dissenting.
In Brown v. Board of Education, 347 U. S. 483 (1954), this Court held that segregation of children in public schools on the basis of race deprives minority group children of equal educational opportunities and therefore denies them the equal protection of the laws under the *782Fourteenth 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 (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 *783of 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.
I
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.”
*784Nowhere 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 mischaracteri-zations 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 a 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 *785note 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 (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., at 201. 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, *786which under our decision in Keyes, forms “a predicate for a finding of the existence of a dual school system,” ibid., and justifies “all-out desegregation.” Id., at 214.
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 *787been 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 identi-fiability 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 *788primary 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 *789identifiability 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 *790within 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.
II
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.
A
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, *791to 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 (1971).
*792Also 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 (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 (1880), actions by an agent or officer of *793the 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 (1880); Shelley v. Kraemer, 334 U. S. 1, 14 (1948).
Under Michigan law a “school district is an agency of the State government.” School District of the City 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 (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. 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 *794Michigan, 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 (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*795ipality .... 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 *796standards 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 (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*797trict 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 flies 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, at 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.” Elementary and Secondary Education and the 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 *798viewed 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.
B
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 (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 (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 (1972). See also Swann v. Charlotte-Mecklenburg Board of Education, 402 U. S. 1, 15 (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.
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 *799unconstitutional 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 *800regular-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 *801headway 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*802portion of whites in county schools may drop as those who can register in private academies'...407 U. S., at 464. 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. See also Green, 391 U. S., at 438 n. 4; Monroe v. Board of Comm’rs, 391 U. S. 450, 459 (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. This is the operative standard re-emphasized in Davis v. School Comm’rs of Mobile County, 402 U. S. 33, 37 (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 *803U. S., at 26. 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, at 18. The evil to be remedied in the dismantling of a dual system is the “[r] acial identification of the system’s schools.” Green, 391 U. S., at 435. The goal is a system without white schools or Negro schools — a system with “just schools.” Id., at 442. 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, supra, at 25; Davis, supra, at 37; Green, supra, at 439.
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. 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. 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. *804What 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. 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, ante, at 670.
Under a Detroit-only decree, Detroit’s schools will clearly remain racially identifiable in comparison with neighboring schools in the metropolitan community. Schools with 66% 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*805dren 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. 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 *806showing a district court may consider this in fashioning a remedy.” 402 U. S., at 20-21.
See also Keyes, 413 U. S., at 202. 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.
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, at 16. 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 *807cases, as in other equitable causes, a remedy which effectively cures the violation is what is required. See Green, 391 U. S., at 439; Davis, 402 U. S., at 37. 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 mal-apportioned in violation of the Fourteenth Amendment. See Reynolds v. Sims, 377 U. S. 533 (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 *808and 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, supra, at 575. See also Gomillion v. Lightfoot, 364 U. S. 339 (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 (DDC 1967). We must do better than “ ‘substitute . . . one segregated school system for another segregated school system.’ ” Wright] 407 U. S., at 456. To suggest, as does the majority, that a Detroit-only plan somehow remedies the effects of de jure segregation of the races is, in my view, to make a solemn mockery of Brown 7’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.
Ill
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 *809adjusting and reconciling public and private needs.” Brown II, 349 U. S., at 300. 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; Green, 391 U. S., at 439. 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. *810Rather 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 to 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 *811Were 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 questions 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 *812easily 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 interdistrict 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 (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. 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 ride buses. 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*813proximately 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% 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, supra, at 30.
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 supra, 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 *814probably 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.
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 *815the 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.
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 (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. 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.
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.
See Mich. Comp. Laws § 388.851 (1970).
See § 388.1179.
See §§ 388.629 and 340.600.
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.
See, e. g., id., § 340.575. See also 1949-1950 Report of the Attorney General 104 (Roth); Yol. 1, 1955 Report of the Attorney General 561 (Kavanagh); 1961-1962 Report of the Attorney General 533 (Kelley).
See Mich. Comp. Laws §§ 211.34 and 340.681.
§ 340.569.
§§ 257.811 (c), 340.361, 340.781, 340.782, 388.371.
§ 340.575.
§ 388.1171.
§340.887 (1).
Op. Atty. Gen. No. 4705 (July 7, 1970), 1969-1970 Report of the Attorney General 156 (Kelley).
See Mich. Comp. Laws § 340.253.
See generally §§340.401-340.415 (consolidations), 340.431-340.449 (annexations).
See 1 Michigan Senate Journal, 1968, p. 423.
See generally Mich. Comp. Laws §§ 340.461-340.468
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 78-L-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 *800we 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.
In fact, the District Court remarked “that this court's task is to enforce constitutional rights not to act as a schoolmaster; the *811court’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.
See, e. g., Mich. Comp. Laws §§340.69, 340.121 (d), 340.359, 340.582, 340.582a, 340.590.
See id., §§ 340.330-340.330u.