

“Q. And did you walk there?

“A. Yes.

“Q. Did Mr. M. ever mention his name?

“A. Yes.” Id., at 27-32.

Justice Brennan,

with whom Justices White and Marshall join, dissenting.

I

It is disturbing to find the Court so splintered on a case that presents such a straightforward issue: Whether the admittedly gender-based classification in Cal. Penal Code Ann. §261.5 (West Supp. 1981) bears a sufficient relationship to the State’s asserted goal of preventing teenage pregnancies to survive the “mid-level” constitutional scrutiny mandated by Craig v. Boren, 429 U. S. 190 (1976).1 Applying the analytical framework provided by our precedents, I am convinced that there is only one proper resolution of this issue: the classification must be declared unconstitutional. I fear that the plurality opinion and Justices Stewart and Blackmun reach the opposite result by placing too much emphasis on the desirability of achieving the State’s asserted statutory goal — prevention of teenage pregnancy — and not enough emphasis on the fundamental question of whether the sex-based discrimination in the California statute is substantially related to the achievement of that goal.2

II

After some uncertainty as to the proper framework for analyzing equal protection challenges to statutes containing gender-based classifications, see ante, at 468, this Court settled upon the proposition that a statute containing a gender-based classification cannot withstand constitutional challenge unless the classification is substantially related to the achievement of an important governmental objective. Kirchberg v. Feenstra, ante, at 459; Wengler v. Druggists Mutual Ins. Co., 446 U. S. 142, 150 (1980); Califano v. Westcott, 443 U. S. 76, 85 (1979); Caban v. Mohammed, 441 U. S. 380, 388 (1979); Orr v. Orr, 440 U. S. 268, 279 (1979); Califano v. Goldfarb, 430 U. S. 199, 210-211 (1977); Califano v. Webster, 430 U. S. 313, 316-317 (1977); Craig v. Boren, supra, at 197. This analysis applies whether the classification discriminates against males or against females. Caban v. Mohammed, supra, at 394; Orr v. Orr, supra, at 278-279; Craig v. Boren, supra, at 204. The burden is on the government to prove both the importance of its asserted objective and the substantial relationship between the classification and that objective. See Kirchberg v. Feenstra, ante, at 461; Wengler v. Druggists Mutual Ins. Co., supra, at 151-152; Caban v. Mohammed, supra, at 393; Craig v. Boren, supra, at 204. And the State cannot meet that burden without showing that a gender-neutral statute would be a less effective means of achieving that goal. Wengler v. Druggists Mutual Ins. Co., supra, at 151-152; Orr v. Orr, supra, at 281, 283.3

The State of California vigorously asserts that the “important governmental objective” to be served by § 261.5 is the prevention of teenage pregnancy. It claims that its statute furthers this goal by deterring sexual activity by males — the class of persons it considers more responsible for causing those pregnancies.4 But even assuming that prevention of teenage pregnancy is an important governmental objective and that it is in fact an objective of § 261.5, see infra, at 494-496, California still has the burden of proving that there are fewer teenage pregnancies under its gender-based statutory rape law than there would be if the law were gender neutral. To meet this burden, the State must show that because its statutory rape law punishes only males, and not females, it more effectively deters minor females from having sexual intercourse.5

The plurality assumes that a gender-neutral statute would be less effective than § 261.5 in deterring sexual activity because a gender-neutral statute would create significant enforcement problems. The plurality thus accepts the State’s assertion that

“a female is surely less likely to report violations of the statute if she herself would be subject to criminal prosecution. In an area already fraught with prosecutorial difficulties, we decline to hold that the Equal Protection Clause requires a legislature to enact a statute so broad that it may well be incapable of enforcement.” Ante, at 473-474 (footnotes omitted).

However, a State’s bare assertion that its gender-based statutory classification substantially furthers an important governmental interest is not enough to meet its burden of proof under Craig v. Boren. Rather, the State must produce evidence that will persuade the court that its assertion is true. See Craig v. Boren, 429 U. S., at 200-204.

The State has not produced such evidence in this case. Moreover, there are at least two serious flaws in the State’s assertion that law enforcement problems created by a gender-neutral statutory rape law would make such a statute less effective than a gender-based statute in deterring sexual activity.

First, the experience of other jurisdictions, and California itself, belies the plurality’s conclusion that a gender-neutral statutory rape law “may well be incapable of enforcement.” There are now at least 37 States that have enacted gender-neutral statutory rape laws. Although most of these laws protect young persons (of either sex) from the sexual exploitation of older individuals, the laws of Arizona, Florida, and Illinois permit prosecution of both minor females and minor males for engaging in mutual sexual conduct.6 California has introduced no evidence that those States have been handicapped by the enforcement problems the plurality finds so persuasive.7 Surely, if those States could provide such evidence, we might expect that California would have introduced it.

In addition, the California Legislature in recent years has revised other sections of the Penal Code to make them gender-neutral. For example, Cal. Penal Code Ann. §§286 (b)(1) and 288a (b)(1) (West Supp. 1981), prohibiting sodomy and oral copulation with a “person who is under 18 years of age,” could cause two minor homosexuals to be subjected to criminal sanctions for engaging in mutually consensual conduct. Again, the State has introduced no evidence to explain why a gender-neutral statutory rape law would be any more difficult to enforce than those statutes.

The second flaw in the State’s assertion is that even assuming that a gender-neutral statute would be more difficult to enforce, the State has still not shown that those enforcement problems would make such a statute less effective than a gender-based statute in deterring minor females from engaging in sexual intercourse.8 Common sense, however, suggests that a gender-neutral statutory rape law is potentially a greater deterrent of sexual activity than a gender-based law, for the simple reason that a gender-neutral law subjects both men and women to criminal sanctions and thus arguably has a deterrent effect on twice as many potential violators. Even if fewer persons were prosecuted under the gender-neutral law, as the State suggests, it would still be true that twice as many persons would be subject to arrest. The State’s failure to prove that a gender-neutral law would be a less effective deterrent than a gender-based law, like the State’s failure to prove that a gender-neutral law would be difficult to enforce, should have led this Court to invalidate § 261.5.

Ill

Until very recently, no California court or commentator had suggested that the purpose of California’s statutory rape law was to protect young women from the risk of pregnancy. Indeed, the historical development of § 261.5 demonstrates that the law was initially enacted on the premise that young women, in contrast to young men, were to be deemed legally incapable of consenting to an act of sexual intercourse.9 Because their chastity was considered particularly precious, those young women were felt to be uniquely in need of the State’s protection.10 In contrast, young men were assumed to be capable of making such decisions for themselves; the law therefore did not offer them any special protection.

It is perhaps because the gender classification in California’s statutory rape law was initially designed to further these outmoded sexual stereotypes, rather than to reduce the incidence of teenage pregnancies, that the State has been unable to demonstrate a substantial relationship beween the classification and its newly asserted goal. Cf. Califano v. Goldfarb, 430 U. S., at 223 (Stevens, J., concurring in judgment). But whatever the reason, the State has not shown that Cal. Penal Code § 261.5 is any more effective than a gender-neutral law would be in deterring minor females from engaging in sexual intercourse. It has therefore not met its burden of proving that the statutory classification is substantially related to the achievement of its asserted goal.

I would hold that § 261.5 violates the Equal Protection Clause of the Fourteenth Amendment, and I would reverse the judgment of the California Supreme Court.

1

The California Supreme Court acknowledged, and indeed the parties do not dispute, that Cal. Penal Code Ann. § 261.5 (West Supp. 1981) discriminates on the basis of sex. Ante, at 467. Because petitioner is male, he faces criminal felony charges and a possible prison term while his female partner remains immune from prosecution. The gender of the participants, not their relative responsibility, determines which of them is subject to criminal sanctions under §261.5.

As the California Supreme Court stated in People v. Hernandez, 61 Cal. 2d 529, 531, 393 P. 2d 673, 674 (1964) (footnote omitted):

“[E]ven in circumstances where a girl’s actual comprehension contradicts the law’s presumption [that a minor female is too innocent and naive to understand the implications and nature of her act], the male is deemed criminally responsible for the act, although himself young and naive and responding to advances which may have been made to him.”

2

2 None of the three opinions upholding the California statute fairly applies the equal protection analysis this Court has so carefully developed since Craig v. Boren, 429 U. S. 190 (1976). The plurality opinion, for example, focusing on the obvious and uncontested fact that only females can become pregnant, suggests that the statutory gender discrimination, rather than being invidious, actually ensures equality of treatment. Since only females are subject to a risk of pregnancy, the plurality opinion concludes that “[a] criminal sanction imposed solely on males . . . serves to roughly ‘equalize’ the deterrents on the sexes.” 'Ante, at 473. Justice Stewart adopts a similar approach. Recognizing that “females can become pregnant as the result of sexual intercourse; males cannot,” Justice Stewart concludes that “[y]oung women and men are not similarly situated with respect to the problems and risks associated with intercourse and pregnancy,” and therefore § 261.5 “is realistically related to the legitimate state purpose of reducing those problems and risks” (emphasis added). Ante, at 478, 479. Justice Blackmun, conceding that some limits must be placed on a State’s power to regulate “the control and direction of young people’s sexual activities,” also finds the statute constitutional. Ante, at 482. He distinguishes the State’s power in the abortion context, where the pregnancy has already occurred, from its power in the present context, where the “problem [is] at its inception.” He then concludes, without explanation, that “the California statutory rape law ... is a sufficiently reasoned and constitutional effort to control the problem at its inception.” Ibid.

All three of these approaches have a common failing. They overlook the fact that the State has not met its burden of proving that the gender discrimination in §261.5 is substantially related to the achievement of the State’s asserted statutory goal. My Brethren seem not to recognize that California has the burden of proving that a gender-neutral statutory rape law would be less effective than §261.5 in deterring sexual activity leading to teenage pregnancy. Because they fail to analyze the issue in these terms, I believe they reach an unsupportable result.

3

Gender-based statutory rape laws were struck down in Navedo v. Preisser, 630 F. 2d 636 (CA8 1980), United States v. Hicks, 625 F. 2d 216 (CA9 1980), and Meloon v. Helgemoe, 564 F. 2d 602 (CA1 1977), cert. denied, 436 U. S. 950 (1978), precisely because the government failed to meet this burden of proof.

4

In a remarkable display of sexual stereotyping, the California Supreme Court stated:

“The Legislature is well within its power in imposing criminal sanctions against males, alone, because they are the only persons who may physiologically cause the result which the law properly seeks to avoid.” 25 Cal. 3d 608, 612, 601 P. 2d 572, 575 (1979) (emphasis in original).

5

Petitioner has not questioned the State’s constitutional power to achieve its asserted objective by criminalizing consensual sexual activity. However, I note that our cases would not foreclose such a privacy challenge.

The State is attempting to reduce the incidence of teenage pregnancy by imposing criminal sanctions on those who engage in consensual sexual activity with minor females. We have stressed, however, that

“[i]f the right of privacy means anything, it is the right of the individual, married or single, to be free from unwarranted governmental intrusion into matters so fundamentally affecting a person as the decision whether to bear or beget a child.” Eisenstadt v. Baird, 405 U. S. 438, 453 (1972) (footnote omitted).

Minors, too, enjoy a right of privacy in connection with decisions affecting procreation. Carey v. Population Services International, 431 U. S. 678, 693 (1977). Thus, despite the suggestion of the plurality to the contrary, ante, at 472-473, n. 8, it is not settled that a State may rely on a pregnancy-prevention justification to make consensual sexual intercourse among minors a criminal act.

6

See-Ariz. Rev. Stat. Ann. § 13-1405 (1978); Fla. Stat. § 794.05 (1979); Ill. Rev. Stat., ch. 38, ¶ 11 — 5 (1979). In addition, eight other States permit both parties to be prosecuted when one of the participants to a consensual act of sexual intercourse is under the age of 16. See Kan. Stat. Ann. §21-3503 (1974); Mass. Gen. Laws Ann., ch. 265, §23 (West Supp. 1981); Mich. Comp. Laws §750.13 (1970); Mont. Code Ann. §§45-5-501 to 45-5-503 (1979); N. H. Rev. Stat. § 632-A:3 (Supp. 1979); Tenn. Code Ann. § 39-3705 (4) (Supp. 1979); Utah Code Ann. § 76-5-401 (Supp. 1979); Vt. Stat. Ann., Tit. 13, §3252 (3) (Supp. 1980).

7

There is a logical reason for this. In contrast to laws governing forcible rape, statutory rape laws apply to consensual sexual activity. Force is not an element of the crime. Since a woman who consents to an act of sexual intercourse is unlikely to report her partner to the police— whether or not she is subject to criminal sanctions — enforcement would not be undermined if the statute were to be made gender neutral. See n. 8, infra.

8

As it is, § 261.5 seems to be an ineffective deterrent of sexual activity. Cf. Carey v. Population Services International, supra, at 695 (substantial reason to doubt that limiting access to contraceptives will substantially discourage early sexual behavior). According to statistics provided by the State, an average of only 61 juvenile males and 352 adult males were arrested for statutory rape each year between 1975 and 1978. Brief for Respondent 19. During each of those years there were approximately one million Californian girls between the ages of 13-17. Cal. Dept. of Finance, Population Projections for California Counties, 1975-2020, with Age/Sex Detail to 2000, Series E-150 (1977). Although the record in this case does not indicate the incidence of sexual intercourse involving those girls during that period, the California State Department of Health estimates that there were almost 50,000 pregnancies among 13-to-17-year-old girls during 1976. Cal. Dept. of Health, Birth and Abortion Records, and Physician Survey of Office Abortions (1976). I think it is fair to speculate from this evidence that a comparison of the number of arrests for statutory rape in California with the number of acts of sexual intercourse involving minor females in that State would likely demonstrate to a male contemplating sexual activity with a minor female that his chances of being arrested are reassuringly low. I seriously question, therefore, whether § 261.5 as enforced has a substantial deterrent effect. See Craig v. Boren, 429 U. S., at 214 (Stevens, J., concurring).

9

California’s statutory rape law had its origins in the Statutes of Westminster enacted during the reign of Edward I at the close of the 13th century (3 Edw. 1, ch. 13 (1275); 13 Edw. 1, ch. 34 (1285)). The age of consent at that time was 12 years, reduced to 10 years in 1576 (18 Eliz. 1, ch. 7, § 4). This statute was part of the common law brought to the United States. Thus, when the first California penal statute was enacted, it contained a provision (1850 Cal. Stats., ch. 99, § 47, p. 234) that proscribed sexual intercourse with females under the age of 10. In 1889, the California statute was amended to make the age of consent 14 (1889 Cal. Stats., ch. 191, § 1, p. 223). In 1897, the age was advanced to 16 (1897 Cal. Stats., ch. 139, § 1, p. 201). In 1913 it was fixed at 18, where it now remains (1913 Cal. Stats., ch. 122, § 1, p. 212).

Because females generally have not reached puberty by the age of 10, it is inconceivable that a statute designed to prevent pregnancy would be directed at acts of sexual intercourse with females under that age.

The only legislative history available, the draftsmen’s notes to the Penal Code of 1872, supports the view that the purpose of California’s statutory rape law was to protect those who were too young to give consent. The draftsmen explained that the “[statutory rape] provision embodies the well settled rule of the existing law; that a girl under ten years of age is incapable of giving any consent to an act of intercourse which can reduce it below the grade of rape.” Code Commissioners’ note, subd. 1, following Cal. Penal Code § 261, p. Ill (1st ed. 1872). There was no mention whatever of pregnancy prevention. See also Note, Forcible and Statutory Rape: An Exploration of the Operation and Objectives of the Consent Standard, 62 Yale L. J. 55, 74-76 (1952).

10

Past decisions of the California courts confirm that the law was designed to protect the State’s young females from their own uninformed decisionmaking. In People v. Verdegreen, 106 Cal. 211, 214-215, 39 P. 607, 608-609 (1895), for example, the California Supreme Court stated: “The obvious purpose of [the statutory rape law] is the protection of society by protecting from violation the virtue of young and unsophisticated girls. ... It is the insidious approach and vile tampering with their persons that primarily undermines the virtue of young girls, and eventually destroys it; and the prevention of this, as much as the principal act, must undoubtedly have been the intent of the legislature.”

As recently as 1964, the California Supreme Court decided People v. Hernandez, 61 Cal. 2d, at 531, 393 P. 2d, at 674, in which it stated that the under-age female

“is presumed too innocent and naive to understand the implications and nature of her act. . . . The law’s concern with her capacity or lack thereof to so understand is explained in part by a popular conception of the social, moral and personal values which are preserved by the abstinence from sexual indulgence on the part of a young woman. An unwise disposition of her sexual favor is deemed to do harm both to herself and the social mores by which the community’s conduct patterns are established. Hence the law of statutory rape intervenes in an effort to avoid such a disposition.”

It was only in deciding Michael M. that the California Supreme Court decided, for the first time in the 130-year history of the statute, that pregnancy prevention had become one of the purposes of the statute.

Justice Stevens,

dissenting.

Local custom and belief — rather than statutory laws of venerable but doubtful ancestry — will determine the volume of sexual activity among unmarried teenagers.1 The empirical evidence cited by the plurality demonstrates the futility of the notion that a statutory prohibition will significantly affect the volume of that activity or provide a meaningful solution to the problems created by it.2 Nevertheless, as a matter of constitutional power, unlike my Brother Brennan, see ante, at 491, n. 5, I would have no doubt about the validity of a state law prohibiting all unmarried teenagers from engaging in sexual intercourse. The societal interests in reducing the incidence of venereal disease and teenage pregnancy are sufficient, in my judgment, to justify a prohibition of conduct that increases the risk of those harms.3

My conclusion that a nondiscriminatory prohibition would be constitutional does not help me answer the question whether a prohibition applicable to only half of the joint participants in the risk-creating conduct is also valid. It cannot be true that the validity of a total ban is an adequate justification for a selective prohibition; otherwise, the constitutional objection to discriminatory rules would be meaningless. The question in this case is whether the difference between males and females justifies this statutory discrimination based entirely on sex.4

The fact that the Court did not immediately acknowledge that the capacity to become pregnant is what primarily differentiates the female from the male5 does not impeach the validity of the plurality’s newly found wisdom. I think the plurality is quite correct in making the assumption that the joint act that this law seeks to prohibit creates a greater risk of harm for the female than for the male. But the plurality surely cannot believe that the risk of pregnancy confronted by the female — any more than the risk of venereal disease confronted by males as well as females — has provided an effective deterrent to voluntary female participation in the risk-creating conduct. Yet the plurality’s decision seems to rest on the assumption that the California Legislature acted on the basis of that rather fanciful notion.

In my judgment, the fact that a class of persons is especially vulnerable to a risk that a statute is designed to avoid is a reason for making the statute applicable to that class. The argument that a special need for protection provides a rational explanation for an exemption is one I simply do not comprehend.6

In this case, the fact that a female confronts a greater risk of harm than a male is a reason for applying the prohibition to her — not a reason for granting her a license to use her own judgment on whether or not to assume the risk. Surely, if we examine the problem from the point of view of society’s interest in preventing the risk-creating conduct from occurring at all, it is irrational to exempt 50% of the potential violators. See dissent of Justice Brennan, ante, at 493-494. And, if we view the government’s interest as that of a parens patriae seeking to protect its subjects from harming themselves, the discrimination is actually perverse. Would a rational parent making rules for the conduct of twin children of opposite sex simultaneously forbid the son and authorize the daughter to engage in conduct that is especially harmful to the daughter? That is the effect of this statutory classification.

If pregnancy or some other special harm is suffered by one of the two participants in the prohibited act, that special harm no doubt would constitute a legitimate mitigating factor in deciding what, if any, punishment might be appropriate in a given case. But from the standpoint of fashioning a general preventive rule — or, indeed, in determining appropriate punishment when neither party in fact has suffered any special harm — I regard a total exemption for the members of the more endangered class as utterly irrational.

In my opinion, the only acceptable justification for a general rule requiring disparate treatment of the two participants in a joint act must be a legislative judgment that one is more guilty than the other. The risk-creating conduct that this statute is designed to prevent requires the participation of two persons — one male and one female.7 In many situations it is probably true that one is the aggressor and the other is either an unwilling, or at least a less willing, participant in the joint act. If a statute authorized punishment of only one participant and required the prosecutor to prove that that participant had been the aggressor, I assume that the discrimination would be valid. Although the question is less clear, I also assume, for the purpose of deciding this case, that it would be permissible to punish only the male participant, if one element of the offense were proof that he had been the aggressor, or at least in some respects the more responsible participant in the joint act. The statute at issue in this case, however, requires no such proof. The question raised by this statute is whether the State, consistently with the Federal Constitution, may always punish the male and never the female when they are equally responsible or when the female is the more responsible of the two.

It would seem to me that an impartial lawmaker could give only one answer to that question. The fact that the California Legislature has decided to apply its prohibition only to the male may reflect a legislative judgment that in the typical case the male is actually the more guilty party. Any such judgment must, in turn, assume that the decision to engage in the risk-creating conduct is always — or at least typically— a male decision. If that assumption is valid, the statutory classification should also be valid. But what is the support for the assumption? It is not contained in the record of this case or in any legislative history or scholarly st,udy that has been called to our attention. I think it is supported to some extent by traditional attitudes toward male-female relationships. But the possibility that such a habitual attitude may reflect nothing more than an irrational prejudice makes it an insufficient justification for discriminatory treatment that is otherwise blatantly unfair. For, as I read this statute, it requires that one, and only one, of two equally guilty wrongdoers be stigmatized by a criminal conviction.

I cannot accept the State’s argument that the constitutionality of the discriminatory rule can be saved by an assumption that prosecutors will commonly invoke this statute only in cases that actually involve a forcible rape, but one that cannot be established by proof beyond a reasonable doubt.8 That assumption implies that a State has a legitimate interest in convicting a defendant on evidence that is constitutionally insufficient. Of course, the State may create a lesser-included offense that would authorize punishment of the more guilty party, but surely the interest in obtaining convictions on inadequate proof cannot justify a statute that punishes one who is equally or less guilty than his partner.9

Nor do I find at all persuasive the suggestion that this discrimination is adequately justified by the desire to encourage females to inform against their male partners. Even if the concept of a wholesale informant’s exemption were an acceptable enforcement device, what is the justification for defining the exempt class entirely by reference to sex rather than by reference to a more neutral criterion such as relative innocence? Indeed, if the exempt class is to be composed entirely of members of one sex, what is there to support the view that the statutory purpose will be better served by granting the informing license to females rather than to males? If a discarded male partner informs on a promiscuous female, a timely threat of prosecution might well prevent the precise harm the statute is intended to minimize.

Finally, even if my logic is faulty and there actually is some speculative basis for treating equally guilty males and females differently, I still believe that any such speculative justification would be outweighed by the paramount interest in evenhanded enforcement of the law. A rule that authorizes punishment of only one of two equally guilty wrongdoers violates the essence of the constitutional requirement that the sovereign must govern impartially.

I respectfully dissent.

1

“Common sense indicates that many young people will engage in sexual activity regardless of what the New York Legislature does; and further, that the incidence of venereal disease and premarital pregnancy is affected by the availability or unavailability of contraceptives. Although young persons theoretically may avoid those harms by practicing total abstention, inevitably many will not.” Carey v. Population Services International, 431 U. S. 678, 714 (Stevens, J., concurring in part and in judgment).

2

If a million teenagers became pregnant in 1976, see ante, at 470, n. 3, there must be countless violations of the California statute. The statistics cited by Justice Brennan also indicate, as he correctly observes, that the statute “seems to be an ineffective deterrent of sexual activity.” See ante, at 493-494, n. 8.

3

See Carey v. Population Services International, supra, at 713 (Stevens, J., concurring in part and in judgment).

4

Equal protection analysis is often said to involve different “levels of scrutiny.” It may be more accurate to say that the burden of sustaining an equal protection challenge is much heavier in some cases than in others. Racial classifications, which are subjected to “strict scrutiny,” are presumptively invalid because there is seldom, if ever, any legitimate reason for treating citizens differently because of their race. On the other hand, most economic classifications are presumptively valid because they are a necessary component of most regulatory programs. In cases involving discrimination between men and women, the natural differences between the sexes are sometimes relevant and sometimes wholly irrelevant. If those differences are obviously irrelevant, the discrimination should be treated as presumptively unlawful in the same way that racial classifications are presumptively unlawful. Cf. Califano v. Goldjarb, 430 U. S. 199, 223 (Stevens, J., concurring in judgment). But if, as in this case, there is an apparent connection between the discrimination and the fact that only women can become pregnant, it may be appropriate to presume that the classification is lawful. This presumption, however, may be overcome by a demonstration that the apparent justification for the discrimination is illusory or wholly inadequate. Thus, instead of applying a “mid-level” form of scrutiny in all sex discrimination cases, perhaps the burden is heavier in some than in others. Nevertheless, as I have previously suggested, the ultimate standard in these, as in all other equal protection cases, is essentially the same. See Craig v. Boren, 429 U. S. 190, 211-212 (Stevens, J., concurring). Professor Cox recently noted that however the level of scrutiny is described, in'the final analysis, “the Court is always deciding whether in its judgment the harm done to the disadvantaged class by the legislative classification is disproportionate to the public purposes the measure is likely to achieve.” Cox, Book Review, 94 Harv. L. Rev. 700, 706 (1981).

5

See General Electric Co. v. Gilbert, 429 U. S. 125, 162 (Stevens, J., dissenting).

6

A hypothetical racial classification will illustrate my point. Assume that skin pigmentation provides some measure of protection against cancer caused by exposure to certain chemicals in the atmosphere and, therefore, that white employees confront a greater risk than black employees in certain industrial settings. Would it be rational to require black employees to wear protective clothing but to exempt whites from that requirement? It seems to me that the greater risk of harm to white workers would be a reason for including them in the requirement — not for granting them an exemption.

7

In light of this indisputable biological fact, I find somewhat puzzling the California Supreme Court’s conclusion, quoted by the plurality, ante, at 467, that males “are the only persons who may physiologically cause the result which the law properly seeks to avoid.” 25 Cal. 3d 608, 612, 601 P. 2d 572, 575 (1979) (emphasis in original). Presumably, the California Supreme Court was referring to the equally indisputable biological fact that only females may become pregnant. However, if pregnancy results from sexual intercourse between two willing participants — and the California statute is directed at such conduct — I would find it difficult to conclude that the pregnancy was “caused” solely by the male participant.

8

According to the State of California:

“The statute is commonly employed in situations involving force, prostitution, pornography or coercion due to status relationships, and the state’s interest in these situations is apparent.” Brief for Respondent 3.

See also id., at 23-25. The State’s interest in these situations is indeed apparent and certainly sufficient to justify statutory prohibition of forcible rape, prostitution, pornography, and nonforeible, but nonetheless coerced, sexual intercourse. However, it is not at all apparent to me how this state interest can justify a statute not specifically directed to any of these offenses.

9

Both Justice Rehnquist and Justice Blackmun apparently attach significance to the testimony at the preliminary hearing indicating that the petitioner struck his partner. See opinion of Rehnquist, J., ante, at 467; opinion of Blackmun, J., ante, at 483-488, n. In light of the fact that the petitioner would be equally guilty of the crime charged in the complaint whether or not that testimony is true, it obviously has no bearing on the legal question presented by this case. The question is not whether “the facts ... fit the crime,” opinion of Blackmun, J., ante, at 487 — that is a question to be answered at trial — but rather, whether the statute defining the crime fits the constitutional requirement that justice be administered in an evenhanded fashion.

As Amended July 3, 1996.

ON WRITS OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT.

No. 94-2107, 976 F.2d 890, affirmed; No. 94-1941, 44 F.3d 1229, reversed and remanded.

GINSBURG, J., delivered the opinion of the Court, in which STEVENS, O'CONNOR, KENNEDY, SOUTER, and BREYER, JJ., joined. REHNQUIST, C. J., filed an opinion concurring in the judgment, post, p. 558. SCALIA, J., filed a dissenting opinion, post, p. 566. THOMAS, J., took no part in the consideration or decision of the case.

JUSTICE GINSBURG delivered the opinion of the Court.

Virginia's public institutions of higher learning include an incomparable military college, Virginia Military Institute (VMI). The United States maintains that the Constitution's equal protection guarantee precludes Virginia from reserving exclusively to men the unique educational opportunities VMI affords. We agree.

I

Founded in 1839, VMI is today the sole single-sex school among Virginia's 15 public institutions of higher learning. VMI's distinctive mission is to produce "citizen-soldiers," men prepared for leadership in civilian life and in military service. VMI pursues this mission through pervasive training of a kind not available anywhere else in Virginia. Assigning prime place to character development, VMI uses an "adversative method" modeled on English public schools and once characteristic of military instruction. VMI constantly endeavors to instill physical and mental discipline in its cadets and impart to them a strong moral code. The school's graduates leave VMI with heightened comprehension of their capacity to deal with duress and stress, and a large sense of accomplishment for completing the hazardous course.

VMI has notably succeeded in its mission to produce leaders; among its alumni are military generals, Members of Congress, and business executives. The school's alumni overwhelmingly perceive that their VMI training helped them to realize their personal goals. VMI's endowment reflects the loyalty of its graduates; VMI has the largest per-student endowment of all public undergraduate institutions in the Nation.

Neither the goal of producing citizen-soldiers nor VMI's implementing methodology is inherently unsuitable to women. And the school's impressive record in producing leaders has made admission desirable to some women. Nevertheless, Virginia has elected to preserve exclusively for men the advantages and opportunities a VMI education affords.

II

A

From its establishment in 1839 as one of the Nation's first state military colleges, see 1839 Va. Acts, ch. 20, VMI has remained financially supported by Virginia and "subject to the control of the [Virginia] General Assembly," Va. Code Ann. § 23-92 (1993). First southern college to teach engineering and industrial chemistry, see H. Wise, Drawing Out the Man: The VMI Story 13 (1978) (The VMI Story), VMI once provided teachers for the Commonwealth's schools, see 1842 Va. Acts, ch. 24, § 2 (requiring every cadet to teach in one of the Commonwealth's schools for a 2-year period). n1 Civil War strife threatened the school's vitality, but a resourceful superintendent regained legislative support by highlighting "VMI's great potential[,] through its technical know-how," to advance Virginia's postwar recovery. The VMI Story 47.

VMI today enrolls about 1,300 men as cadets. n2 Its academic offerings in the liberal arts, sciences, and engineering are also available at other public colleges and universities in Virginia. But VMI's mission is special. It is the mission of the school"'to produce educated and honorable men, prepared for the varied work of civil life, imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready as citizen-soldiers to defend their country in time of national peril.'" 766 F. Supp. 1407, 1425 (WD Va. 1991) (quoting Mission Study Committee of the VMI Board of Visitors, Report, May 16, 1986).In contrast to the federal service academies, institutions maintained "to prepare cadets for career service in the armed forces," VMI's program "is directed at preparation for both military and civilian life"; "only about 15% of VMI cadets enter career military service." 766 F. Supp., at 1432.

VMI produces its "citizen-soldiers" through "an adversative, or doubting, model of education" which features "physical rigor, mental stress, absolute equality of treatment, absence of privacy, minute regulation of behavior, and indoctrination in desirable values." Id., at 1421. As one Commandant of Cadets described it, the adversative method "'dissects the young student,'" and makes him aware of his "'limits and capabilities,'" so that he knows "'how far he can go with his anger, . . . how much he can take under stress, . . . exactly what he can do when he is physically exhausted.'" Id., at 1421-1422 (quoting Col. N. Bissell).

VMI cadets live in spartan barracks where surveillance is constant and privacy nonexistent; they wear uniforms, eat together in the mess hall, and regularly participate in drills. Id., at 1424, 1432. Entering students are incessantly exposed to the rat line, "an extreme form of the adversative model," comparable in intensity to Marine Corps boot camp. Id., at 1422. Tormenting and punishing, the rat line bonds new cadets to their fellow sufferers and, when they have completed the 7-month experience, to their former tormentors. Ibid.

VMI's "adversative model" is further characterized by a hierarchical "class system" of privileges and responsibilities, a "dyke system" for assigning a senior class mentor to each entering class "rat," and a stringently enforced "honor code," which prescribes that a cadet "'does not lie, cheat, steal nor tolerate those who do.'" Id., at 1422-1423.

VMI attracts some applicants because of its reputation as an extraordinarily challenging military school, and "because its alumni are exceptionally close to the school." Id., at 1421. "Women have no opportunity anywhere to gain the benefits of [the system of education at VMI]." Ibid.

B

In 1990, prompted by a complaint filed with the Attorney General by a female high-school student seeking admission to VMI, the United States sued the Commonwealth of Virginia and VMI, alleging that VMI's exclusively male admission policy violated the Equal Protection Clause of the Fourteenth Amendment. Id., at 1408. n3 Trial of the action consumed six days and involved an array of expert witnesses on each side. Ibid.

In the two years preceding the lawsuit, the District Court noted, VMI had received inquiries from 347 women, but had responded to none of them. Id., at 1436. "Some women, at least," the court said, "would want to attend the school if they had the opportunity." Id., at 1414. The court further recognized that, with recruitment, VMI could "achieve at least 10% female enrollment" -- "a sufficient 'critical mass' to provide the female cadets with a positive educational experience." Id., at 1437-1438. And it was also established that "some women are capable of all of the individual activities required of VMI cadets." Id., at 1412. In addition, experts agreed that if VMI admitted women, "the VMI ROTC experience would become a better training program from the perspective of the armed forces, because it would provide training in dealing with a mixed-gender army." Id., at 1441.

The District Court ruled in favor of VMI, however, and rejected the equal protection challenge pressed by the United States. That court correctly recognized that Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982), was the closest guide. 766 F. Supp., at 1410. There, this Court underscored that a party seeking to uphold government action based on sex must establish an "exceedingly persuasive justification" for the classification. Mississippi Univ. for Women, 458 U.S. at 724 (internal quotation marks omitted). To succeed, the defender of the challenged action must show "at least that the classification serves important governmental objectives and that the discriminatory means employed are substantially related to the achievement of those objectives." Ibid. (internal quotation marks omitted).

The District Court reasoned that education in "a single gender environment, be it male or female," yields substantial benefits. 766 F. Supp., at 1415. VMI's school for men brought diversity to an otherwise coeducational Virginia system, and that diversity was "enhanced by VMI's unique method of instruction." Ibid. If single-gender education for males ranks as an important governmental objective, it becomes obvious, the District Court concluded, that the only means of achieving the objective "is to exclude women from the all-male institution -- VMI." Ibid.

"Women are [indeed] denied a unique educational opportunity that is available only at VMI," the District Court acknowledged. Id., at 1432. But "[VMI's] single-sex status would be lost, and some aspects of the [school's] distinctive method would be altered" if women were admitted, id., at 1413: "Allowance for personal privacy would have to be made," id., at 1412; "physical education requirements would have to be altered, at least for the women," id., at 1413; the adversative environment could not survive unmodified, id., at 1412-1413. Thus, "sufficient constitutional justification" had been shown, the District Court held, "for continuing [VMI's] single-sex policy." Id., at 1413.

The Court of Appeals for the Fourth Circuit disagreed and vacated the District Court's judgment. The appellate court held: "The Commonwealth of Virginia has not . . . advanced any state policy by which it can justify its determination, under an announced policy of diversity, to afford VMI's unique type of program to men and not to women." 976 F.2d 890, 892 (1992).

The appeals court greeted with skepticism Virginia's assertion that it offers single-sex education at VMI as a facet of the Commonwealth's overarching and undisputed policy to advance "autonomy and diversity." The court underscored Virginia's nondiscrimination commitment: "'It is extremely important that [colleges and universities] deal with faculty, staff, and students without regard to sex, race, or ethnic origin.'" Id., at 899 (quoting 1990 Report of the Virginia Commission on the University of the 21st Century). "That statement, " the Court of Appeals said, "is the only explicit one that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions." 976 F.2d, at 899. Furthermore, the appeals court observed, in urging "diversity" to justify an all-male VMI, the Commonwealth had supplied "no explanation for the movement away from [single-sex education] in Virginia by public colleges and universities." Ibid. In short, the court concluded, "[a] policy of diversity which aims to provide an array of educational opportunities, including single-gender institutions, must do more than favor one gender." Ibid.

The parties agreed that "some women can meet the physical standards now imposed on men," id., at 896, and the court was satisfied that "neither the goal of producing citizen soldiers nor VMI's implementing methodology is inherently unsuitable to women," id., at 899. The Court of Appeals, however, accepted the District Court's finding that "at least these three aspects of VMI's program -- physical training, the absence of privacy, and the adversative approach -- would be materially affected by coeducation." Id., at 896-897. Remanding the case, the appeals court assigned to Virginia, in the first instance, responsibility for selecting a remedial course. The court suggested these options for the Commonwealth: Admit women to VMI; establish parallel institutions or programs; or abandon state support, leaving VMI free to pursue its policies as a private institution. Id., at 900. In May 1993, this Court denied certiorari. See 508 U.S. 946; see also ibid. (opinion of SCALIA, J., noting the interlocutory posture of the litigation).

C

In response to the Fourth Circuit's ruling, Virginia proposed a parallel program for women: Virginia Women's Institute for Leadership (VWIL). The 4-year, state-sponsored undergraduate program would be located at Mary Baldwin College, a private liberal arts school for women, and would be open, initially, to about 25 to 30 students. Although VWIL would share VMI's mission -- to produce "citizen-soldiers" -- the VWIL program would differ, as does Mary Baldwin College, from VMI in academic offerings, methods of education, and financial resources. See 852 F. Supp. 471, 476-477 (WD Va. 1994).

The average combined SAT score of entrants at Mary Baldwin is about 100 points lower than the score for VMI freshmen. See id., at 501. Mary Baldwin's faculty holds "significantly fewer Ph. D.'s than the faculty at VMI," id., at 502, and receives significantly lower salaries, see Tr. 158 (testimony of James Lott, Dean of Mary Baldwin College), reprinted in 2 App. in Nos. 94-1667 and 94-1717 (CA4) (hereinafter Tr.). While VMI offers degrees in liberal arts, the sciences, and engineering, Mary Baldwin, at the time of trial, offered only bachelor of arts degrees. See 852 F. Supp., at 503. A VWIL student seeking to earn an engineering degree could gain one, without public support, by attending Washington University in St. Louis, Missouri, for two years, paying the required private tuition. See ibid.

Experts in educating women at the college level composed the Task Force charged with designing the VWIL program; Task Force members were drawn from Mary Baldwin's own faculty and staff. Id., at 476. Training its attention on methods of instruction appropriate for "most women," the Task Force determined that a military model would be "wholly inappropriate" for VWIL. Ibid.; see 44 F.3d 1229, 1233 (CA4 1995).

VWIL students would participate in ROTC programs and a newly established, "largely ceremonial" Virginia Corps of Cadets, id., at 1234, but the VWIL House would not have a military format, 852 F. Supp., at 477, and VWIL would not require its students to eat meals together or to wear uniforms during the schoolday, id., at 495. In lieu of VMI's adversative method, the VWIL Task Force favored "a cooperative method which reinforces self-esteem." Id., at 476. In addition to the standard bachelor of arts program offered at Mary Baldwin, VWIL students would take courses in leadership, complete an off-campus leadership externship, participate in community service projects, and assist in arranging a speaker series. See 44 F.3d, at 1234.

Virginia represented that it will provide equal financial support for in-state VWIL students and VMI cadets, 852 F. Supp., at 483, and the VMI Foundation agreed to supply a $ 5.4625 million endowment for the VWIL program, id., at 499. Mary Baldwin's own endowment is about $ 19 million; VMI's is $ 131 million. Id., at 503. Mary Baldwin will add $ 35 million to its endowment based on future commitments; VMI will add $ 220 million. Ibid. The VMI Alumni Association has developed a network of employers interested in hiring VMI graduates. The Association has agreed to open its network to VWIL graduates, id., at 499, but those graduates will not have the advantage afforded by a VMI degree.

D

Virginia returned to the District Court seeking approval of its proposed remedial plan, and the court decided the plan met the requirements of the Equal Protection Clause. Id., at 473. The District Court again acknowledged evidentiary support for these determinations: "The VMI methodology could be used to educate women and, in fact, some women . . . may prefer the VMI methodology to the VWIL methodology." Id., at 481. But the "controlling legal principles," the District Court decided, "do not require the Commonwealth to provide a mirror image VMI for women." Ibid. The court anticipated that the two schools would "achieve substantially similar outcomes." Ibid. It concluded: "If VMI marches to the beat of a drum, then Mary Baldwin marches to the melody of a fife and when the march is over, both will have arrived at the same destination." Id., at 484.

A divided Court of Appeals affirmed the District Court's judgment. 44 F.3d 1229 (CA4 1995). This time, the appellate court determined to give "greater scrutiny to the selection of means than to the [Commonwealth's] proffered objective." Id., at 1236. The official objective or purpose, the court said, should be reviewed deferentially. Ibid. Respect for the "legislative will," the court reasoned, meant that the judiciary should take a "cautious approach," inquiring into the "legitimacy" of the governmental objective and refusing approval for any purpose revealed to be "pernicious." Ibid.

"Providing the option of a single-gender college education may be considered a legitimate and important aspect of a public system of higher education," the appeals court observed, id., at 1238; that objective, the court added, is "not pernicious," id., at 1239. Moreover, the court continued, the adversative method vital to a VMI education "has never been tolerated in a sexually heterogeneous environment." Ibid. The method itself "was not designed to exclude women," the court noted, but women could not be accommodated in the VMI program, the court believed, for female participation in VMI's adversative training "would destroy . . . any sense of decency that still permeates the relationship between the sexes." Ibid.

Having determined, deferentially, the legitimacy of Virginia's purpose, the court considered the question of means. Exclusion of "men at Mary Baldwin College and women at VMI," the court said, was essential to Virginia's purpose, for without such exclusion, the Commonwealth could not "accomplish [its] objective of providing single-gender education." Ibid.

The court recognized that, as it analyzed the case, means merged into end, and the merger risked "bypassing any equal protection scrutiny." Id., at 1237. The court therefore added another inquiry, a decisive test it called "substantive comparability." Ibid. The key question, the court said, was whether men at VMI and women at VWIL would obtain "substantively comparable benefits at their institution or through other means offered by the State." Ibid. Although the appeals court recognized that the VWIL degree "lacks the historical benefit and prestige" of a VMI degree, it nevertheless found the educational opportunities at the two schools "sufficiently comparable." Id., at 1241.

Senior Circuit Judge Phillips dissented. The court, in his judgment, had not held Virginia to the burden of showing an "'exceedingly persuasive [justification]' " for the Commonwealth's action. Id., at 1247 (quoting Mississippi Univ. for Women, 458 U.S. at 724). In Judge Phillips' view, the court had accepted "rationalizations compelled by the exigencies of this litigation," and had not confronted the Commonwealth's "actual overriding purpose." 44 F.3d, at 1247. That purpose, Judge Phillips said, was clear from the historical record; it was "not to create a new type of educational opportunity for women, . . . nor to further diversify the Commonwealth's higher education system[,] . . . but [was] simply . . . to allow VMI to continue to exclude women in order to preserve its historic character and mission." Ibid.

Judge Phillips suggested that the Commonwealth would satisfy the Constitution's equal protection requirement if it "simultaneously opened single-gender undergraduate institutions having substantially comparable curricular and extra-curricular programs, funding, physical plant, administration and support services, and faculty and library resources." Id., at 1250. But he thought it evident that the proposed VWIL program, in comparison to VMI, fell "far short . . . from providing substantially equal tangible and intangible educational benefits to men and women." Ibid.

The Fourth Circuit denied rehearing en banc. 52 F.3d 90 (1995). Circuit Judge Motz, joined by Circuit Judges Hall, Murnaghan, and Michael, filed a dissenting opinion. n4 Judge Motz agreed with Judge Phillips that Virginia had not shown an "'exceedingly persuasive justification'" for the disparate opportunities the Commonwealth supported. Id., at 92 (quoting Mississippi Univ. for Women, 458 U.S. at 724). She asked: "[H]ow can a degree from a yet to be implemented supplemental program at Mary Baldwin be held 'substantively comparable' to a degree from a venerable Virginia military institution that was established more than 150 years ago?" 52 F.3d, at 93. "Women need not be guaranteed equal 'results,'" Judge Motz said, "but the Equal Protection Clause does require equal opportunity . . . [and] that opportunity is being denied here." Ibid.

III

The cross-petitions in this case present two ultimate issues. First, does Virginia's exclusion of women from the educational opportunities provided by VMI -- extraordinary opportunities for military training and civilian leadership development -- deny to women "capable of all of the individual activities required of VMI cadets," 766 F. Supp., at 1412, the equal protection of the laws guaranteed by the Fourteenth Amendment? Second, if VMI's "unique" situation, id., at 1413 -- as Virginia's sole single-sex public institution of higher education -- offends the Constitution's equal protection principle, what is the remedial requirement?

IV

We note, once again, the core instruction of this Court's pathmarking decisions in J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 136-137, 128 L. Ed. 2d 89, 114 S. Ct. 1419, and n. 6 (1994), and Mississippi Univ. for Women, 458 U.S. at 724 (internal quotation marks omitted): Parties who seek to defend gender-based government action must demonstrate an "exceedingly persuasive justification" for that action.

Today's skeptical scrutiny of official action denying rights or opportunities based on sex responds to volumes of history. As a plurality of this Court acknowledged a generation ago, "our Nation has had a long and unfortunate history of sex discrimination." Frontiero v. Richardson, 411 U.S. 677, 684, 36 L. Ed. 2d 583, 93 S. Ct. 1764 (1973). Through a century plus three decades and more of that history, women did not count among voters composing "We the People"; n5 not until 1920 did women gain a constitutional right to the franchise. Id., at 685. And for a half century thereafter, it remained the prevailing doctrine that government, both federal and state, could withhold from women opportunities accorded men so long as any "basis in reason" could be conceived for the discrimination. See, e. g., Goesaert v. Cleary, 335 U.S. 464, 467, 93 L. Ed. 163, 69 S. Ct. 198 (1948) (rejecting challenge of female tavern owner and her daughter to Michigan law denying bartender licenses to females -- except for wives and daughters of male tavern owners; Court would not "give ear" to the contention that "an unchivalrous desire of male bartenders to . . . monopolize the calling" prompted the legislation).

In 1971, for the first time in our Nation's history, this Court ruled in favor of a woman who complained that her State had denied her the equal protection of its laws. Reed v. Reed, 404 U.S. 71, 73, 30 L. Ed. 2d 225, 92 S. Ct. 251 (holding unconstitutional Idaho Code prescription that, among "'several persons claiming and equally entitled to administer [a decedent's estate], males must be preferred to females'"). Since Reed, the Court has repeatedly recognized that neither federal nor state government acts compatibly with the equal protection principle when a law or official policy denies to women, simply because they are women, full citizenship stature -- equal opportunity to aspire, achieve, participate in and contribute to society based on their individual talents and capacities. See, e. g., Kirchberg v. Feenstra, 450 U.S. 455, 462-463, 67 L. Ed. 2d 428, 101 S. Ct. 1195 (1981) (affirming invalidity of Louisiana law that made husband "head and master" of property jointly owned with his wife, giving him unilateral right to dispose of such property without his wife's consent); Stanton v. Stanton, 421 U.S. 7, 43 L. Ed. 2d 688, 95 S. Ct. 1373 (1975) (invalidating Utah requirement that parents support boys until age 21, girls only until age 18).

Without equating gender classifications, for all purposes, to classifications based on race or national origin, n6 the Court, in post-Reed decisions, has carefully inspected official action that closes a door or denies opportunity to women (or to men). See J. E. B., 511 U.S. at 152 (KENNEDY, J., concurring in judgment) (case law evolving since 1971 "reveal[s] a strong presumption that gender classifications are invalid"). To summarize the Court's current directions for cases of official classification based on gender: Focusing on the differential treatment or denial of opportunity for which relief is sought, the reviewing court must determine whether the proffered justification is "exceedingly persuasive." The burden of justification is demanding and it rests entirely on the State. See Mississippi Univ. for Women, 458 U.S. at 724. The State must show "at least that the [challenged] classification serves 'important governmental objectives and that the discriminatory means employed' are 'substantially related to the achievement of those objectives.'" Ibid. (quoting Wengler v. Druggists Mut. Ins. Co., 446 U.S. 142, 150, 64 L. Ed. 2d 107, 100 S. Ct. 1540 (1980)). The justification must be genuine, not hypothesized or invented post hoc in response to litigation. And it must not rely on overbroad generalizations about the different talents, capacities, or preferences of males and females. See Weinberger v. Wiesenfeld, 420 U.S. 636, 643, 648, 43 L. Ed. 2d 514, 95 S. Ct. 1225 (1975); Califano v. Goldfarb, 430 U.S. 199, 223-224, 51 L. Ed. 2d 270, 97 S. Ct. 1021 (1977) (STEVENS, J., concurring in judgment).

The heightened review standard our precedent establishes does not make sex a proscribed classification. Supposed "inherent differences" are no longer accepted as a ground for race or national origin classifications. See Loving v. Virginia, 388 U.S. 1, 18 L. Ed. 2d 1010, 87 S. Ct. 1817 (1967). Physical differences between men and women, however, are enduring: "The two sexes are not fungible; a community made up exclusively of one [sex] is different from a community composed of both." Ballard v. United States, 329 U.S. 187, 193, 91 L.

Ed. 181, 67 S. Ct. 261 (1946).

"Inherent differences" between men and women, we have come to appreciate, remain cause for celebration, but not for denigration of the members of either sex or for artificial constraints on an individual's opportunity. Sex classifications may be used to compensate women "for particular economic disabilities [they have] suffered," Califano v. Webster, 430 U.S. 313, 320, 51 L. Ed. 2d 360, 97 S. Ct. 1192 (1977) (per curiam), to "promote equal employment opportunity," see California Fed. Sav. & Loan Assn. v. Guerra, 479 U.S. 272, 289, 93 L. Ed. 2d 613, 107 S. Ct. 683 (1987), to advance full development of the talent and capacities of our Nation's people. n7 But such classifications may not be used, as they once were, see Goesaert, 335 U.S. at 467, to create or perpetuate the legal, social, and economic inferiority of women.

Measuring the record in this case against the review standard just described, we conclude that Virginia has shown no "exceedingly persuasive justification" for excluding all women from the citizen-soldier training afforded by VMI. We therefore affirm the Fourth Circuit's initial judgment, which held that Virginia had violated the Fourteenth Amendment's Equal Protection Clause. Because the remedy proffered by Virginia -- the Mary Baldwin VWIL program -- does not cure the constitutional violation, i.e., it does not provide equal opportunity, we reverse the Fourth Circuit's final judgment in this case.

V

The Fourth Circuit initially held that Virginia had advanced no state policy by which it could justify, under equal protection principles, its determination "to afford VMI's unique type of program to men and not to women." 976 F.2d, at 892. Virginia challenges that "liability" ruling and asserts two justifications in defense of VMI's exclusion of women. First, the Commonwealth contends, "single-sex education provides important educational benefits," Brief for Cross-Petitioners 20, and the option of single-sex education contributes to "diversity in educational approaches," id., at 25. Second, the Commonwealth argues, "the unique VMI method of character development and leadership training," the school's adversative approach, would have to be modified were VMI to admit women. Id., at 33-36 (internal quotation marks omitted). We consider these two justifications in turn.

A

Single-sex education affords pedagogical benefits to at least some students, Virginia emphasizes, and that reality is uncontested in this litigation. n8 Similarly, it is not disputed that diversity among public educational institutions can serve the public good. But Virginia has not shown that VMI was established, or has been maintained, with a view to diversifying, by its categorical exclusion of women, educational opportunities within the Commonwealth. In cases of this genre, our precedent instructs that "benign" justifications proffered in defense of categorical exclusions will not be accepted automatically; a tenable justification must describe actual state purposes, not rationalizations for actions in fact differently grounded. See Wiesenfeld, 420 U.S. at 648, and n. 16 ("mere recitation of a benign [or] compensatory purpose" does not block "inquiry into the actual purposes" of government-maintained gender-based classifications); Goldfarb, 430 U.S. at 212-213 (rejecting government-proffered purposes after "inquiry into the actual purposes") (internal quotation marks omitted).

Mississippi Univ. for Women is immediately in point. There the State asserted, in justification of its exclusion of men from a nursing school, that it was engaging in "educational affirmative action" by "compensating for discrimination against women." 458 U.S. at 727. Undertaking a "searching analysis," id., at 728, the Court found no close resemblance between "the alleged objective" and "the actual purpose underlying the discriminatory classification," id., at 730. Pursuing a similar inquiry here, we reach the same conclusion.

Neither recent nor distant history bears out Virginia's alleged pursuit of diversity through single-sex educational options. In 1839, when the Commonwealth established VMI, a range of educational opportunities for men and women was scarcely contemplated. Higher education at the time was considered dangerous for women; n9 reflecting widely held views about women's proper place, the Nation's first universities and colleges -- for example, Harvard in Massachusetts, William and Mary in Virginia -- admitted only men. See E. Farello, A History of the Education of Women in the United States 163 (1970). VMI was not at all novel in this respect: In admitting no women, VMI followed the lead of the Commonwealth's flagship school, the University of Virginia, founded in 1819.

"No struggle for the admission of women to a state university," a historian has recounted, "was longer drawn out, or developed more bitterness, than that at the University of Virginia." 2 T. Woody, A History of Women's Education in the United States 254 (1929) (History of Women's Education). In 1879, the State Senate resolved to look into the possibility of higher education for women, recognizing that Virginia "'has never, at any period of her history,'" provided for the higher education of her daughters, though she "'has liberally provided for the higher education of her sons.'" Ibid. (quoting 10 Educ. J. Va. 212 (1879)). Despite this recognition, no new opportunities were instantly open to women. n10

Virginia eventually provided for several women's seminaries and colleges. Farmville Female Seminary became a public institution in 1884. See supra, at 521, n. 2. Two women's schools, Mary Washington College and James Madison University, were founded in 1908; another, Radford University, was founded in 1910. 766 F. Supp., at 1418-1419. By the mid-1970's, all four schools had become coeducational. Ibid.

Debate concerning women's admission as undergraduates at the main university continued well past the century's midpoint. Familiar arguments were rehearsed. If women were admitted, it was feared, they "would encroach on the rights of men; there would be new problems of government, perhaps scandals; the old honor system would have to be changed; standards would be lowered to those of other coeducational schools; and the glorious reputation of the university, as a school for men, would be trailed in the dust." 2 History of Women's Education 255.

Ultimately, in 1970, "the most prestigious institution of higher education in Virginia," the University of Virginia, introduced coeducation and, in 1972, began to admit women on an equal basis with men. See Kirstein v. Rector and Visitors of Univ. of Virginia, 309 F. Supp. 184, 186 (ED Va. 1970). A three-judge Federal District Court confirmed: "Virginia may not now deny to women, on the basis of sex, educational opportunities at the Charlottesville campus that are not afforded in other institutions operated by the State." Id., at 187.

Virginia describes the current absence of public single-sex higher education for women as "an historical anomaly." Brief for Cross-Petitioners 30. But the historical record indicates action more deliberate than anomalous: First, protection of women against higher education; next, schools for women far from equal in resources and stature to schools for men; finally, conversion of the separate schools to coeducation. The state legislature, prior to the advent of this controversy, had repealed "all Virginia statutes requiring individual institutions to admit only men or women." 766 F. Supp., at 1419. And in 1990, an official commission, "legislatively established to chart the future goals of higher education in Virginia," reaffirmed the policy "'of affording broad access" while maintaining "autonomy and diversity.'" 976 F.2d, at 898-899 (quoting Report of the Virginia Commission on the University of the 21st Century). Significantly, the Commission reported:"'Because colleges and universities provide opportunities for students to develop values and learn from role models, it is extremely important that they deal with faculty, staff, and students without regard to sex, race, or ethnic origin.'" Id., at 899 (emphasis supplied by Court of Appeals deleted).This statement, the Court of Appeals observed, "is the only explicit one that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions." Ibid.

Our 1982 decision in Mississippi Univ. for Women prompted VMI to reexamine its male-only admission policy. See 766 F. Supp., at 1427-1428. Virginia relies on that reexamination as a legitimate basis for maintaining VMI's single-sex character. See Reply Brief for Cross-Petitioners 6. A Mission Study Committee, appointed by the VMI Board of Visitors, studied the problem from October 1983 until May 1986, and in that month counseled against "change of VMI status as a single-sex college." See 766 F. Supp. at 1429 (internal quotation marks omitted). Whatever internal purpose the Mission Study Committee served -- and however well meaning the framers of the report -- we can hardly extract from that effort any commonwealth policy evenhandedly to advance diverse educational options. As the District Court observed, the Committee's analysis "primarily focused on anticipated difficulties in attracting females to VMI," and the report, overall, supplied "very little indication of how the conclusion was reached." Ibid.

In sum, we find no persuasive evidence in this record that VMI's male-only admission policy "is in furtherance of a state policy of 'diversity.'" See 976 F.2d, at 899. No such policy, the Fourth Circuit observed, can be discerned from the movement of all other public colleges and universities in Virginia away from single-sex education. See ibid. That court also questioned "how one institution with autonomy, but with no authority over any other state institution, can give effect to a state policy of diversity among institutions." Ibid. A purpose genuinely to advance an array of educational options, as the Court of Appeals recognized, is not served by VMI's historic and constant plan -- a plan to "afford a unique educational benefit only to males." Ibid. However "liberally" this plan serves the Commonwealth's sons, it makes no provision whatever for her daughters. That is not equal protection.

B

Virginia next argues that VMI's adversative method of training provides educational benefits that cannot be made available, unmodified, to women. Alterations to accommodate women would necessarily be "radical," so "drastic," Virginia asserts, as to transform, indeed "destroy," VMI's program. See Brief for Cross-Petitioners 34-36. Neither sex would be favored by the transformation, Virginia maintains: Men would be deprived of the unique opportunity currently available to them; women would not gain that opportunity because their participation would "eliminate the very aspects of [the] program that distinguish [VMI] from . . . other institutions of higher education in Virginia." Id., at 34.

The District Court forecast from expert witness testimony, and the Court of Appeals accepted, that coeducation would materially affect "at least these three aspects of VMI's program -- physical training, the absence of privacy, and the adversative approach." 976 F.2d, at 896-897. And it is uncontested that women's admission would require accommodations, primarily in arranging housing assignments and physical training programs for female cadets. See Brief for Cross-Respondent 11, 29-30. It is also undisputed, however, that "the VMI methodology could be used to educate women." 852 F. Supp., at 481. The District Court even allowed that some women may prefer it to the methodology a women's college might pursue. See ibid. "Some women, at least, would want to attend [VMI] if they had the opportunity," the District Court recognized, 766 F. Supp., at 1414, and "some women," the expert testimony established, "are capable of all of the individual activities required of VMI cadets," id., at 1412. The parties, furthermore, agree that "some women can meet the physical standards [VMI] now impose[s] on men." 976 F.2d, at 896. In sum, as the Court of Appeals stated, "neither the goal of producing citizen soldiers," VMI's raison d'etre, "nor VMI's implementing methodology is inherently unsuitable to women." Id., at 899.

In support of its initial judgment for Virginia, a judgment rejecting all equal protection objections presented by the United States, the District Court made "findings" on "gender-based developmental differences." 766 F. Supp., at 1434-1435. These "findings" restate the opinions of Virginia's expert witnesses, opinions about typically male or typically female "tendencies." Id., at 1434. For example, "males tend to need an atmosphere of adversativeness," while "females tend to thrive in a cooperative atmosphere." Ibid. "I'm not saying that some women don't do well under [the] adversative model," VMI's expert on educational institutions testified, "undoubtedly there are some [women] who do"; but educational experiences must be designed "around the rule," this expert maintained, and not "around the exception." Ibid. (internal quotation marks omitted).

The United States does not challenge any expert witness estimation on average capacities or preferences of men and women. Instead, the United States emphasizes that time and again since this Court's turning point decision in Reed v. Reed, 404 U.S. 71, 30 L. Ed. 2d 225, 92 S. Ct. 251 (1971), we have cautioned reviewing courts to take a "hard look" at generalizations or "tendencies" of the kind pressed by Virginia, and relied upon by the District Court. See O'Connor, Portia's Progress, 66 N. Y. U. L. Rev. 1546, 1551 (1991). State actors controlling gates to opportunity, we have instructed, may not exclude qualified individuals based on "fixed notions concerning the roles and abilities of males and females." Mississippi Univ. for Women, 458 U.S. at 725; see J. E. B., 511 U.S. at 139, n. 11 (equal protection principles, as applied to gender classifications, mean state actors may not rely on "overbroad" generalizations to make "judgments about people that are likely to . . . perpetuate historical patterns of discrimination").

It may be assumed, for purposes of this decision, that most women would not choose VMI's adversative method. As Fourth Circuit Judge Motz observed, however, in her dissent from the Court of Appeals' denial of rehearing en banc, it is also probable that "many men would not want to be educated in such an environment." 52 F.3d, at 93. (On that point, even our dissenting colleague might agree.) Education, to be sure, is not a "one size fits all" business. The issue, however, is not whether "women -- or men -- should be forced to attend VMI"; rather, the question is whether the Commonwealth can constitutionally deny to women who have the will and capacity, the training and attendant opportunities that VMI uniquely affords. Ibid.

The notion that admission of women would downgrade VMI's stature, destroy the adversative system and, with it, even the school, n11 is a judgment hardly proved, n12 a prediction hardly different from other "self-fulfilling prophec[ies]," see Mississippi Univ. for Women, 458 U.S. at 730, once routinely used to deny rights or opportunities. When women first sought admission to the bar and access to legal education, concerns of the same order were expressed. For example, in 1876, the Court of Common Pleas of Hennepin County, Minnesota, explained why women were thought ineligible for the practice of law. Women train and educate the young, the court said, which"forbids that they shall bestow that time (early and late) and labor, so essential in attaining to the eminence to which the true lawyer should ever aspire. It cannot therefore be said that the opposition of courts to the admission of females to practice . . . is to any extent the outgrowth of . . . 'old fogyism[.]' . . . It arises rather from a comprehension of the magnitude of the responsibilities connected with the successful practice of law, and a desire to grade up the profession." In re Application of Martha Angle Dorsett to Be Admitted to Practice as Attorney and Counselor at Law (Minn. C. P. Hennepin Cty., 1876), in The Syllabi, Oct. 21, 1876, pp. 5, 6 (emphasis added).A like fear, according to a 1925 report, accounted for Columbia Law School's resistance to women's admission, although"the faculty . . . never maintained that women could not master legal learning . . . . No, its argument has been . . . more practical. If women were admitted to the Columbia Law School, [the faculty] said, then the choicer, more manly and red-blooded graduates of our great universities would go to the Harvard Law School!" The Nation, Feb. 18, 1925, p. 173.

Medical faculties similarly resisted men and women as partners in the study of medicine. See R. Morantz-Sanchez, Sympathy and Science: Women Physicians in American Medicine 51-54, 250 (1985); see also M. Walsh, "Doctors Wanted: No Women Need Apply" 121-122 (1977) (quoting E. Clarke, Medical Education of Women, 4 Boston Med. & Surg. J. 345, 346 (1869) ("'God forbid that I should ever see men and women aiding each other to display with the scalpel the secrets of the reproductive system . . . .'")); cf. supra, at 536-537, n. 9. More recently, women seeking careers in policing encountered resistance based on fears that their presence would "undermine male solidarity," see F. Heidensohn, Women in Control? 201 (1992); deprive male partners of adequate assistance, see id., at 184-185; and lead to sexual misconduct, see C. Milton et al., Women in Policing 32-33 (1974). Field studies did not confirm these fears. See Heidensohn, supra, at 92-93; P. Bloch & D. Anderson, Policewomen on Patrol: Final Report (1974).

Women's successful entry into the federal military academies, n13 and their participation in the Nation's military forces, n14 indicate that Virginia's fears for the future of VMI may not be solidly grounded. n15 The Commonwealth's justification for excluding all women from "citizen-soldier" training for which some are qualified, in any event, cannot rank as "exceedingly persuasive," as we have explained and applied that standard.

Virginia and VMI trained their argument on "means" rather than "end," and thus misperceived our precedent. Single-sex education at VMI serves an "important governmental objective," they maintained, and exclusion of women is not only "substantially related," it is essential to that objective. By this notably circular argument, the "straightforward" test Mississippi Univ. for Women described, see 458 U.S. at 724-725, was bent and bowed.

The Commonwealth's misunderstanding and, in turn, the District Court's, is apparent from VMI's mission: to produce "citizen-soldiers," individuals"'imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready . . . to defend their country in time of national peril.'" 766 F. Supp., at 1425 (quoting Mission Study Committee of the VMI Board of Visitors, Report, May 16, 1986).Surely that goal is great enough to accommodate women, who today count as citizens in our American democracy equal in stature to men. Just as surely, the Commonwealth's great goal is not substantially advanced by women's categorical exclusion, in total disregard of their individual merit, from the Commonwealth's premier "citizen-soldier" corps. n16 Virginia, in sum, "has fallen far short of establishing the 'exceedingly persuasive justification,'" Mississippi Univ. for Women, 458 U.S. at 731, that must be the solid base for any gender-defined classification.

VI

In the second phase of the litigation, Virginia presented its remedial plan -- maintain VMI as a male-only college and create VWIL as a separate program for women. The plan met District Court approval. The Fourth Circuit, in turn, deferentially reviewed the Commonwealth's proposal and decided that the two single-sex programs directly served Virginia's reasserted purposes: single-gender education, and "achieving the results of an adversative method in a military environment." See 44 F.3d, at 1236, 1239. Inspecting the VMI and VWIL educational programs to determine whether they "afforded to both genders benefits comparable in substance, [if] not in form and detail," id., at 1240, the Court of Appeals concluded that Virginia had arranged for men and women opportunities "sufficiently comparable" to survive equal protection evaluation, id., at 1240-1241. The United States challenges this "remedial" ruling as pervasively misguided.

A

A remedial decree, this Court has said, must closely fit the constitutional violation; it must be shaped to place persons unconstitutionally denied an opportunity or advantage in "the position they would have occupied in the absence of [discrimination]. " See Milliken v. Bradley, 433 U.S. 267, 280, 53 L. Ed. 2d 745, 97 S. Ct. 2749 (1977) (internal quotation marks omitted). The constitutional violation in this case is the categorical exclusion of women from an extraordinary educational opportunity afforded men. A proper remedy for an unconstitutional exclusion, we have explained, aims to "eliminate [so far as possible] the discriminatory effects of the past" and to "bar like discrimination in the future." Louisiana v. United States, 380 U.S. 145, 154, 13 L. Ed. 2d 709, 85 S. Ct. 817 (1965).

Virginia chose not to eliminate, but to leave untouched, VMI's exclusionary policy. For women only, however, Virginia proposed a separate program, different in kind from VMI and unequal in tangible and intangible facilities. n17 Having violated the Constitution's equal protection requirement, Virginia was obliged to show that its remedial proposal "directly addressed and related to" the violation, see Milliken, 433 U.S. at 282, i. e., the equal protection denied to women ready, willing, and able to benefit from educational opportunities of the kind VMI offers. Virginia described VWIL as a "parallel program," and asserted that VWIL shares VMI's mission of producing "citizen-soldiers" and VMI's goals of providing "education, military training, mental and physical discipline, character . . . and leadership development." Brief for Respondents 24 (internal quotation marks omitted). If the VWIL program could not "eliminate the discriminatory effects of the past," could it at least "bar like discrimination in the future"? See Louisiana, 380 U.S. at 154. A comparison of the programs said to be "parallel" informs our answer. In exposing the character of, and differences in, the VMI and VWIL programs, we recapitulate facts earlier presented. See supra, at 520-523, 526-527.

VWIL affords women no opportunity to experience the rigorous military training for which VMI is famed. See 766 F. Supp., at 1413-1414 ("No other school in Virginia or in the United States, public or private, offers the same kind of rigorous military training as is available at VMI."); id., at 1421 (VMI "is known to be the most challenging military school in the United States"). Instead, the VWIL program "deemphasize[s]" military education, 44 F.3d, at 1234, and uses a "cooperative method" of education "which reinforces self-esteem," 852 F. Supp., at 476.

VWIL students participate in ROTC and a "largely ceremonial" Virginia Corps of Cadets, see 44 F.3d, at 1234, but Virginia deliberately did not make VWIL a military institute. The VWIL House is not a military-style residence and VWIL students need not live together throughout the 4-year program, eat meals together, or wear uniforms during the school day. See 852 F. Supp., at 477, 495. VWIL students thus do not experience the "barracks" life "crucial to the VMI experience," the spartan living arrangements designed to foster an "egalitarian ethic." See 766 F. Supp., at 1423-1424. "The most important aspects of the VMI educational experience occur in the barracks," the District Court found, id., at 1423, yet Virginia deemed that core experience nonessential, indeed inappropriate, for training its female citizen-soldiers.

VWIL students receive their "leadership training" in seminars, externships, and speaker series, see 852 F. Supp., at 477, episodes and encounters lacking the "physical rigor, mental stress, . . . minute regulation of behavior, and indoctrination in desirable values" made hallmarks of VMI's citizen-soldier training, see 766 F. Supp., at 1421. n18 Kept away from the pressures, hazards, and psychological bonding characteristic of VMI's adversative training, see id., at 1422, VWIL students will not know the "feeling of tremendous accomplishment" commonly experienced by VMI's successful cadets, id., at 1426.

Virginia maintains that these methodological differences are "justified pedagogically," based on "important differences between men and women in learning and developmental needs," "psychological and sociological differences" Virginia describes as "real" and "not stereotypes." Brief for Respondents 28 (internal quotation marks omitted). The Task Force charged with developing the leadership program for women, drawn from the staff and faculty at Mary Baldwin College, "determined that a military model and, especially VMI's adversative method, would be wholly inappropriate for educating and training most women." 852 F. Supp., at 476 (emphasis added). See also 44 F.3d, at 1233-1234 (noting Task Force conclusion that, while "some women would be suited to and interested in [a VMI-style experience]," VMI's adversative method "would not be effective for women as a group") (emphasis added). The Commonwealth embraced the Task Force view, as did expert witnesses who testified for Virginia. See 852 F. Supp., at 480-481.

As earlier stated, see supra, at 541-542, generalizations about "the way women are," estimates of what is appropriate for most women, no longer justify denying opportunity to women whose talent and capacity place them outside the average description. Notably, Virginia never asserted that VMI's method of education suits most men. It is also revealing that Virginia accounted for its failure to make the VWIL experience "the entirely militaristic experience of VMI" on the ground that VWIL "is planned for women who do not necessarily expect to pursue military careers." 852 F. Supp., at 478. By that reasoning, VMI's "entirely militaristic" program would be inappropriate for men in general or as a group, for "only about 15% of VMI cadets enter career military service." See 766 F. Supp., at 1432.

In contrast to the generalizations about women on which Virginia rests, we note again these dispositive realities: VMI's "implementing methodology" is not "inherently unsuitable to women," 976 F.2d, at 899; "some women . . . do well under [the] adversative model," 766 F. Supp., at 1434 (internal quotation marks omitted); "some women, at least, would want to attend [VMI] if they had the opportunity," id., at 1414; "some women are capable of all of the individual activities required of VMI cadets," id., at 1412, and "can meet the physical standards [VMI] now impose[s] on men," 976 F.2d, at 896. It is on behalf of these women that the United States has instituted this suit, and it is for them that a remedy must be crafted, n19 a remedy that will end their exclusion from a state-supplied educational opportunity for which they are fit, a decree that will "bar like discrimination in the future." Louisiana, 380 U.S. at 154.

B

In myriad respects other than military training, VWIL does not qualify as VMI's equal. VWIL's student body, faculty, course offerings, and facilities hardly match VMI's. Nor can the VWIL graduate anticipate the benefits associated with VMI's 157-year history, the school's prestige, and its influential alumni network.

Mary Baldwin College, whose degree VWIL students will gain, enrolls first-year women with an average combined SAT score about 100 points lower than the average score for VMI freshmen. 852 F. Supp., at 501. The Mary Baldwin faculty holds "significantly fewer Ph. D.'s," id., at 502, and receives substantially lower salaries, see Tr. 158 (testimony of James Lott, Dean of Mary Baldwin College), than the faculty at VMI.

Mary Baldwin does not offer a VWIL student the range of curricular choices available to a VMI cadet. VMI awards baccalaureate degrees in liberal arts, biology, chemistry, civil engineering, electrical and computer engineering, and mechanical engineering. See 852 F. Supp., at 503; Virginia Military Institute: More than an Education 11 (Govt. exh. 75, lodged with Clerk of this Court). VWIL students attend a school that "does not have a math and science focus," 852 F. Supp., at 503; they cannot take at Mary Baldwin any courses in engineering or the advanced math and physics courses VMI offers, see id., at 477.

For physical training, Mary Baldwin has "two multi-purpose fields" and "one gymnasium." Id., at 503. VMI has "an NCAA competition level indoor track and field facility; a number of multi-purpose fields; baseball, soccer and lacrosse fields; an obstacle course; large boxing, wrestling and martial arts facilities; an 11-laps-to-the-mile indoor running course; an indoor pool; indoor and outdoor rifle ranges; and a football stadium that also contains a practice field and outdoor track." Ibid.

Although Virginia has represented that it will provide equal financial support for in-state VWIL students and VMI cadets, id., at 483, and the VMI Foundation has agreed to endow VWIL with $ 5.4625 million, id., at 499, the difference between the two schools' financial reserves is pronounced. Mary Baldwin's endowment, currently about $ 19 million, will gain an additional $ 35 million based on future commitments; VMI's current endowment, $ 131 million -- the largest public college per-student endowment in the Nation -- will gain $ 220 million. Id., at 503.

The VWIL student does not graduate with the advantage of a VMI degree. Her diploma does not unite her with the legions of VMI "graduates [who] have distinguished themselves" in military and civilian life. See 976 F.2d, at 892-893. "[VMI] alumni are exceptionally close to the school," and that closeness accounts, in part, for VMI's success in attracting applicants. See 766 F. Supp., at 1421. A VWIL graduate cannot assume that the "network of business owners, corporations, VMI graduates and non-graduate employers . . . interested in hiring VMI graduates," 852 F. Supp., at 499, will be equally responsive to her search for employment, see 44 F.3d, at 1250 (Phillips, J., dissenting) ("the powerful political and economic ties of the VMI alumni network cannot be expected to open" for graduates of the fledgling VWIL program).

Virginia, in sum, while maintaining VMI for men only, has failed to provide any "comparable single-gender women's institution." Id., at 1241. Instead, the Commonwealth has created a VWIL program fairly appraised as a "pale shadow" of VMI in terms of the range of curricular choices and faculty stature, funding, prestige, alumni support and influence. See id., at 1250 (Phillips, J., dissenting).

Virginia's VWIL solution is reminiscent of the remedy Texas proposed 50 years ago, in response to a state trial court's 1946 ruling that, given the equal protection guarantee, African Americans could not be denied a legal education at a state facility. See Sweatt v. Painter, 339 U.S. 629, 94 L. Ed. 1114, 70 S. Ct. 848 (1950). Reluctant to admit African Americans to its flagship University of Texas Law School, the State set up a separate school for Heman Sweatt and other black law students. Id., at 632. As originally opened, the new school had no independent faculty or library, and it lacked accreditation. Id., at 633. Nevertheless, the state trial and appellate courts were satisfied that the new school offered Sweatt opportunities for the study of law "substantially equivalent to those offered by the State to white students at the University of Texas." Id., at 632 (internal quotation marks omitted).

Before this Court considered the case, the new school had gained "a faculty of five full-time professors; a student body of 23; a library of some 16,500 volumes serviced by a full-time staff; a practice court and legal aid association; and one alumnus who had become a member of the Texas Bar." Id., at 633. This Court contrasted resources at the new school with those at the school from which Sweatt had been excluded. The University of Texas Law School had a full-time faculty of 16, a student body of 850, a library containing over 65,000 volumes, scholarship funds, a law review, and moot court facilities. Id., at 632-633.

More important than the tangible features, the Court emphasized, are "those qualities which are incapable of objective measurement but which make for greatness" in a school, including "reputation of the faculty, experience of the administration, position and influence of the alumni, standing in the community, traditions and prestige." Id., at 634. Facing the marked differences reported in the Sweatt opinion, the Court unanimously ruled that Texas had not shown "substantial equality in the [separate] educational opportunities" the State offered. Id., at 633. Accordingly, the Court held, the Equal Protection Clause required Texas to admit African Americans to the University of Texas Law School. Id., at 636. In line with Sweatt, we rule here that Virginia has not shown substantial equality in the separate educational opportunities the Commonwealth supports at VWIL and VMI.

C

When Virginia tendered its VWIL plan, the Fourth Circuit did not inquire whether the proposed remedy, approved by the District Court, placed women denied the VMI advantage in "the position they would have occupied in the absence of [discrimination]." Milliken, 433 U.S. at 280 (internal quotation marks omitted). Instead, the Court of Appeals considered whether the Commonwealth could provide, with fidelity to the equal protection principle, separate and unequal educational programs for men and women.

The Fourth Circuit acknowledged that "the VWIL degree from Mary Baldwin College lacks the historical benefit and prestige of a degree from VMI." 44 F.3d, at 1241. The Court of Appeals further observed that VMI is "an ongoing and successful institution with a long history," and there remains no "comparable single-gender women's institution." Ibid. Nevertheless, the appeals court declared the substantially different and significantly unequal VWIL program satisfactory. The court reached that result by revising the applicable standard of review. The Fourth Circuit displaced the standard developed in our precedent, see supra, at 532-534, and substituted a standard of its own invention.

We have earlier described the deferential review in which the Court of Appeals engaged, see supra, at 528-529, a brand of review inconsistent with the more exacting standard our precedent requires, see supra, at 532-534. Quoting in part from Mississippi Univ. for Women, the Court of Appeals candidly described its own analysis as one capable of checking a legislative purpose ranked as "pernicious," but generally according "deference to [the] legislative will." 44 F.3d, at 1235, 1236. Recognizing that it had extracted from our decisions a test yielding "little or no scrutiny of the effect of a classification directed at [single-gender education]," the Court of Appeals devised another test, a "substantive comparability" inquiry, id., at 1237, and proceeded to find that new test satisfied, id., at 1241.

The Fourth Circuit plainly erred in exposing Virginia's VWIL plan to a deferential analysis, for "all gender-based classifications today" warrant "heightened scrutiny." See J. E. B., 511 U.S. at 136. Valuable as VWIL may prove for students who seek the program offered, Virginia's remedy affords no cure at all for the opportunities and advantages withheld from women who want a VMI education and can make the grade. See supra, at 549-554. n20 In sum, Virginia's remedy does not match the constitutional violation; the Commonwealth has shown no "exceedingly persuasive justification" for withholding from women qualified for the experience premier training of the kind VMI affords.

VII

A generation ago, "the authorities controlling Virginia higher education," despite long established tradition, agreed "to innovate and favorably entertained the [then] relatively new idea that there must be no discrimination by sex in offering educational opportunity." Kirstein, 309 F. Supp., at 186. Commencing in 1970, Virginia opened to women "educational opportunities at the Charlottesville campus that [were] not afforded in other [state-operated] institutions." Id., at 187; see supra, at 538. A federal court approved the Commonwealth's innovation, emphasizing that the University of Virginia "offered courses of instruction . . . not available elsewhere." 309 F. Supp., at 187. The court further noted: "There exists at Charlottesville a 'prestige' factor [not paralleled in] other Virginia educational institutions." Ibid.

VMI, too, offers an educational opportunity no other Virginia institution provides, and the school's "prestige" -- associated with its success in developing "citizen-soldiers" -- is unequaled. Virginia has closed this facility to its daughters and, instead, has devised for them a "parallel program," with a faculty less impressively credentialed and less well paid, more limited course offerings, fewer opportunities for military training and for scientific specialization. Cf. Sweatt, 339 U.S. at 633. VMI, beyond question, "possesses to a far greater degree" than the VWIL program "those qualities which are incapable of objective measurement but which make for greatness in a . . . school," including "position and influence of the alumni, standing in the community, traditions and prestige." Id., at 634. Women seeking and fit for a VMI-quality education cannot be offered anything less, under the Commonwealth's obligation to afford them genuinely equal protection.

A prime part of the history of our Constitution, historian Richard Morris recounted, is the story of the extension of constitutional rights and protections to people once ignored or excluded. n21 VMI's story continued as our comprehension of "We the People" expanded. See supra, at 546, n. 16. There is no reason to believe that the admission of women capable of all the activities required of VMI cadets would destroy the Institute rather than enhance its capacity to serve the "more perfect Union."

* * *

For the reasons stated, the initial judgment of the Court of Appeals, 976 F.2d 890 (CA4 1992), is affirmed, the final judgment of the Court of Appeals, 44 F.3d 1229 (CA4 1995), is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

JUSTICE THOMAS took no part in the consideration or decision of this case.

CHIEF JUSTICE REHNQUIST, concurring in the judgment.

The Court holds first that Virginia violates the Equal Protection Clause by maintaining the Virginia Military Institute's (VMI's) all-male admissions policy, and second that establishing the Virginia Women's Institute for Leadership (VWIL) program does not remedy that violation. While I agree with these conclusions, I disagree with the Court's analysis and so I write separately.

I

Two decades ago in Craig v. Boren, 429 U.S. 190, 197, 50 L. Ed. 2d 397, 97 S. Ct. 451 (1976), we announced that "to withstand constitutional challenge, . . . classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives." We have adhered to that standard of scrutiny ever since. See Califano v. Goldfarb, 430 U.S. 199, 210-211, 51 L. Ed. 2d 270, 97 S. Ct. 1021 (1977); Califano v. Webster, 430 U.S. 313, 316-317, 51 L. Ed. 2d 360, 97 S. Ct. 1192 (1977); Orr v. Orr, 440 U.S. 268, 279, 59 L. Ed. 2d 306, 99 S. Ct. 1102 (1979); Caban v. Mohammed, 441 U.S. 380, 388, 60 L. Ed. 2d 297, 99 S. Ct. 1760 (1979); Davis v. Passman, 442 U.S. 228, 234-235, 235, n. 9, 60 L. Ed. 2d 846, 99 S. Ct. 2264 (1979); Personnel Administrator of Mass. v. Feeney, 442 U.S. 256, 273, 60 L. Ed. 2d 870, 99 S. Ct. 2282 (1979); Califano v. Westcott, 443 U.S. 76, 85, 61 L. Ed. 2d 382, 99 S. Ct. 2655 (1979); Wengler v. Druggists Mut. Ins. Co., 446 U.S. 142, 150, 64 L. Ed. 2d 107, 100 S. Ct. 1540 (1980); Kirchberg v. Feenstra, 450 U.S. 455, 459-460, 67 L. Ed. 2d 428, 101 S. Ct. 1195 (1981); Michael M. v. Superior Court, Sonoma Cty., 450 U.S. 464, 469, 67 L. Ed. 2d 437, 101 S. Ct. 1200 (1981); Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 724, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982); Heckler v. Mathews, 465 U.S. 728, 744, 79 L. Ed. 2d 646, 104 S. Ct. 1387 (1984); J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 137, n. 6, 128 L. Ed. 2d 89, 114 S. Ct. 1419 (1994). While the majority adheres to this test today, ante, at 524, 533, it also says that the Commonwealth must demonstrate an "'exceedingly persuasive justification'" to support a gender-based classification. See ante, at 524, 529, 530, 531, 533, 534, 545, 546, 556. It is unfortunate that the Court thereby introduces an element of uncertainty respecting the appropriate test.

While terms like "important governmental objective" and "substantially related" are hardly models of precision, they have more content and specificity than does the phrase "exceedingly persuasive justification." That phrase is best confined, as it was first used, as an observation on the difficulty of meeting the applicable test, not as a formulation of the test itself. See, e. g., Feeney, supra, at 273 ("These precedents dictate that any state law overtly or covertly designed to prefer males over females in public employment require an exceedingly persuasive justification"). To avoid introducing potential confusion, I would have adhered more closely to our traditional, "firmly established," Hogan, supra, at 723; Heckler, supra, at 744, standard that a gender-based classification "must bear a close and substantial relationship to important governmental objectives." Feeney, supra, at 273.

Our cases dealing with gender discrimination also require that the proffered purpose for the challenged law be the actual purpose. See ante, at 533, 535-536. It is on this ground that the Court rejects the first of two justifications Virginia offers for VMI's single-sex admissions policy, namely, the goal of diversity among its public educational institutions. While I ultimately agree that the Commonwealth has not carried the day with this justification, I disagree with the Court's method of analyzing the issue.

VMI was founded in 1839, and, as the Court notes, ante, at 536-537, admission was limited to men because under the then-prevailing view men, not women, were destined for higher education. However misguided this point of view may be by present-day standards, it surely was not unconstitutional in 1839. The adoption of the Fourteenth Amendment, with its Equal Protection Clause, was nearly 30 years in the future. The interpretation of the Equal Protection Clause to require heightened scrutiny for gender discrimination was yet another century away.

Long after the adoption of the Fourteenth Amendment, and well into this century, legal distinctions between men and women were thought to raise no question under the Equal Protection Clause. The Court refers to our decision in Goesaert v. Cleary, 335 U.S. 464, 93 L. Ed. 163, 69 S. Ct. 198 (1948). Likewise representing that now abandoned view was Hoyt v. Florida, 368 U.S. 57, 7 L. Ed. 2d 118, 82 S. Ct. 159 (1961), where the Court upheld a Florida system of jury selection in which men were automatically placed on jury lists, but women were placed there only if they expressed an affirmative desire to serve. The Court noted that despite advances in women's opportunities, the "woman is still regarded as the center of home and family life." Id., at 62.

Then, in 1971, we decided Reed v. Reed, 404 U.S. 71, 30 L. Ed. 2d 225, 92 S. Ct. 251, which the Court correctly refers to as a seminal case. But its facts have nothing to do with admissions to any sort of educational institution. An Idaho statute governing the administration of estates and probate preferred men to women if the other statutory qualifications were equal. The statute's purpose, according to the Idaho Supreme Court, was to avoid hearings to determine who was better qualified as between a man and a woman both applying for letters of administration. This Court held that such a rule violated the Fourteenth Amendment because "a mandatory preference to members of either sex over members of the other, merely to accomplish the elimination of hearings," was an "arbitrary legislative choice forbidden by the Equal Protection Clause." Id., at 76. The brief opinion in Reed made no mention of either Goesaert or Hoyt.

Even at the time of our decision in Reed v. Reed, therefore, Virginia and VMI were scarcely on notice that its holding would be extended across the constitutional board. They were entitled to believe that "one swallow doesn't make a summer" and await further developments. Those developments were 11 years in coming. In Mississippi Univ. for Women v. Hogan, supra, a case actually involving a single-sex admissions policy in higher education, the Court held that the exclusion of men from a nursing program violated the Equal Protection Clause. This holding did place Virginia on notice that VMI's men-only admissions policy was open to serious question.

The VMI Board of Visitors, in response, appointed a Mission Study Committee to examine "the legality and wisdom of VMI's single-sex policy in light of" Hogan. 766 F. Supp. 1407, 1427 (WD Va. 1991). But the committee ended up cryptically recommending against changing VMI's status as a single-sex college. After three years of study, the committee found "'no information'" that would warrant a change in VMI's status. Id., at 1429. Even the District Court, ultimately sympathetic to VMI's position, found that "the Report provided very little indication of how [its] conclusion was reached" and that "the one and one-half pages in the committee's final report devoted to analyzing the information it obtained primarily focuses on anticipated difficulties in attracting females to VMI." Ibid. The reasons given in the report for not changing the policy were the changes that admission of women to VMI would require, and the likely effect of those changes on the institution. That VMI would have to change is simply not helpful in addressing the constitutionality of the status after Hogan.

Before this Court, Virginia has sought to justify VMI's single-sex admissions policy primarily on the basis that diversity in education is desirable, and that while most of the public institutions of higher learning in the Commonwealth are coeducational, there should also be room for single-sex institutions. I agree with the Court that there is scant evidence in the record that this was the real reason that Virginia decided to maintain VMI as men only. * But, unlike the majority, I would consider only evidence that postdates our decision in Hogan, and would draw no negative inferences from the Commonwealth's actions before that time. I think that after Hogan, the Commonwealth was entitled to reconsider its policy with respect to VMI, and not to have earlier justifications, or lack thereof, held against it.

Even if diversity in educational opportunity were the Commonwealth's actual objective, the Commonwealth's position would still be problematic. The difficulty with its position is that the diversity benefited only one sex; there was single-sex public education available for men at VMI, but no corresponding single-sex public education available for women. When Hogan placed Virginia on notice that VMI's admissions policy possibly was unconstitutional, VMI could have dealt with the problem by admitting women; but its governing body felt strongly that the admission of women would have seriously harmed the institution's educational approach. Was there something else the Commonwealth could have done to avoid an equal protection violation? Since the Commonwealth did nothing, we do not have to definitively answer that question.

I do not think, however, that the Commonwealth's options were as limited as the majority may imply. The Court cites, without expressly approving it, a statement from the opinion of the dissenting judge in the Court of Appeals, to the effect that the Commonwealth could have "simultaneously opened single-gender undergraduate institutions having substantially comparable curricular and extra-curricular programs, funding, physical plant, administration and support services, and faculty and library resources." Ante, at 529-530 (internal quotation marks omitted). If this statement is thought to exclude other possibilities, it is too stringent a requirement. VMI had been in operation for over a century and a half, and had an established, successful, and devoted group of alumni. No legislative wand could instantly call into existence a similar institution for women; and it would be a tremendous loss to scrap VMI's history and tradition. In the words of Grover Cleveland's second inaugural address, the Commonwealth faced a condition, not a theory. And it was a condition that had been brought about, not through defiance of decisions construing gender bias under the Equal Protection Clause, but, until the decision in Hogan, a condition that had not appeared to offend the Constitution. Had Virginia made a genuine effort to devote comparable public resources to a facility for women, and followed through on such a plan, it might well have avoided an equal protection violation. I do not believe the Commonwealth was faced with the stark choice of either admitting women to VMI, on the one hand, or abandoning VMI and starting from scratch for both men and women, on the other.

But, as I have noted, neither the governing board of VMI nor the Commonwealth took any action after 1982. If diversity in the form of single-sex, as well as coeducational, institutions of higher learning were to be available to Virginians, that diversity had to be available to women as well as to men.

The dissent criticizes me for "disregarding the four all-women's private colleges in Virginia (generously assisted by public funds)." Post, at 595. The private women's colleges are treated by the Commonwealth exactly as all other private schools are treated, which includes the provision of tuition-assistance grants to Virginia residents. Virginia gives no special support to the women's single-sex education. But obviously, the same is not true for men's education. Had the Commonwealth provided the kind of support for the private women's schools that it provides for VMI, this may have been a very different case. For in so doing, the Commonwealth would have demonstrated that its interest in providing a single-sex education for men was to some measure matched by an interest in providing the same opportunity for women.

Virginia offers a second justification for the single-sex admissions policy: maintenance of the adversative method. I agree with the Court that this justification does not serve an important governmental objective. A State does not have substantial interest in the adversative methodology unless it is pedagogically beneficial. While considerable evidence shows that a single-sex education is pedagogically beneficial for some students, see 766 F. Supp., at 1414, and hence a State may have a valid interest in promoting that methodology, there is no similar evidence in the record that an adversative method is pedagogically beneficial or is any more likely to produce character traits than other methodologies.

II

The Court defines the constitutional violation in these cases as "the categorical exclusion of women from an extraordinary educational opportunity afforded to men." Ante, at 547. By defining the violation in this way, and by emphasizing that a remedy for a constitutional violation must place the victims of discrimination in "'the position they would have occupied in the absence of [discrimination],'" ibid., the Court necessarily implies that the only adequate remedy would be the admission of women to the all-male institution. As the foregoing discussion suggests, I would not define the violation in this way; it is not the "exclusion of women" that violates the Equal Protection Clause, but the maintenance of an all-men school without providing any -- much less a comparable -- institution for women.

Accordingly, the remedy should not necessarily require either the admission of women to VMI or the creation of a VMI clone for women. An adequate remedy in my opinion might be a demonstration by Virginia that its interest in educating men in a single-sex environment is matched by its interest in educating women in a single-sex institution. To demonstrate such, the Commonwealth does not need to create two institutions with the same number of faculty Ph. D.'s, similar SAT scores, or comparable athletic fields. See ante, at 551-552. Nor would it necessarily require that the women's institution offer the same curriculum as the men's; one could be strong in computer science, the other could be strong in liberal arts. It would be a sufficient remedy, I think, if the two institutions offered the same quality of education and were of the same overall caliber.

If a State decides to create single-sex programs, the State would, I expect, consider the public's interest and demand in designing curricula. And rightfully so. But the State should avoid assuming demand based on stereotypes; it must not assume a priori, without evidence, that there would be no interest in a women's school of civil engineering, or in a men's school of nursing.

In the end, the women's institution Virginia proposes, VWIL, fails as a remedy, because it is distinctly inferior to the existing men's institution and will continue to be for the foreseeable future. VWIL simply is not, in any sense, the institution that VMI is. In particular, VWIL is a program appended to a private college, not a self-standing institution; and VWIL is substantially underfunded as compared to VMI. I therefore ultimately agree with the Court that Virginia has not provided an adequate remedy.

JUSTICE SCALIA, dissenting.

Today the Court shuts down an institution that has served the people of the Commonwealth of Virginia with pride and distinction for over a century and a half. To achieve that desired result, it rejects (contrary to our established practice) the factual findings of two courts below, sweeps aside the precedents of this Court, and ignores the history of our people. As to facts: It explicitly rejects the finding that there exist "gender-based developmental differences" supporting Virginia's restriction of the "adversative" method to only a men's institution, and the finding that the all-male composition of the Virginia Military Institute (VMI) is essential to that institution's character. As to precedent: It drastically revises our established standards for reviewing sex-based classifications. And as to history: It counts for nothing the long tradition, enduring down to the present, of men's military colleges supported by both States and the Federal Government.

Much of the Court's opinion is devoted to deprecating the closed-mindedness of our forebears with regard to women's education, and even with regard to the treatment of women in areas that have nothing to do with education. Closed minded they were -- as every age is, including our own, with regard to matters it cannot guess, because it simply does not consider them debatable. The virtue of a democratic system with a First Amendment is that it readily enables the people, over time, to be persuaded that what they took for granted is not so, and to change their laws accordingly. That system is destroyed if the smug assurances of each age are removed from the democratic process and written into the Constitution. So to counterbalance the Court's criticism of our ancestors, let me say a word in their praise: They left us free to change. The same cannot be said of this most illiberal Court, which has embarked on a course of inscribing one after another of the current preferences of the society (and in some cases only the counter majoritarian preferences of the society's law-trained elite) into our Basic Law. Today it enshrines the notion that no substantial educational value is to be served by an all-men's military academy -- so that the decision by the people of Virginia to maintain such an institution denies equal protection to women who cannot attend that institution but can attend others. Since it is entirely clear that the Constitution of the United States -- the old one -- takes no sides in this educational debate, I dissent.

I

I shall devote most of my analysis to evaluating the Court's opinion on the basis of our current equal protection jurisprudence, which regards this Court as free to evaluate everything under the sun by applying one of three tests: "rational basis" scrutiny, intermediate scrutiny, or strict scrutiny. These tests are no more scientific than their names suggest, and a further element of randomness is added by the fact that it is largely up to us which test will be applied in each case. Strict scrutiny, we have said, is reserved for state "classifications based on race or national origin and classifications affecting fundamental rights," Clark v. Jeter, 486 U.S. 456, 461, 100 L. Ed. 2d 465, 108 S. Ct. 1910 (1988) (citation omitted). It is my position that the term "fundamental rights" should be limited to "interest[s] traditionally protected by our society," Michael H. v. Gerald D., 491 U.S. 110, 122, 105 L. Ed. 2d 91, 109 S. Ct. 2333 (1989) (plurality opinion of SCALIA, J.); but the Court has not accepted that view, so that strict scrutiny will be applied to the deprivation of whatever sort of right we consider "fundamental." We have no established criterion for "intermediate scrutiny" either, but essentially apply it when it seems like a good idea to load the dice. So far it has been applied to content-neutral restrictions that place an incidental burden on speech, to disabilities attendant to illegitimacy, and to discrimination on the basis of sex. See, e. g., Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 662, 129 L. Ed. 2d 497, 114 S. Ct. 2445 (1994); Mills v. Habluetzel, 456 U.S. 91, 98-99, 71 L. Ed. 2d 770, 102 S. Ct. 1549 (1982); Craig v. Boren, 429 U.S. 190, 197, 50 L. Ed. 2d 397, 97 S. Ct. 451 (1976).

I have no problem with a system of abstract tests such as rational basis, intermediate, and strict scrutiny (though I think we can do better than applying strict scrutiny and intermediate scrutiny whenever we feel like it). Such formulas are essential to evaluating whether the new restrictions that a changing society constantly imposes upon private conduct comport with that "equal protection" our society has always accorded in the past. But in my view the function of this Court is to preserve our society's values regarding (among other things) equal protection, not to revise them; to prevent backsliding from the degree of restriction the Constitution imposed upon democratic government, not to prescribe, on our own authority, progressively higher degrees. For that reason it is my view that, whatever abstract tests we may choose to devise, they cannot supersede -- and indeed ought to be crafted so as to reflect -- those constant and unbroken national traditions that embody the people's understanding of ambiguous constitutional texts. More specifically, it is my view that "when a practice not expressly prohibited by the text of the Bill of Rights bears the endorsement of a long tradition of open, widespread, and unchallenged use that dates back to the beginning of the Republic, we have no proper basis for striking it down." Rutan v. Republican Party of Ill., 497 U.S. 62, 95, 111 L. Ed. 2d 52, 110 S. Ct. 2729 (1990) (SCALIA, J., dissenting). The same applies, mutatis mutandis, to a practice asserted to be in violation of the post-Civil War Fourteenth Amendment. See, e. g., Burnham v. Superior Court of Cal., County of Marin, 495 U.S. 604, 109 L. Ed. 2d 631, 110 S. Ct. 2105 (1990) (plurality opinion of SCALIA, J.) (Due Process Clause); J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 156-163, 128 L. Ed. 2d 89, 114 S. Ct. 1419 (1994) (SCALIA, J., dissenting) (Equal Protection Clause); Planned Parenthood of Southeastern Pa. v. Casey, 505 U.S. 833, 979-984, 1000-1001, 120 L. Ed. 2d 674, 112 S. Ct. 2791 (1992) (SCALIA, J., dissenting) (various alleged "penumbras").

The all-male constitution of VMI comes squarely within such a governing tradition. Founded by the Commonwealth of Virginia in 1839 and continuously maintained by it since, VMI has always admitted only men. And in that regard it has not been unusual. For almost all of VMI's more than a century and a half of existence, its single-sex status reflected the uniform practice for government-supported military colleges. Another famous Southern institution, The Citadel, has existed as a state-funded school of South Carolina since 1842. And all the federal military colleges -- West Point, the Naval Academy at Annapolis, and even the Air Force Academy, which was not established until 1954 -- admitted only males for most of their history. Their admission of women in 1976 (upon which the Court today relies, see ante, at 544-545, nn. 13, 15) came not by court decree, but because the people, through their elected representatives, decreed a change. See, e. g., § 803(a), 89 Stat. 537, note following 10 U.S.C. § 4342. In other words, the tradition of having government-funded military schools for men is as well rooted in the traditions of this country as the tradition of sending only men into military combat. The people may decide to change the one tradition, like the other, through democratic processes; but the assertion that either tradition has been unconstitutional through the centuries is not law, but politics-smuggled-into-law.

And the same applies, more broadly, to single-sex education in general, which, as I shall discuss, is threatened by today's decision with the cutoff of all state and federal support. Government-run nonmilitary educational institutions for the two sexes have until very recently also been part of our national tradition. "[It is] coeducation, historically, [that] is a novel educational theory. From grade school through high school, college, and graduate and professional training, much of the Nation's population during much of our history has been educated in sexually segregated classrooms." Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 736, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982) (Powell, J., dissenting); see id., at 736-739. These traditions may of course be changed by the democratic decisions of the people, as they largely have been.

Today, however, change is forced upon Virginia, and reversion to single-sex education is prohibited nationwide, not by democratic processes but by order of this Court. Even while bemoaning the sorry, bygone days of "fixed notions" concerning women's education, see ante, at 536-537, and n. 10, 537-539, 542-545, the Court favors current notions so fixedly that it is willing to write them into the Constitution of the United States by application of custom-built "tests." This is not the interpretation of a Constitution, but the creation of one.

II

To reject the Court's disposition today, however, it is not necessary to accept my view that the Court's made-up tests cannot displace longstanding national traditions as the primary determinant of what the Constitution means. It is only necessary to apply honestly the test the Court has been applying to sex-based classifications for the past two decades. It is well settled, as JUSTICE O'CONNOR stated some time ago for a unanimous Court, that we evaluate a statutory classification based on sex under a standard that lies "between the extremes of rational basis review and strict scrutiny." Clark v. Jeter, 486 U.S. at 461. We have denominated this standard "intermediate scrutiny" and under it have inquired whether the statutory classification is "substantially related to an important governmental objective." Ibid. See, e. g., Heckler v. Mathews, 465 U.S. 728, 744, 79 L. Ed. 2d 646, 104 S. Ct. 1387 (1984); Wengler v. Druggists Mut. Ins. Co., 446 U.S. 142, 150, 64 L. Ed. 2d 107, 100 S. Ct. 1540 (1980); Craig v. Boren, 429 U.S. at 197.

Before I proceed to apply this standard to VMI, I must comment upon the manner in which the Court avoids doing so. Notwithstanding our above-described precedents and their "'firmly established principles,'" Heckler, supra, at 744 (quoting Hogan, supra, at 723), the United States urged us to hold in this litigation "that strict scrutiny is the correct constitutional standard for evaluating classifications that deny opportunities to individuals based on their sex." Brief for United States in No. 94-2107, p. 16. (This was in flat contradiction of the Government's position below, which was, in its own words, to "state unequivocally that the appropriate standard in this case is 'intermediate scrutiny.'" 2 Record, Doc. No. 88, p. 3 (emphasis added).) The Court, while making no reference to the Government's argument, effectively accepts it.

Although the Court in two places recites the test as stated in Hogan, see ante, at 524, 532-533, which asks whether the State has demonstrated "that the classification serves important governmental objectives and that the discriminatory means employed are substantially related to the achievement of those objectives," 458 U.S. at 724 (internal quotation marks omitted), the Court never answers the question presented in anything resembling that form. When it engages in analysis, the Court instead prefers the phrase "exceedingly persuasive justification" from Hogan. The Court's nine invocations of that phrase, see ante, at 524, 529, 530, 531, 533, 534, 545, 546, 556, and even its fanciful description of that imponderable as "the core instruction" of the Court's decisions in J. E. B. v. Alabama ex rel. T. B., 511 U.S. 127, 128 L. Ed. 2d 89, 114 S. Ct. 1419 (1994), and Hogan, supra, see ante, at 531, would be unobjectionable if the Court acknowledged that whether a "justification" is "exceedingly persuasive" must be assessed by asking "[whether] the classification serves important governmental objectives and [whether] the discriminatory means employed are substantially related to the achievement of those objectives." Instead, however, the Court proceeds to interpret "exceedingly persuasive justification" in a fashion that contradicts the reasoning of Hogan and our other precedents.

That is essential to the Court's result, which can only be achieved by establishing that intermediate scrutiny is not survived if there are some women interested in attending VMI, capable of undertaking its activities, and able to meet its physical demands. Thus, the Court summarizes its holding as follows:

"In contrast to the generalizations about women on which Virginia rests, we note again these dispositive realities: VMI's implementing methodology is not inherently unsuitable to women; some women do well under the adversative model; some women, at least, would want to attend VMI if they had the opportunity; some women are capable of all of the individual activities required of VMI cadets and can meet the physical standards VMI now imposes on men." Ante, at 550 (internal quotation marks, citations, and punctuation omitted; emphasis added). Similarly, the Court states that "the Commonwealth's justification for excluding all women from 'citizen-soldier' training for which some are qualified . . . cannot rank as 'exceedingly persuasive' . . . ." Ante, at 545. n1

-----------------------------------
Footnotes

n1 Accord, ante, at 541 ("In sum . . ., neither the goal of producing citizen-soldiers, VMI's raison d'etre, nor VMI's implementing methodology is inherently unsuitable to women") (internal quotation marks omitted; emphasis added); ante, at 542 ("The question is whether the Commonwealth can constitutionally deny to women who have the will and capacity, the training and attendant opportunities that VMI uniquely affords"); ante, at 547-548 (the "violation" is that "equal protection [has been] denied to women ready, willing, and able to benefit from educational opportunities of the kind VMI offers"); ante, at 550 ("As earlier stated, see supra, at 541-542, generalizations about 'the way women are,' estimates of what is appropriate for most women, no longer justify denying opportunity to women whose talent and capacity place them outside the average description").

Only the amorphous "exceedingly persuasive justification" phrase, and not the standard elaboration of intermediate scrutiny, can be made to yield this conclusion that VMI's single-sex composition is unconstitutional because there exist several women (or, one would have to conclude under the Court's reasoning, a single woman) willing and able to undertake VMI's program. Intermediate scrutiny has never required a least-restrictive-means analysis, but only a "substantial relation" between the classification and the state interests that it serves. Thus, in Califano v. Webster, 430 U.S. 313, 51 L. Ed. 2d 360, 97 S. Ct. 1192 (1977) (per curiam), we upheld a congressional statute that provided higher Social Security benefits for women than for men. We reasoned that "women . . . as such have been unfairly hindered from earning as much as men," but we did not require proof that each woman so benefited had suffered discrimination or that each disadvantaged man had not; it was sufficient that even under the former congressional scheme "women on the average received lower retirement benefits than men." Id., at 318, and n. 5 (emphasis added). The reasoning in our other intermediate-scrutiny cases has similarly required only a substantial relation between end and means, not a perfect fit. In Rostker v. Goldberg, 453 U.S. 57, 69 L. Ed. 2d 478, 101 S. Ct. 2646 (1981), we held that selective-service registration could constitutionally exclude women, because even "assuming that a small number of women could be drafted for noncombat roles, Congress simply did not consider it worth the added burdens of including women in draft and registration plans." Id., at 81. In Metro Broadcasting, Inc. v. FCC, 497 U.S. 547, 579, 582-583, 111 L. Ed. 2d 445, 110 S. Ct. 2997 (1990), overruled on other grounds, Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227, 132 L. Ed. 2d 158, 115 S. Ct. 2097 (1995), we held that a classification need not be accurate "in every case" to survive intermediate scrutiny so long as, "in the aggregate," it advances the underlying objective. There is simply no support in our cases for the notion that a sex-based classification is invalid unless it relates to characteristics that hold true in every instance.

Not content to execute a de facto abandonment of the intermediate scrutiny that has been our standard for sex-based classifications for some two decades, the Court purports to reserve the question whether, even in principle, a higher standard (i. e., strict scrutiny) should apply. "The Court has," it says, "thus far reserved most stringent judicial scrutiny for classifications based on race or national origin . . .," ante, at 532, n. 6 (emphasis added); and it describes our earlier cases as having done no more than decline to "equate gender classifications, for all purposes, to classifications based on race or national origin," ante, at 532 (emphasis added). The wonderful thing about these statements is that they are not actually false -- just as it would not be actually false to say that "our cases have thus far reserved the 'beyond a reasonable doubt' standard of proof for criminal cases," or that "we have not equated tort actions, for all purposes, to criminal prosecutions." But the statements are misleading, insofar as they suggest that we have not already categorically held strict scrutiny to be inapplicable to sex-based classifications. See, e. g., Heckler v. Mathews, 465 U.S. 728, 79 L. Ed. 2d 646, 104 S. Ct. 1387 (1984) (upholding state action after applying only intermediate scrutiny); Michael M. v. Superior Court, Sonoma Cty., 450 U.S. 464, 67 L. Ed. 2d 437, 101 S. Ct. 1200 (1981) (same) (plurality and both concurring opinions); Califano v. Webster, supra (same) (per curiam). And the statements are irresponsible, insofar as they are calculated to destabilize current law. Our task is to clarify the law -- not to muddy the waters, and not to exact overcompliance by intimidation. The States and the Federal Government are entitled to know before they act the standard to which they will be held, rather than be compelled to guess about the outcome of Supreme Court peek-a-boo.

The Court's intimations are particularly out of place because it is perfectly clear that, if the question of the applicable standard of review for sex-based classifications were to be regarded as an appropriate subject for reconsideration, the stronger argument would be not for elevating the standard to strict scrutiny, but for reducing it to rational-basis review. The latter certainly has a firmer foundation in our past jurisprudence: Whereas no majority of the Court has ever applied strict scrutiny in a case involving sex-based classifications, we routinely applied rational-basis review until the 1970's, see, e. g., Hoyt v. Florida, 368 U.S. 57, 7 L. Ed. 2d 118, 82 S. Ct. 159 (1961); Goesaert v. Cleary, 335 U.S. 464, 93 L. Ed. 163, 69 S. Ct. 198 (1948). And of course normal, rational-basis review of sex-based classifications would be much more in accord with the genesis of heightened standards of judicial review, the famous footnote in United States v. Carolene Products Co., 304 U.S. 144, 82 L. Ed. 1234, 58 S. Ct. 778 (1938), which said (intimatingly) that we did not have to inquire in the case at hand"whether prejudice against discrete and insular minorities may be a special condition, which tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more searching judicial inquiry." Id., at 152-153, n. 4.It is hard to consider women a "discrete and insular minority" unable to employ the "political processes ordinarily to be relied upon," when they constitute a majority of the electorate. And the suggestion that they are incapable of exerting that political power smacks of the same paternalism that the Court so roundly condemns. See, e. g., ante, at 536-537, 542-545 (and accompanying notes). Moreover, a long list of legislation proves the proposition false. See, e. g., Equal Pay Act of 1963, 29 U.S.C. § 206(d); Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-2; Title IX of the Education Amendments of 1972, 20 U.S.C. § 1681; Women's Business Ownership Act of 1988, Pub. L. 100-533, 102 Stat. 2689; Violence Against Women Act of 1994, Pub. L. 103-322, Title IV, 108 Stat. 1902.

III

With this explanation of how the Court has succeeded in making its analysis seem orthodox -- and indeed, if intimations are to be believed, even overly generous to VMI -- I now proceed to describe how the analysis should have been conducted. The question to be answered, I repeat, is whether the exclusion of women from VMI is "substantially related to an important governmental objective."

A

It is beyond question that Virginia has an important state interest in providing effective college education for its citizens. That single-sex instruction is an approach substantially related to that interest should be evident enough from the long and continuing history in this country of men's and women's colleges. But beyond that, as the Court of Appeals here stated: "That single-gender education at the college level is beneficial to both sexes is a fact established in this case." 44 F.3d 1229, 1238 (CA4 1995) (emphasis added).

The evidence establishing that fact was overwhelming -- indeed, "virtually uncontradicted" in the words of the court that received the evidence, 766 F. Supp. 1407, 1415 (WD Va. 1991). As an initial matter, Virginia demonstrated at trial that "[a] substantial body of contemporary scholarship and research supports the proposition that, although males and females have significant areas of developmental overlap, they also have differing developmental needs that are deep-seated." Id., at 1434. While no one questioned that for many students a coeducational environment was nonetheless not inappropriate, that could not obscure the demonstrated benefits of single-sex colleges. For example, the District Court stated as follows:

"One empirical study in evidence, not questioned by any expert, demonstrates that single-sex colleges provide better educational experiences than coeducational institutions. Students of both sexes become more academically involved, interact with faculty frequently, show larger increases in intellectual self-esteem and are more satisfied with practically all aspects of college experience (the sole exception is social life) compared with their counterparts in coeducational institutions. Attendance at an all-male college substantially increases the likelihood that a student will carry out career plans in law, business and college teaching, and also has a substantial positive effect on starting salaries in business. Women's colleges increase the chances that those who attend will obtain positions of leadership, complete the baccalaureate degree, and aspire to higher degrees." Id., at 1412.See also id., at 1434-1435 (factual findings). "In the light of this very substantial authority favoring single-sex education," the District Court concluded that "the VMI Board's decision to maintain an all-male institution is fully justified even without taking into consideration the other unique features of VMI's teaching and training." Id., at 1412. This finding alone, which even this Court cannot dispute, see ante, at 535, should be sufficient to demonstrate the constitutionality of VMI's all-male composition.

But besides its single-sex constitution, VMI is different from other colleges in another way. It employs a "distinctive educational method," sometimes referred to as the "adversative, or doubting, model of education." 766 F. Supp., at 1413, 1421. "Physical rigor, mental stress, absolute equality of treatment, absence of privacy, minute regulation of behavior, and indoctrination in desirable values are the salient attributes of the VMI educational experience." Id., at 1421. No one contends that this method is appropriate for all individuals; education is not a "one size fits all" business. Just as a State may wish to support junior colleges, vocational institutes, or a law school that emphasizes case practice instead of classroom study, so too a State's decision to maintain within its system one school that provides the adversative method is "substantially related" to its goal of good education. Moreover, it was uncontested that "if the state were to establish a women's VMI-type [i. e., adversative] program, the program would attract an insufficient number of participants to make the program work," 44 F.3d, at 1241; and it was found by the District Court that if Virginia were to include women in VMI, the school "would eventually find it necessary to drop the adversative system altogether," 766 F. Supp., at 1413. Thus, Virginia's options were an adversative method that excludes women or no adversative method at all.

There can be no serious dispute that, as the District Court found, single-sex education and a distinctive educational method "represent legitimate contributions to diversity in the Virginia higher education system." Ibid. As a theoretical matter, Virginia's educational interest would have been best served (insofar as the two factors we have mentioned are concerned) by six different types of public colleges -- an all-men's, an all-women's, and a coeducational college run in the "adversative method," and an all-men's, an all-women's, and a coeducational college run in the "traditional method." But as a practical matter, of course, Virginia's financial resources, like any State's, are not limitless, and the Commonwealth must select among the available options. Virginia thus has decided to fund, in addition to some 14 coeducational 4-year colleges, one college that is run as an all-male school on the adversative model: the Virginia Military Institute.

Virginia did not make this determination regarding the make-up of its public college system on the unrealistic assumption that no other colleges exist. Substantial evidence in the District Court demonstrated that the Commonwealth has long proceeded on the principle that "'higher education resources should be viewed as a whole -- public and private'" -- because such an approach enhances diversity and because "'it is academic and economic waste to permit unwarranted duplication.'" Id., at 1420-1421 (quoting 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia). It is thus significant that, whereas there are "four all-female private [colleges] in Virginia," there is only "one private all-male college," which "indicates that the private sector is providing for the [former] form of education to a much greater extent that it provides for all-male education." 766 F. Supp., at 1420-1421. In these circumstances, Virginia's election to fund one public all-male institution and one on the adversative model -- and to concentrate its resources in a single entity that serves both these interests in diversity -- is substantially related to the Commonwealth's important educational interests.

B

The Court today has no adequate response to this clear demonstration of the conclusion produced by application of intermediate scrutiny. Rather, it relies on a series of contentions that are irrelevant or erroneous as a matter of law, foreclosed by the record in this litigation, or both.

1. I have already pointed out the Court's most fundamental error, which is its reasoning that VMI's all-male composition is unconstitutional because "some women are capable of all of the individual activities required of VMI cadets," 766 F. Supp., at 1412, and would prefer military training on the adversative model. See supra, at 571-574. This unacknowledged adoption of what amounts to (at least) strict scrutiny is without antecedent in our sex-discrimination cases and by itself discredits the Court's decision.

2. The Court suggests that Virginia's claimed purpose in maintaining VMI as an all-male institution -- its asserted interest in promoting diversity of educational options -- is not "genuine," but is a pretext for discriminating against women. Ante, at 539; see ante, at 535-540. To support this charge, the Court would have to impute that base motive to VMI's Mission Study Committee, which conducted a 3-year study from 1983 to 1986 and recommended to VMI's Board of Visitors that the school remain all male. The committee, a majority of whose members consisted of non-VMI graduates, "read materials on education and on women in the military," "made site visits to single-sex and newly coeducational institutions" including West Point and the Naval Academy, and "considered the reasons that other institutions had changed from single-sex to coeducational status"; its work was praised as "thorough" in the accreditation review of VMI conducted by the Southern Association of Colleges and Schools. See 766 F. Supp., at 1413, 1428; see also id., at 1427-1430 (detailed findings of fact concerning the Mission Study Committee). The Court states that "whatever internal purpose the Mission Study Committee served -- and however well meaning the framers of the report -- we can hardly extract from that effort any commonwealth policy evenhandedly to advance diverse educational options." Ante, at 539. But whether it is part of the evidence to prove that diversity was the Commonwealth's objective (its short report said nothing on that particular subject) is quite separate from whether it is part of the evidence to prove that antifeminism was not. The relevance of the Mission Study Committee is that its very creation, its sober 3-year study, and the analysis it produced utterly refute the claim that VMI has elected to maintain its all-male student-body composition for some misogynistic reason.

The Court also supports its analysis of Virginia's "actual state purposes" in maintaining VMI's student body as all male by stating that there is no explicit statement in the record "'in which the Commonwealth has expressed itself'" concerning those purposes. Ante, at 535, 539 (quoting 976 F.2d 890, 899 (CA4 1992)); see also ante, at 525. That is wrong on numerous grounds. First and foremost, in its implication that such an explicit statement of "actual purposes" is needed. The Court adopts, in effect, the argument of the United States that since the exclusion of women from VMI in 1839 was based on the "assumptions" of the time "that men alone were fit for military and leadership roles," and since "before this litigation was initiated, Virginia never sought to supply a valid, contemporary rationale for VMI's exclusionary policy," "that failure itself renders the VMI policy invalid." Brief for United States in No. 94-2107, at 10. This is an unheard-of doctrine. Each state decision to adopt or maintain a governmental policy need not be accompanied -- in anticipation of litigation and on pain of being found to lack a relevant state interest -- by a lawyer's contemporaneous recitation of the State's purposes. The Constitution is not some giant Administrative Procedure Act, which imposes upon the States the obligation to set forth a "statement of basis and purpose" for their sovereign Acts, see 5 U.S.C. § 553(c). The situation would be different if what the Court assumes to have been the 1839 policy had been enshrined and remained enshrined in legislation -- a VMI charter, perhaps, pronouncing that the institution's purpose is to keep women in their place. But since the 1839 policy was no more explicitly recorded than the Court contends the present one is, the mere fact that today's Commonwealth continues to fund VMI "is enough to answer [the United States'] contention that the [classification] was the 'accidental by-product of a traditional way of thinking about females.'" Michael M., 450 U.S. at 471, n. 6 (plurality opinion) (quoting Califano v. Webster, 430 U.S. at 320) (internal quotation marks omitted).

It is, moreover, not true that Virginia's contemporary reasons for maintaining VMI are not explicitly recorded. It is hard to imagine a more authoritative source on this subject than the 1990 Report of the Virginia Commission on the University of the 21st Century (1990 Report). As the parties stipulated, that report "notes that the hallmarks of Virginia's educational policy are 'diversity and autonomy.'" Stipulations of Fact 37, reprinted in Lodged Materials from the Record 64 (Lodged Materials). It said: "The formal system of higher education in Virginia includes a great array of institutions: state-supported and independent, two-year and senior, research and highly specialized, traditionally black and single-sex." 1990 Report, quoted in relevant part at Lodged Materials 64-65 (emphasis added). n2 The Court's only response to this is repeated reliance on the Court of Appeals' assertion that "'the only explicit [statement] that we have found in the record in which the Commonwealth has expressed itself with respect to gender distinctions'" (namely, the statement in the 1990 Report that the Commonwealth's institutions must "deal with faculty, staff, and students without regard to sex") had nothing to do with the purpose of diversity. Ante, at 525, 539 (quoting 976 F.2d, at 899). This proves, I suppose, that the Court of Appeals did not find a statement dealing with sex and diversity in the record; but the pertinent question (accepting the need for such a statement) is whether it was there. And the plain fact, which the Court does not deny, is that it was.

n2 This statement is supported by other evidence in the record demonstrating, by reference to both public and private institutions, that Virginia actively seeks to foster its "'rich heritage of pluralism and diversity in higher education,'" 1969 Report of the Virginia Commission on Constitutional Revision, quoted in relevant part at Lodged Materials 53; that Virginia views "'one special characteristic of the Virginia system [as being] its diversity,'" 1989 Virginia Plan for Higher Education, quoted in relevant part at Lodged Materials 64; and that in the Commonwealth's view "higher education resources should be viewed as a whole -- public and private" because "'Virginia needs the diversity inherent in a dual system of higher education,'" 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia, quoted in 766 F. Supp. 1407, 1420 (WD Va. 1991). See also Budget Initiatives for 1990-1992 of State Council of Higher Education for Virginia 10 (June 21, 1989) (Budget Initiatives), quoted at n. 3, infra. It should be noted (for this point will be crucial to my later discussion) that these official reports quoted here, in text and footnote, regard the Commonwealth's educational system -- public and private -- as a unitary one.

The Court contends that "[a] purpose genuinely to advance an array of educational options . . . is not served" by VMI. Ante, at 539-540. It relies on the fact that all of Virginia's other public colleges have become coeducational. Ibid.; see also ante, at 521, n. 2. The apparent theory of this argument is that unless Virginia pursues a great deal of diversity, its pursuit of some diversity must be a sham. This fails to take account of the fact that Virginia's resources cannot support all possible permutations of schools, see supra, at 578, and of the fact that Virginia coordinates its public educational offerings with the offerings of in-state private educational institutions that the Commonwealth provides money for its residents to attend and otherwise assists -- which include four women's colleges. n3

n3 The Commonwealth provides tuition assistance, scholarship grants, guaranteed loans, and work-study funds for residents of Virginia who attend private colleges in the Commonwealth. See, e. g., Va. Code Ann. §§ 23-38.11 to 23-38.19 (1993 and Supp. 1995) (Tuition Assistance Grant Act); §§ 23-38.30 to 23-38.44:3 (Virginia Student Assistance Authorities); Va. Code Ann. §§ 23-38.45 to 23-38.53 (1993) (College Scholarship Assistance Act); §§ 23-38.53:1 to 23-38.53:3 (Virginia Scholars Program); §§ 23-38.70, 23-38.71 (Virginia Work-Study Program). These programs involve substantial expenditures: for example, Virginia appropriated $ 4,413,750 (not counting federal funds it also earmarked) for the College Scholarship Assistance Program for both 1996 and 1997, and for the Tuition Assistance Grant Program appropriated $ 21,568,000 for 1996 and $ 25,842,000 for 1997. See 1996 Va. Appropriations Act, ch. 912, pt. 1, § 160.

In addition, as the parties stipulated in the District Court, the Commonwealth provides other financial support and assistance to private institutions -- including single-sex colleges -- through low-cost building loans, state-funded services contracts, and other programs. See, e. g., Va. Code Ann. §§ 23-30.39 to 23.30.58 (1993) (Educational Facilities Authority Act). The State Council of Higher Education for Virginia, in a 1989 document not created for purposes of this litigation but introduced into evidence, has described these various programs as a "means by which the Commonwealth can provide funding to its independent institutions, thereby helping to maintain a diverse system of higher education." Budget Initiatives 10.

Finally, the Court unreasonably suggests that there is some pretext in Virginia's reliance upon decentralized decisionmaking to achieve diversity -- its granting of substantial autonomy to each institution with regard to student-body composition and other matters, see 766 F. Supp., at 1419. The Court adopts the suggestion of the Court of Appeals that it is not possible for "one institution with autonomy, but with no authority over any other state institution, [to] give effect to a state policy of diversity among institutions." Ante, at 539 (internal quotation marks omitted). If it were impossible for individual human beings (or groups of human beings) to act autonomously in effective pursuit of a common goal, the game of soccer would not exist. And where the goal is diversity in a free market for services, that tends to be achieved even by autonomous actors who act out of entirely selfish interests and make no effort to cooperate. Each Virginia institution, that is to say, has a natural incentive to make itself distinctive in order to attract a particular segment of student applicants. And of course none of the institutions is entirely autonomous; if and when the legislature decides that a particular school is not well serving the interest of diversity -- if it decides, for example, that a men's school is not much needed -- funding will cease. n4

n4 The Court, unfamiliar with the Commonwealth's policy of diverse and independent institutions, and in any event careless of state and local traditions, must be forgiven by Virginians for quoting a reference to "'the Charlottesville campus'" of the University of Virginia. See ante, at 538. The University of Virginia, an institution even older than VMI, though not as old as another of the Commonwealth's universities, the College of William and Mary, occupies the portion of Charlottesville known, not as the "campus," but as "the grounds." More importantly, even if it were a "campus," there would be no need to specify "the Charlottesville campus," as one might refer to the Bloomington or Indianapolis campus of Indiana University. Unlike university systems with which the Court is perhaps more familiar, such as those in New York (e. g., the State University of New York at Binghamton or Buffalo), Illinois (University of Illinois at Urbana-Champaign or at Chicago), and California (University of California, Los Angeles, or University of California, Berkeley), there is only one University of Virginia. It happens (because Thomas Jefferson lived near there) to be located at Charlottesville. To many Virginians it is known, simply, as "the University," which suffices to distinguish it from the Commonwealth's other institutions offering 4-year college instruction, which include Christopher Newport College, Clinch Valley College, the College of William and Mary, George Mason University, James Madison University, Longwood College, Mary Washington University, Norfolk State University, Old Dominion University, Radford University, Virginia Commonwealth University, Virginia Polytechnic Institute and State University, Virginia State University -- and, of course, VMI.

3. In addition to disparaging Virginia's claim that VMI's single-sex status serves a state interest in diversity, the Court finds fault with Virginia's failure to offer education based on the adversative training method to women. It dismisses the District Court's "'findings' on 'gender-based developmental differences'" on the ground that "these 'findings' restate the opinions of Virginia's expert witnesses, opinions about typically male or typically female 'tendencies.'" Ante, at 541 (quoting 766 F. Supp., at 1434-1435). How remarkable to criticize the District Court on the ground that its findings rest on the evidence (i. e., the testimony of Virginia's witnesses)! That is what findings are supposed to do. It is indefensible to tell the Commonwealth that "the burden of justification is demanding and it rests entirely on [you]," ante, at 533, and then to ignore the District Court's findings because they rest on the evidence put forward by the Commonwealth -- particularly when, as the District Court said, "the evidence in the case . . . is virtually uncontradicted," 766 F. Supp., at 1415 (emphasis added).

Ultimately, in fact, the Court does not deny the evidence supporting these findings. See ante, at 541-546. It instead makes evident that the parties to this litigation could have saved themselves a great deal of time, trouble, and expense by omitting a trial. The Court simply dispenses with the evidence submitted at trial -- it never says that a single finding of the District Court is clearly erroneous -- in favor of the Justices' own view of the world, which the Court proceeds to support with (1) references to observations of someone who is not a witness, nor even an educational expert, nor even a judge who reviewed the record or participated in the judgment below, but rather a judge who merely dissented from the Court of Appeals' decision not to rehear this litigation en banc, see ante, at 542, (2) citations of nonevidentiary materials such as amicus curiae briefs filed in this Court, see ante, at 544-545, nn. 13, 14, and (3) various historical anecdotes designed to demonstrate that Virginia's support for VMI as currently constituted reminds the Justices of the "bad old days," see ante, at 542-544.

It is not too much to say that this approach to the litigation has rendered the trial a sham. But treating the evidence as irrelevant is absolutely necessary for the Court to reach its conclusion. Not a single witness contested, for example, Virginia's "substantial body of 'exceedingly persuasive' evidence . . . that some students, both male and female, benefit from attending a single-sex college" and "[that] for those students, the opportunity to attend a single-sex college is a valuable one, likely to lead to better academic and professional achievement." 766 F. Supp., at 1411-1412. Even the United States' expert witness "called himself a 'believer in single-sex education,'" although it was his "personal, philosophical preference," not one "born of educational-benefit considerations," "that single-sex education should be provided only by the private sector." Id., at 1412.

4. The Court contends that Virginia, and the District Court, erred, and "misperceived our precedent," by "training their argument on 'means' rather than 'end,'" ante, at 545. The Court focuses on "VMI's mission," which is to produce individuals "imbued with love of learning, confident in the functions and attitudes of leadership, possessing a high sense of public service, advocates of the American democracy and free enterprise system, and ready . . . to defend their country in time of national peril." 766 F. Supp., at 1425 (quoting Mission Study Committee of the VMI Board of Visitors, Report, May 16, 1986). "Surely," the Court says, "that goal is great enough to accommodate women." Ante, at 545.

This is lawmaking by indirection. What the Court describes as "VMI's mission" is no less the mission of all Virginia colleges. Which of them would the Old Dominion continue to fund if they did not aim to create individuals "imbued with love of learning, etc.," right down to being ready "to defend their country in time of national peril"? It can be summed up as "learning, leadership, and patriotism." To be sure, those general educational values are described in a particularly martial fashion in VMI's mission statement, in accordance with the military, adversative, and all-male character of the institution. But imparting those values in that fashion -- i. e., in a military, adversative, all-male environment -- is the distinctive mission of VMI. And as I have discussed (and both courts below found), that mission is not "great enough to accommodate women."

The Court's analysis at least has the benefit of producing foreseeable results. Applied generally, it means that whenever a State's ultimate objective is "great enough to accommodate women" (as it always will be), then the State will be held to have violated the Equal Protection Clause if it restricts to men even one means by which it pursues that objective -- no matter how few women are interested in pursuing the objective by that means, no matter how much the single-sex program will have to be changed if both sexes are admitted, and no matter how beneficial that program has theretofore been to its participants.

5. The Court argues that VMI would not have to change very much if it were to admit women. See, e. g., ante, at 540-542. The principal response to that argument is that it is irrelevant: If VMI's single-sex status is substantially related to the government's important educational objectives, as I have demonstrated above and as the Court refuses to discuss, that concludes the inquiry. There should be no debate in the federal judiciary over "how much" VMI would be required to change if it admitted women and whether that would constitute "too much" change.

But if such a debate were relevant, the Court would certainly be on the losing side. The District Court found as follows: "The evidence establishes that key elements of the adversative VMI educational system, with its focus on barracks life, would be fundamentally altered, and the distinctive ends of the system would be thwarted, if VMI were forced to admit females and to make changes necessary to accommodate their needs and interests." 766 F. Supp., at 1411. Changes that the District Court's detailed analysis found would be required include new allowances for personal privacy in the barracks, such as locked doors and coverings on windows, which would detract from VMI's approach of regulating minute details of student behavior, "contradict the principle that everyone is constantly subject to scrutiny by everyone else," and impair VMI's "total egalitarian approach" under which every student must be "treated alike"; changes in the physical training program, which would reduce "the intensity and aggressiveness of the current program"; and various modifications in other respects of the adversative training program that permeates student life. See id., at 1412-1413, 1435-1443. As the Court of Appeals summarized it, "the record supports the district court's findings that at least these three aspects of VMI's program -- physical training, the absence of privacy, and the adversative approach -- would be materially affected by coeducation, leading to a substantial change in the egalitarian ethos that is a critical aspect of VMI's training." 976 F.2d, at 896-897.

In the face of these findings by two courts below, amply supported by the evidence, and resulting in the conclusion that VMI would be fundamentally altered if it admitted women, this Court simply pronounces that "the notion that admission of women would downgrade VMI's stature, destroy the adversative system and, with it, even the school, is a judgment hardly proved." Ante, at 542 (footnote omitted). The point about "downgrading VMI's stature" is a straw man; no one has made any such claim. The point about "destroying the adversative system" is simply false; the District Court not only stated that "evidence supports this theory," but specifically concluded that while "without a doubt" VMI could assimilate women, "it is equally without a doubt that VMI's present methods of training and education would have to be changed" by a "move away from its adversative new cadet system." 766 F. Supp., at 1413, and n. 8, 1440. And the point about "destroying the school," depending upon what that ambiguous phrase is intended to mean, is either false or else sets a standard much higher than VMI had to meet. It sufficed to establish, as the District Court stated, that VMI would be "significantly different" upon the admission of women, 766 F. Supp., at 1412, and "would eventually find it necessary to drop the adversative system altogether," id., at 1413. n5

n5 The Court's do-it-yourself approach to factfinding, which throughout is contrary to our well-settled rule that we will not "undertake to review concurrent findings of fact by two courts below in the absence of a very obvious and exceptional showing of error," Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U.S. 271, 275, 93 L. Ed. 672, 69 S. Ct. 535 (1949) (and cases cited), is exemplified by its invocation of the experience of the federal military academies to prove that not much change would occur. See ante, at 542, n. 11; 544-545, and n. 15; 550-551, n. 19. In fact, the District Court noted that "the West Point experience" supported the theory that a coeducational VMI would have to "adopt a [different] system," for West Point found it necessary upon becoming coeducational to "move away" from its adversative system. 766 F. Supp., at 1413, 1440. "Without a doubt . . . VMI's present methods of training and education would have to be changed as West Point's were." Id., at 1413, n. 8; accord, 976 F.2d 890, 896-897 (CA4 1992) (upholding District Court's findings that "the unique characteristics of VMI's program," including its "unique methodology," "would be destroyed by coeducation").

6. Finally, the absence of a precise "all-women's analogue" to VMI is irrelevant. In Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982), we attached no constitutional significance to the absence of an all-male nursing school. As Virginia notes, if a program restricted to one sex is necessarily unconstitutional unless there is a parallel program restricted to the other sex, "the opinion in Hogan could have ended with its first footnote, which observed that 'Mississippi maintains no other single-sex public university or college.'" Brief for Cross-Petitioners in No. 94-2107, at 38 (quoting Mississippi Univ. for Women v. Hogan, supra, at 720, n. 1).

Although there is no precise female-only analogue to VMI, Virginia has created during this litigation the Virginia Women's Institute for Leadership (VWIL), a state-funded all-women's program run by Mary Baldwin College. I have thus far said nothing about VWIL because it is, under our established test, irrelevant, so long as VMI's all-male character is "substantially related" to an important state goal. But VWIL now exists, and the Court's treatment of it shows how far reaching today's decision is.

VWIL was carefully designed by professional educators who have long experience in educating young women. The program rejects the proposition that there is a "difference in the respective spheres and destinies of man and woman," Bradwell v. State, 83 U.S. 130, 16 Wall. 130, 141, 21 L. Ed. 442 (1873), and is designed to "provide an all-female program that will achieve substantially similar outcomes [to VMI's] in an all-female environment," 852 F. Supp. 471, 481 (WD Va. 1994). After holding a trial where voluminous evidence was submitted and making detailed findings of fact, the District Court concluded that "there is a legitimate pedagogical basis for the different means employed [by VMI and VWIL] to achieve the substantially similar ends." Ibid. The Court of Appeals undertook a detailed review of the record and affirmed. 44 F.3d 1229 (CA4 1995). n6 But it is Mary Baldwin College, which runs VWIL, that has made the point most succinctly:

"It would have been possible to develop the VWIL program to more closely resemble VMI, with adversative techniques associated with the rat line and barracks-like living quarters. Simply replicating an existing program would have required far less thought, research, and educational expertise. But such a facile approach would have produced a paper program with no real prospect of successful implementation." Brief for Mary Baldwin College as Amicus Curiae 5.It is worth noting that none of the United States' own experts in the remedial phase of this litigation was willing to testify that VMI's adversative method was an appropriate methodology for educating women. This Court, however, does not care. Even though VWIL was carefully designed by professional educators who have tremendous experience in the area, and survived the test of adversarial litigation, the Court simply declares, with no basis in the evidence, that these professionals acted on "'overbroad' generalizations," ante, at 542, 550.

n6 The Court is incorrect in suggesting that the Court of Appeals applied a "deferential" "brand of review inconsistent with the more exacting standard our precedent requires." Ante, at 555. That court "inquired (1) whether the state's objective is 'legitimate and important,' and (2) whether 'the requisite direct, substantial relationship between objective and means is present,'" 44 F.3d, at 1235 (quoting Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 725, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982). To be sure, such review is "deferential" to a degree that the Court's new standard is not, for it is intermediate scrutiny. (The Court cannot evade this point or prove the Court of Appeals too deferential by stating that that court "devised another test, a 'substantive comparability' inquiry,'" ante, at 555 (quoting 44 F.3d, at 1237), for as that court explained, its "substantive comparability" inquiry was an "additional step" that it engrafted on "the traditional test" of intermediate scrutiny, ibid. (emphasis added).)

C

A few words are appropriate in response to the concurrence, which finds VMI unconstitutional on a basis that is more moderate than the Court's but only at the expense of being even more implausible. The concurrence offers three reasons: First, that there is "scant evidence in the record," ante, at 562, that diversity of educational offering was the real reason for Virginia's maintaining VMI. "Scant" has the advantage of being an imprecise term. I have cited the clearest statements of diversity as a goal for higher education in the 1990 Report, the 1989 Virginia Plan for Higher Education, the Budget Initiatives prepared in 1989 by the State Council of Higher Education for Virginia, the 1974 Report of the General Assembly Commission on Higher Education to the General Assembly of Virginia, and the 1969 Report of the Virginia Commission on Constitutional Revision. See supra, at 579, 581-582, and n. 2, 583, n. 3. There is no evidence to the contrary, once one rejects (as the concurrence rightly does) the relevance of VMI's founding in days when attitudes toward the education of women were different. Is this conceivably not enough to foreclose rejecting as clearly erroneous the District Court's determination regarding "the Commonwealth's objective of educational diversity"? 766 F. Supp., at 1413. Especially since it is absurd on its face even to demand "evidence" to prove that the Commonwealth's reason for maintaining a men's military academy is that a men's military academy provides a distinctive type of educational experience (i. e., fosters diversity). What other purpose would the Commonwealth have? One may argue, as the Court does, that this type of diversity is designed only to indulge hostility toward women -- but that is a separate point, explicitly rejected by the concurrence, and amply refuted by the evidence I have mentioned in discussing the Court's opinion. n7 What is now under discussion -- the concurrence's making central to the disposition of this litigation the supposedly "scant" evidence that Virginia maintained VMI in order to offer a diverse educational experience -- is rather like making crucial to the lawfulness of the United States Army record "evidence" that its purpose is to do battle. A legal culture that has forgotten the concept of res ipsa loquitur deserves the fate that it today decrees for VMI.

n7 The concurrence states that it "read[s] the Court" not "as saying that the diversity rationale is a pretext" for discriminating against women, but as saying merely that the diversity rationale is not genuine. Ante, at 562, n. The Court itself makes no such disclaimer, which would be difficult to credit inasmuch as the foundation for its conclusion that the diversity rationale is not "genuine," ante, at 539, is its antecedent discussion of Virginia's "deliberate" actions over the past century and a half, based on "familiar arguments," that sought to enforce once "widely held views about women's proper place," ante, at 537, 538.

Second, the concurrence dismisses out of hand what it calls Virginia's "second justification for the single-sex admissions policy: maintenance of the adversative method." Ante, at 564. The concurrence reasons that "this justification does not serve an important governmental objective" because, whatever the record may show about the pedagogical benefits of single-sex education, "there is no similar evidence in the record that an adversative method is pedagogically beneficial or is any more likely to produce character traits than other methodologies." Ibid. That is simply wrong. See, e. g., 766 F. Supp., at 1426 (factual findings concerning character traits produced by VMI's adversative methodology); id., at 1434 (factual findings concerning benefits for many college-age men of an adversative approach in general). In reality, the pedagogical benefits of VMI's adversative approach were not only proved, but were a given in this litigation. The reason the woman applicant who prompted this suit wanted to enter VMI was assuredly not that she wanted to go to an all-male school; it would cease being all-male as soon as she entered. She wanted the distinctive adversative education that VMI provided, and the battle was joined (in the main) over whether VMI had a basis for excluding women from that approach. The Court's opinion recognizes this, and devotes much of its opinion to demonstrating that "'some women . . . do well under [the] adversative model'" and that "it is on behalf of these women that the United States has instituted this suit." Ante, at 550 (quoting 766 F. Supp., at 1434). Of course, in the last analysis it does not matter whether there are any benefits to the adversative method. The concurrence does not contest that there are benefits to single-sex education, and that alone suffices to make Virginia's case, since admission of a woman will even more surely put an end to VMI's single-sex education than it will to VMI's adversative methodology.

A third reason the concurrence offers in support of the judgment is that the Commonwealth and VMI were not quick enough to react to the "further developments" in this Court's evolving jurisprudence. Ante, at 561. Specifically, the concurrence believes it should have been clear after Hogan that "the difficulty with [Virginia's] position is that the diversity benefited only one sex; there was single-sex public education available for men at VMI, but no corresponding single-sex public education available for women." Ante, at 562. If only, the concurrence asserts, Virginia had "made a genuine effort to devote comparable public resources to a facility for women, and followed through on such a plan, it might well have avoided an equal protection violation." Ante, at 563. That is to say, the concurrence believes that after our decision in Hogan (which held a program of the Mississippi University for Women to be unconstitutional -- without any reliance on the fact that there was no corresponding Mississippi all-men's program), the Commonwealth should have known that what this Court expected of it was . . . yes!, the creation of a state all-women's program. Any lawyer who gave that advice to the Commonwealth ought to have been either disbarred or committed. (The proof of that pudding is today's 6-Justice majority opinion.) And any Virginia politician who proposed such a step when there were already four 4-year women's colleges in Virginia (assisted by state support that may well exceed, in the aggregate, what VMI costs, see n. 3, supra) ought to have been recalled.

In any event, "diversity in the form of single-sex, as well as coeducational, institutions of higher learning" is "available to women as well as to men" in Virginia. Ante, at 564. The concurrence is able to assert the contrary only by disregarding the four all-women's private colleges in Virginia (generously assisted by public funds) and the Commonwealth's longstanding policy of coordinating public with private educational offerings, see supra, at 579, 581-582, and n. 2, 583-584, and n. 3. According to the concurrence, the reason Virginia's assistance to its four all-women's private colleges does not count is that "the private women's colleges are treated by the State exactly as all other private schools are treated." Ante, at 564. But if Virginia cannot get credit for assisting women's education if it only treats women's private schools as it does all other private schools, then why should it get blame for assisting men's education if it only treats VMI as it does all other public schools? This is a great puzzlement.

IV

As is frequently true, the Court's decision today will have consequences that extend far beyond the parties to the litigation. What I take to be the Court's unease with these consequences, and its resulting unwillingness to acknowledge them, cannot alter the reality.

A

Under the constitutional principles announced and applied today, single-sex public education is unconstitutional. By going through the motions of applying a balancing test -- asking whether the State has adduced an "exceedingly persuasive justification" for its sex-based classification -- the Court creates the illusion that government officials in some future case will have a clear shot at justifying some sort of single-sex public education. Indeed, the Court seeks to create even a greater illusion than that: It purports to have said nothing of relevance to other public schools at all. "We address specifically and only an educational opportunity recognized . . . as 'unique.'" Ante, at 534, n. 7.

The Supreme Court of the United States does not sit to announce "unique" dispositions. Its principal function is to establish precedent -- that is, to set forth principles of law that every court in America must follow. As we said only this Term, we expect both ourselves and lower courts to adhere to the "rationale upon which the Court based the results of its earlier decisions." Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 66-67, 134 L. Ed. 2d 252, 116 S. Ct. 1114 (1996) (emphasis added). That is the principal reason we publish our opinions.

And the rationale of today's decision is sweeping: for sex-based classifications, a redefinition of intermediate scrutiny that makes it indistinguishable from strict scrutiny. See supra, at 571-574. Indeed, the Court indicates that if any program restricted to one sex is "unique," it must be opened to members of the opposite sex "who have the will and capacity" to participate in it. Ante, at 542. I suggest that the single-sex program that will not be capable of being characterized as "unique" is not only unique but nonexistent. n8

n8 In this regard, I note that the Court -- which I concede is under no obligation to do so -- provides no example of a program that would pass muster under its reasoning today: not even, for example, a football or wrestling program. On the Court's theory, any woman ready, willing, and physically able to participate in such a program would, as a constitutional matter, be entitled to do so.

In any event, regardless of whether the Court's rationale leaves some small amount of room for lawyers to argue, it ensures that single-sex public education is functionally dead. The costs of litigating the constitutionality of a single-sex education program, and the risks of ultimately losing that litigation, are simply too high to be embraced by public officials. Any person with standing to challenge any sex-based classification can haul the State into federal court and compel it to establish by evidence (presumably in the form of expert testimony) that there is an "exceedingly persuasive justification" for the classification. Should the courts happen to interpret that vacuous phrase as establishing a standard that is not utterly impossible of achievement, there is considerable risk that whether the standard has been met will not be determined on the basis of the record evidence -- indeed, that will necessarily be the approach of any court that seeks to walk the path the Court has trod today. No state official in his right mind will buy such a high-cost, high-risk lawsuit by commencing a single-sex program. The enemies of single-sex education have won; by persuading only seven Justices (five would have been enough) that their view of the world is enshrined in the Constitution, they have effectively imposed that view on all 50 States.

This is especially regrettable because, as the District Court here determined, educational experts in recent years have increasingly come to "support [the] view that substantial educational benefits flow from a single-gender environment, be it male or female, that cannot be replicated in a coeducational setting." 766 F. Supp., at 1415 (emphasis added). "The evidence in this case," for example, "is virtually uncontradicted" to that effect. Ibid. Until quite recently, some public officials have attempted to institute new single-sex programs, at least as experiments. In 1991, for example, the Detroit Board of Education announced a program to establish three boys-only schools for inner-city youth; it was met with a lawsuit, a preliminary injunction was swiftly entered by a District Court that purported to rely on Hogan, see Garrett v. Board of Ed. of School Dist. of Detroit, 775 F. Supp. 1004, 1006 (ED Mich. 1991), and the Detroit Board of Education voted to abandon the litigation and thus abandon the plan, see Detroit Plan to Aid Blacks with All-Boy Schools Abandoned, Los Angeles Times, Nov. 8, 1991, p. A4, col. 1. Today's opinion assures that no such experiment will be tried again.

B

There are few extant single-sex public educational programs. The potential of today's decision for widespread disruption of existing institutions lies in its application to private single-sex education. Government support is immensely important to private educational institutions. Mary Baldwin College -- which designed and runs VWIL -- notes that private institutions of higher education in the 1990-1991 school year derived approximately 19 percent of their budgets from federal, state, and local government funds, not including financial aid to students. See Brief for Mary Baldwin College as Amicus Curiae 22, n. 13 (citing U.S. Dept. of Education, National Center for Education Statistics, Digest of Education Statistics, p. 38 and Note (1993)). Charitable status under the tax laws is also highly significant for private educational institutions, and it is certainly not beyond the Court that rendered today's decision to hold that a donation to a single-sex college should be deemed contrary to public policy and therefore not deductible if the college discriminates on the basis of sex. See Note, The Independent Sector and the Tax Laws: Defining Charity in an Ideal Democracy, 64 S. Cal. L. Rev. 461, 476 (1991). See also Bob Jones Univ. v. United States, 461 U.S. 574, 76 L. Ed. 2d 157, 103 S. Ct. 2017 (1983).

The Court adverts to private single-sex education only briefly, and only to make the assertion (mentioned above) that "we address specifically and only an educational opportunity recognized by the District Court and the Court of Appeals as 'unique.'" Ante, at 534, n. 7. As I have already remarked, see supra, at 596, that assurance assures nothing, unless it is to be taken as a promise that in the future the Court will disclaim the reasoning it has used today to destroy VMI. The Government, in its briefs to this Court, at least purports to address the consequences of its attack on VMI for public support of private single-sex education. It contends that private colleges that are the direct or indirect beneficiaries of government funding are not thereby necessarily converted into state actors to which the Equal Protection Clause is then applicable. See Brief for United States in No. 94-2107, at 35-37 (discussing Rendell-Baker v. Kohn, 457 U.S. 830, 73 L. Ed. 2d 418, 102 S. Ct. 2764 (1982), and Blum v. Yaretsky, 457 U.S. 991, 73 L. Ed. 2d 534, 102 S. Ct. 2777 (1982)). That is true. It is also virtually meaningless.

The issue will be not whether government assistance turns private colleges into state actors, but whether the government itself would be violating the Constitution by providing state support to single-sex colleges. For example, in Norwood v. Harrison, 413 U.S. 455, 37 L. Ed. 2d 723, 93 S. Ct. 2804 (1973), we saw no room to distinguish between state operation of racially segregated schools and state support of privately run segregated schools. "Racial discrimination in state-operated schools is barred by the Constitution and 'it is also axiomatic that a state may not induce, encourage or promote private persons to accomplish what it is constitutionally forbidden to accomplish.'" Id., at 465 (quoting Lee v. Macon County Bd. of Ed., 267 F. Supp. 458, 475-476 (MD Ala. 1967)); see also Cooper v. Aaron, 358 U.S. 1, 19, 3 L. Ed. 2d 5, 78 S. Ct. 1401 (1958) ("State support of segregated schools through any arrangement, management, funds, or property cannot be squared with the [Fourteenth] Amendment's command that no State shall deny to any person within its jurisdiction the equal protection of the laws"); Grove City College v. Bell, 465 U.S. 555, 565, 79 L. Ed. 2d 516, 104 S. Ct. 1211 (1984) (case arising under Title IX of the Education Amendments of 1972 and stating that "the economic effect of direct and indirect assistance often is indistinguishable"). When the Government was pressed at oral argument concerning the implications of these cases for private single-sex education if government-provided single-sex education is unconstitutional, it stated that the implications will not be so disastrous, since States can provide funding to racially segregated private schools, "depending on the circumstances," Tr. of Oral Arg. 56. I cannot imagine what those "circumstances" might be, and it would be as foolish for private-school administrators to think that that assurance from the Justice Department will outlive the day it was made, as it was for VMI to think that the Justice Department's "unequivocal" support for an intermediate-scrutiny standard in this litigation would survive the Government's loss in the courts below.

The only hope for state-assisted single-sex private schools is that the Court will not apply in the future the principles of law it has applied today. That is a substantial hope, I am happy and ashamed to say. After all, did not the Court today abandon the principles of law it has applied in our earlier sex-classification cases? And does not the Court positively invite private colleges to rely upon our ad-hocery by assuring them this litigation is "unique"? I would not advise the foundation of any new single-sex college (especially an all-male one) with the expectation of being allowed to receive any government support; but it is too soon to abandon in despair those single-sex colleges already in existence. It will certainly be possible for this Court to write a future opinion that ignores the broad principles of law set forth today, and that characterizes as utterly dispositive the opinion's perceptions that VMI was a uniquely prestigious all-male institution, conceived in chauvinism, etc., etc. I will not join that opinion.

* * *

Justice Brandeis said it is "one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country." New State Ice Co. v. Liebmann, 285 U.S. 262, 311, 76 L. Ed. 747, 52 S. Ct. 371 (1932) (dissenting opinion). But it is one of the unhappy incidents of the federal system that a self-righteous Supreme Court, acting on its Members' personal view of what would make a "'more perfect Union,'" ante, at 558 (a criterion only slightly more restrictive than a "more perfect world"), can impose its own favored social and economic dispositions nationwide. As today's disposition, and others this single Term, show, this places it beyond the power of a "single courageous State," not only to introduce novel dispositions that the Court frowns upon, but to reintroduce, or indeed even adhere to, disfavored dispositions that are centuries old. See, e. g., BMW of North America, Inc. v. Gore, 517 U.S. 559, 134 L. Ed. 2d 809, 116 S. Ct. 1589 (1996); Romer v. Evans, 517 U.S. 620, 134 L. Ed. 2d 855, 116 S. Ct. 1620 (1996). The sphere of self-government reserved to the people of the Republic is progressively narrowed.

In the course of this dissent, I have referred approvingly to the opinion of my former colleague, Justice Powell, in Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982). Many of the points made in his dissent apply with equal force here -- in particular, the criticism of judicial opinions that purport to be "narrow" but whose "logic" is "sweeping." Id., at 745-746, n. 18. But there is one statement with which I cannot agree. Justice Powell observed that the Court's decision in Hogan, which struck down a single-sex program offered by the Mississippi University for Women, had thereby "left without honor . . . an element of diversity that has characterized much of American education and enriched much of American life." Id., at 735. Today's decision does not leave VMI without honor; no court opinion can do that.

In an odd sort of way, it is precisely VMI's attachment to such old-fashioned concepts as manly "honor" that has made it, and the system it represents, the target of those who today succeed in abolishing public single-sex education. The record contains a booklet that all first-year VMI students (the so-called "rats") were required to keep in their possession at all times. Near the end there appears the following period piece, entitled "The Code of a Gentleman":

"Without a strict observance of the fundamental Code of Honor, no man, no matter how 'polished,' can be considered a gentleman. The honor of a gentleman demands the inviolability of his word, and the incorruptibility of his principles. He is the descendant of the knight, the crusader; he is the defender of the defense-less and the champion of justice . . . or he is not a Gentleman.

"A Gentleman . . .

"Does not discuss his family affairs in public or with acquaintances.

"Does not speak more than casually about his girl friend.

"Does not go to a lady's house if he is affected by alcohol. He is temperate in the use of alcohol.

"Does not lose his temper; nor exhibit anger, fear, hate, embarrassment, ardor or hilarity in public.

"Does not hail a lady from a club window.

"A gentleman never discusses the merits or demerits of a lady.

"Does not mention names exactly as he avoids the mention of what things cost.

"Does not borrow money from a friend, except in dire need. Money borrowed is a debt of honor, and must be repaid as promptly as possible. Debts incurred by a deceased parent, brother, sister or grown child are assumed by honorable men as a debt of honor.

" Does not display his wealth, money or possessions.

"Does not put his manners on and off, whether in the club or in a ballroom. He treats people with courtesy, no matter what their social position may be.

"Does not slap strangers on the back nor so much as lay a finger on a lady.

"Does not 'lick the boots of those above' nor 'kick the face of those below him on the social ladder.'

"Does not take advantage of another's helplessness or ignorance and assumes that no gentleman will take advantage of him.

"A Gentleman respects the reserves of others, but demands that others respect those which are his.



"A Gentleman can become what he wills to be. . . ."I do not know whether the men of VMI lived by this code; perhaps not. But it is powerfully impressive that a public institution of higher education still in existence sought to have them do so. I do not think any of us, women included, will be better off for its destruction.

n1 During the Civil War, school teaching became a field dominated by women. See A. Scott, The Southern Lady: From Pedestal to Politics, 1830-1930, p. 82 (1970).

n2 Historically, most of Virginia's public colleges and universities were single sex; by the mid-1970's, however, all except VMI had become coeducational. 766 F. Supp. 1407, 1418-1419 (WD Va. 1991). For example, Virginia's legislature incorporated Farmville Female Seminary Association in 1839, the year VMI opened. 1839 Va. Acts, ch. 167. Originally providing instruction in "English, Latin, Greek, French, and piano" in a "home atmosphere," R. Sprague, Longwood College: A History 7-8, 15 (1989) (Longwood College), Farmville Female Seminary became a public institution in 1884 with a mission to train "white female teachers for public schools," 1884 Va. Acts, ch. 311. The school became Longwood College in 1949, Longwood College 136, and introduced coeducation in 1976, id., at 133.

n3 The District Court allowed the VMI Foundation and the VMI Alumni Association to intervene as defendants. 766 F. Supp., at 1408.

n4 Six judges voted to rehear the case en banc, four voted against rehearing, and three were recused. The Fourth Circuit's local Rule permits rehearing en banc only on the vote of a majority of the Circuit's judges in regular active service (currently 13) without regard to recusals. See 52 F.3d, at 91, and n. 1.

n5 As Thomas Jefferson stated the view prevailing when the Constitution was new: "Were our State a pure democracy . . . there would yet be excluded from their deliberations . . . women, who, to prevent depravation of morals and ambiguity of issue, could not mix promiscuously in the public meetings of men." Letter from Thomas Jefferson to Samuel Kercheval (Sept. 5, 1816), in 10 Writings of Thomas Jefferson 45-46, n. 1 (P. Ford ed. 1899).

n6 The Court has thus far reserved most stringent judicial scrutiny for classifications based on race or national origin, but last Term observed that strict scrutiny of such classifications is not inevitably "fatal in fact." Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 237, 132 L. Ed. 2d 158, 115 S. Ct. 2097 (1995) (internal quotation marks omitted).

n7 Several amici have urged that diversity in educational opportunities is an altogether appropriate governmental pursuit and that single-sex schools can contribute importantly to such diversity. Indeed, it is the mission of some single-sex schools "to dissipate, rather than perpetuate, traditional gender classifications." See Brief for Twenty-six Private Women's Colleges as Amici Curiae 5. We do not question the Commonwealth's prerogative evenhandedly to support diverse educational opportunities. We address specifically and only an educational opportunity recognized by the District Court and the Court of Appeals as "unique," see 766 F. Supp., at 1413, 1432; 976 F.2d, at 892, an opportunity available only at Virginia's premier military institute, the Commonwealth's sole single-sex public university or college. Cf. Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 720, n. 1, 73 L. Ed. 2d 1090, 102 S. Ct. 3331 (1982) ("Mississippi maintains no other single-sex public university or college. Thus, we are not faced with the question of whether States can provide 'separate but equal' undergraduate institutions for males and females.").

n8 On this point, the dissent sees fire where there is no flame. See post, at 596-598, 598-600. "Both men and women can benefit from a single-sex education," the District Court recognized, although "the beneficial effects" of such education, the court added, apparently "are stronger among women than among men." 766 F. Supp., at 1414. The United States does not challenge that recognition. Cf. C. Jencks & D. Riesman, The Academic Revolution 297-298 (1968):

"The pluralistic argument for preserving all-male colleges is uncomfortably similar to the pluralistic argument for preserving all-white colleges . . . . The all-male college would be relatively easy to defend if it emerged from a world in which women were established as fully equal to men. But it does not. It is therefore likely to be a witting or unwitting device for preserving tacit assumptions of male superiority -- assumptions for which women must eventually pay."

n9 Dr. Edward H. Clarke of Harvard Medical School, whose influential book, Sex in Education, went through 17 editions, was perhaps the most well-known speaker from the medical community opposing higher education for women. He maintained that the physiological effects of hard study and academic competition with boys would interfere with the development of girls' reproductive organs. See E. Clarke, Sex in Education 38-39, 62-63 (1873); id., at 127 ("identical education of the two sexes is a crime before God and humanity, that physiology protests against, and that experience weeps over"); see also H. Maudsley, Sex in Mind and in Education 17 (1874) ("It is not that girls have not ambition, nor that they fail generally to run the intellectual race [in coeducational settings], but it is asserted that they do it at a cost to their strength and health which entails life-long suffering, and even incapacitates them for the adequate performance of the natural functions of their sex."); C. Meigs, Females and Their Diseases 350 (1848) (after five or six weeks of "mental and educational discipline," a healthy woman would "lose . . . the habit of menstruation" and suffer numerous ills as a result of depriving her body for the sake of her mind).

n10 Virginia's Superintendent of Public Instruction dismissed the coeducational idea as "'repugnant to the prejudices of the people'" and proposed a female college similar in quality to Girton, Smith, or Vassar. 2 History of Women's Education 254 (quoting Dept. of Interior, 1 Report of Commissioner of Education, H. R. Doc. No. 5, 58th Cong., 2d Sess., 438 (1904)).

n11 See post, at 566, 598-599, 603. Forecasts of the same kind were made regarding admission of women to the federal military academies. See, e. g., Hearings on H. R. 9832 et al. before Subcommittee No. 2 of the House Committee on Armed Services, 93d Cong., 2d Sess., 137 (1975) (statement of Lt. Gen. A. P. Clark, Superintendent of U.S. Air Force Academy) ("It is my considered judgment that the introduction of female cadets will inevitably erode this vital atmosphere."); id., at 165 (statement of Hon. H. H. Callaway, Secretary of the Army) ("Admitting women to West Point would irrevocably change the Academy. . . . The Spartan atmosphere -- which is so important to producing the final product -- would surely be diluted, and would in all probability disappear.").

n12 See 766 F. Supp., at 1413 (describing testimony of expert witness David Riesman: "If VMI were to admit women, it would eventually find it necessary to drop the adversative system altogether, and adopt a system that provides more nurturing and support for the students."). Such judgments have attended, and impeded, women's progress toward full citizenship stature throughout our Nation's history. Speaking in 1879 in support of higher education for females, for example, Virginia State Senator C. T. Smith of Nelson recounted that legislation proposed to protect the property rights of women had encountered resistance. 10 Educ. J. Va. 213 (1879). A Senator opposing the measures objected that "there [was] no formal call for the [legislation]," and "depicted in burning eloquence the terrible consequences such laws would produce." Ibid. The legislation passed, and a year or so later, its sponsor, C. T. Smith, reported that "not one of [the forecast "terrible consequences"] has or ever will happen, even unto the sounding of Gabriel's trumpet." Ibid. See also supra, at 537-538.

n13 Women cadets have graduated at the top of their class at every federal military academy. See Brief for Lieutenant Colonel Rhonda Cornum et al. as Amici Curiae 11, n. 25; cf. Defense Advisory Committee on Women in the Services, Report on the Integration and Performance of Women at West Point 64 (1992).

n14 Brief for Lieutenant Colonel Rhonda Cornum, supra, at 5-9 (reporting the vital contributions and courageous performance of women in the military); see Mintz, President Nominates 1st Woman to Rank of Three-Star General, Washington Post, Mar. 27, 1996, p. A19, col. 1 (announcing President's nomination of Marine Corps Major General Carol Mutter to rank of Lieutenant General; Mutter will head corps manpower and planning); Tousignant, A New Era for the Old Guard, Washington Post, Mar. 23, 1996, p. C1, col. 2 (reporting admission of Sergeant Heather Johnsen to elite Infantry unit that keeps round-the-clock vigil at Tomb of the Unknowns in Arlington National Cemetery).

n15 Inclusion of women in settings where, traditionally, they were not wanted inevitably entails a period of adjustment. As one West Point cadet squad leader recounted: "The classes of '78 and '79 see the women as women, but the classes of '80 and '81 see them as classmates." U.S. Military Academy, A. Vitters, Report of Admission of Women (Project Athena II) 84 (1978) (internal quotation marks omitted).

n16 VMI has successfully managed another notable change. The school admitted its first African-American cadets in 1968. See The VMI Story 347-349 (students no longer sing "Dixie," salute the Confederate flag or the tomb of General Robert E. Lee at ceremonies and sports events). As the District Court noted, VMI established a program on "retention of black cadets" designed to offer academic and social-cultural support to "minority members of a dominantly white and tradition-oriented student body." 766 F. Supp., at 1436-1437. The school maintains a "special recruitment program for blacks" which, the District Court found, "has had little, if any, effect on VMI's method of accomplishing its mission." Id., at 1437.

n17 As earlier observed, see supra, at 529, Judge Phillips, in dissent, measured Virginia's plan against a paradigm arrangement, one that "could survive equal protection scrutiny": single-sex schools with "substantially comparable curricular and extra-curricular programs, funding, physical plant, administration and support services, . . . faculty[,] and library resources." 44 F.3d 1229, 1250 (CA4 1995). Cf. Bray v. Lee, 337 F. Supp. 934 (Mass. 1972) (holding inconsistent with the Equal Protection Clause admission of males to Boston's Boys Latin School with a test score of 120 or higher (up to a top score of 200) while requiring a score, on the same test, of at least 133 for admission of females to Girls Latin School, but not ordering coeducation). Measuring VMI/VWIL against the paradigm, Judge Phillips said, "reveals how far short the [Virginia] plan falls from providing substantially equal tangible and intangible educational benefits to men and women." 44 F.3d, at 1250.

n18 Both programs include an honor system. Students at VMI are expelled forthwith for honor code violations, see 766 F. Supp., at 1423; the system for VWIL students, see 852 F. Supp., at 496-497, is less severe, see Tr. 414-415 (testimony of Mary Baldwin College President Cynthia Tyson).

n19 Admitting women to VMI would undoubtedly require alterations necessary to afford members of each sex privacy from the other sex in living arrangements, and to adjust aspects of the physical training programs. See Brief for Petitioner 27-29; cf. note following 10 U.S.C. § 4342 (academic and other standards for women admitted to the Military, Naval, and Air Force Academies "shall be the same as those required for male individuals, except for those minimum essential adjustments in such standards required because of physiological differences between male and female individuals"). Experience shows such adjustments are manageable. See U.S. Military Academy, A. Vitters, N. Kinzer, & J. Adams, Report of Admission of Women (Project Athena I-IV) (1977-1980) (4-year longitudinal study of the admission of women to West Point); Defense Advisory Committee on Women in the Services, Report on the Integration and Performance of Women at West Point 17-18 (1992).

n20 Virginia's prime concern, it appears, is that "placing men and women into the adversative relationship inherent in the VMI program . . . would destroy, at least for that period of the adversative training, any sense of decency that still permeates the relationship between the sexes." 44 F.3d, at 1239; see supra, at 540-546. It is an ancient and familiar fear. Compare In re Lavinia Goodell, 39 Wis. 232, 246 (1875) (denying female applicant's motion for admission to the bar of its court, Wisconsin Supreme Court explained: "Discussions are habitually necessary in courts of justice, which are unfit for female ears. The habitual presence of women at these would tend to relax the public sense of decency and propriety."), with Levine, Closing Comments, 6 Law & Inequality 41 (1988) (presentation at Eighth Circuit Judicial Conference, Colorado Springs, Colo., July 17, 1987) (footnotes omitted): "Plato questioned whether women should be afforded equal opportunity to become guardians, those elite Rulers of Platonic society. Ironically, in that most undemocratic system of government, the Republic, women's native ability to serve as guardians was not seriously questioned. The concern was over the wrestling and exercise class in which all candidates for guardianship had to participate, for rigorous physical and mental training were prerequisites to attain the exalted status of guardian. And in accord with Greek custom, those exercise classes were conducted in the nude. Plato concluded that their virtue would clothe the women's nakedness and that Platonic society would not thereby be deprived of the talent of qualified citizens for reasons of mere gender." For Plato's full text on the equality of women, see 2 The Dialogues of Plato 302-312 (B. Jowett transl., 4th ed. 1953). Virginia, not bound to ancient Greek custom in its "rigorous physical and mental training" programs, could more readily make the accommodations necessary to draw on "the talent of [all] qualified citizens." Cf. supra, at 550-551, n. 19.

- - - - - - - - - - - - End Footnotes- - - - - - - - - - - - - - - - - - - - - - - - - - - - Footnotes - - - - - - - - - - - - - - -

n21 R. Morris, The Forging of the Union, 1781-1789, p. 193 (1987); see id., at 191, setting out letter to a friend from Massachusetts patriot (later second President) John Adams, on the subject of qualifications for voting in his home State: "It is dangerous to open so fruitful a source of controversy and altercation as would be opened by attempting to alter the qualifications of voters; there will be no end of it. New claims will arise; women will demand a vote; lads from twelve to twenty-one will think their rights not enough attended to; and every man who has not a farthing, will demand an equal voice with any other, in all acts of state. It tends to confound and destroy all distinctions, and prostrate all ranks to one common level." Letter from John Adams to James Sullivan (May 26, 1776), in 9 Works of John Adams 378 (C. Adams ed. 1854).

n22 The dissent equates our conclusion that VMI's "asserted interest in promoting diversity" is not "'genuine,'" with a "charge" that the diversity rationale is "a pretext for discriminating against women." Post, at 579-580. Of course, those are not the same thing. I do not read the Court as saying that the diversity rationale is a pretext for discrimination, and I would not endorse such a proposition. We may find that diversity was not the Commonwealth's real reason without suggesting, or having to show, that the real reason was "antifeminism," post, at 580. Our cases simply require that the proffered purpose for the challenged gender classification be the actual purpose, although not necessarily recorded. See ante, at 533, 535-536. The dissent also says that the interest in diversity is so transparent that having to articulate it is "absurd on its face." Post, at 592. Apparently, that rationale was not obvious to the Mission Study Committee which failed to list it among its reasons for maintaining VMI's all-men admissions policy.