Tallulah Morgan v. John J. Kerrigan, Boston Home and School Association, Defendant-Intervenor-Appellant, Kevin H. White, Etc.

COFFIN, Chief Judge. These appeals present varied challenges to orders of the district court implementing a plan of desegregation for the public schools of Boston. The consolidated cases concern the remedy phase of litigation initiated by plaintiffs-appellees, representing a class of all black public school students and their parents, against, principally, the Boston School Committee and the Superintendent of Boston Public Schools. The liability phase came to an end in 1974 with a district court finding of substantial segregation in the entire school system intentionally brought about and maintained by official action over the years. Morgan v. Hennigan, 379 F.Supp. 410 (D.Mass.1974). 1 We affirmed, Morgan v. Kerrigan, 509 F.2d 580 (1st Cir. 1974), and the Supreme Court denied certiorari, 421 U.S. 963, 95 S.Ct. 1950, 44 L.Ed.2d 449 (1975). *406 While the liability issues were- being considered on appeal, the district court, after its decision on June 21, 1974, began its exploration of appropriate remedies. The period from June, 1974, to May, 1975, was occupied with the addition of parties to the litigation, 2 hearings as to the nature, scope and objectives of a plan, submission and criticism of various plans, consideration of all proposals and preparation of a plan by a panel of masters, and, finally, the issuance of a revised plan by the district court on May 10, 1975, followed by a Memorandum Decision and Remedial Order. Morgan v. Kerrigan, 401 F.Supp. 216 (D.Mass.1975). On June 17, 1975, we denied appellants’ motion for a stay pending appeal, Morgan v. Kerrigan, 1 Cir., 523 F.2d 917 (1975), but devised a consolidated and expedited briefing schedule, in which all parties have faithfully cooperated. During the summer of 1975 the court, counsel, school officials, teachers, parent organizations, and federal, state, and city agencies and officials worked intensively to prepare for the September school opening in compliance with the district court’s order, commonly referred to as Phase II. 3 The schools are now functioning in accordance with the court’s plan and orders. The issues to be considered here are both procedural and substantive and require some further background of events. After hearings during the summer and early fall of 1974, the district court, on October 31, 1974, entered an order establishing guidelines 4 and date (December 16) for a plan of desegregation to be filed by the School Committee. Such a plan was prepared by the staff but, on the deadline date, the School Committee voted not to submit it. 5 It was, however, filed by the Committee’s counsel. This plan called for six districts, with varying learning approaches available within each district, and “magnet” or special purpose high schools, but left with parents the choice of schools for their children. On January 20, 1975, plaintiffs submitted a plan, accepting the six districts identified in the staff plan of December 16, 1974, but proposing a mandatory allocation of students among the schools to achieve minority enrollments that were neither too small (e. g., a minimum of 29.4 percent in elementary schools) nor too large (e. g., a maximum of 60.6 percent in elementary schools). Also on January 20, 1975, the Boston Home and School Association (Association) filed a plan which was based on the theory that segregation in certain schools was the result of “existing residential separateness” and a policy of neighborhood school assignments, rather than of any official actions of the School Committee. To support its approach, the Association offered evidence of demographic patterns, which the district court *407 refused to accept on the grounds that the evidence was irrelevant at the remedy stage of the case and that the issue raised by the offer had been litigated and finally decided in the liability phase of these proceedings. The court’s refusal to consider the Association’s plan is not in issue, but the Association’s contention that the court should reopen the proceedings to consider, for purposes of tailoring remedies, the impact of demographic conditions on particular schools is one of the principal issues before us. The School Committee’s plan, finally submitted on January 27, 1975, see note 5 supra, also kept the six districts or zones, and allowed parents several options, ranging from electing to have their child remain in a racially mixed school, to choosing a citywide or zoned magnet school, to any school within the zone. Should the school chosen by the parents be dominantly black or white, the desegregative remedy would be a once-a-week (for elementary schools) or a once-every-two-week (for middle level schools) visit by paired black and white schools to a “third site” resource center for training and experience in race relations. With these three plans on the table, the court appointed two experts to assist in evaluating plans and a panel of four masters to consider the plans — commencing with the School Committee’s January 27 plan — hold hearings, and “make recommendations to the Court”. 6 The masters held hearings for over two weeks, and, after hearing argument addressed to a draft report, issued their final report on March 31, 1975. The masters found the School Committee plan inadequate, in large part because of its reliance on parental free choice; rejected plaintiffs’ plan as being educationally deficient, unwieldly and arbitrary; rejected the December 16 plan as being vague and unduly burdensome to minorities; and proposed a ten district system, one being a citywide district with magnet schools and specially appealing programs, with each of the other nine districts and some of the magnet schools being paired with specific colleges, labor and business organizations for assistance in program enrichment. Mandatory busing was estimated by the masters to affect 10,700 to 14,900 students. The court then called for hearings on objections to the masters’ report, which commenced on April 10. In the meantime, the court had called for updated enrollment data from the School Committee. The court issued its Draft Revision of the Masters’ Report on April 17, heard comments on April 18, and issued its desegregation plan on May 10. Its plan reduced the number of districts from the ten recommended by the masters to nine, redrew district lines, and reflected — on the basis of the new data — less racial disparity generally in school assignments than did the masters’ plan. The plan not only precisely set forth the new districts but called for changes in the school system’s administrative hierarchy, and established a system of community participation in district and citywide councils. The mandatory busing estimated by the court would affect 21,000 students. During this time of accelerated activity, the court also appointed an ad hoc committee of three attorneys to assist in obtaining support from colleges and universities, and ordered that school personnel meet and confer with personnel from the designated college or- university. Late in June the court authorized the court-appointed experts to resolve some remaining issues relating to facilities utilization, program allocation, and enrollment limits. Each of these actions by the court *408 has been challenged, as well as the court’s plan itself. This skeletal recitation of chronology masks a year of increasingly intensive activity in collecting and updating data, preparing, evaluating, and amending plans embracing a wide variety of approaches, and, finally, devising procedures and taking action to put into effect a plan calling for a dramatically different educational system affecting some 80,000 students. The pressures of time, the problems of developing reliable data, the clash among radically differing approaches, the resistance of the School Committee, the sheer numbers of parents, students, teachers, and administrators to be informed and oriented were all part of the massive problems of implementation. While we appreciate the labors that have taken place by all concerned, we also appreciate the necessity of giving the most careful consideration to the issues before us. Some are of large significance. Some are of little moment. For purposes of clarity, we summarize them, not necessarily in terms of importance, but in terms of their breadth and specificity. A. Broad challenges to the court’s plan. 1. The School Committee contends that its free choice-third site plan was constitutionally sufficient and should have been adopted. 2. The Mayor contends that, the masters’ plan being constitutionally sufficient, the Equal Educational Opportunities Act of 1974, 20 U.S.C. § 1701 et seq., deprived the court of power to issue its own plan. 3. The Association contends that the court’s plan was erroneous because the court did not confine the remedy to eliminating the demonstrable effects of the School Committee’s unlawful actions. 4. The Association, the Mayor, and by implication, the School Committee, contend that the court’s plan is defective in not having given weight to the impact of the departure of white students to other schools outside the Boston school system. B. Detailed challenges to the court’s plan. 1. The School Committee and the Association contend that the court erred in a mechanical resort to ratios. The May- or and the Association object particularly to the use of ratios in effecting minority assignments to the high standard examination schools. 2. The School Committee challenges the appointment of three masters and one expert and objects to the compensation of all four masters. 3. The School Committee objects to the court’s encroachment on its functions: a. In specifying magnet programs; b. In requiring the participation of colleges and universities; c. In requiring the systematic involvement of a Citywide Coordinating Council and Community District Advisory Councils; d. In requiring the hiring of additional supervisory personnel; and e. In giving the supervisory power to court-appointed experts. * * * * * * ' A. Broad challenges to the court’s plan. Four sweeping criticisms have been leveled at the court’s plan. The simplest is that the School Committee’s plan passed constitutional muster and the court could not justifiably require more. A similar argument, taking another point of reference, is that since the masters’ plan was constitutionally sufficient, the court could not, by reason of the Equal Educational Opportunities Act of 1974, require more. The remaining two broad scale attacks do not assert a barrier to going beyond any other plan but rely on the necessity for making further inquiries into remedial issues before a final plan is implemented. One type of prerequisite inquiry would be to ascertain the prior impact of official segre-gative action so that the remedy could be restricted to removing that impact. The other asserted prerequisite would be an inquiry into the likelihood of “white flight” so that a remedy may be tailored which, by minimizing such flight, would *409 assure a maximum achievable co-education among the races. 1. Sufficiency of the School Committee’s Plan. Upon a finding that a school system has been operated in contravention of the equal protection clause of the Fourteenth Amendment to the constitution, the burden falls upon the local school authorities to present a plan of action to the district court to remedy the violations. Swann v. Charlotte-Mecklenberg Board of Education, 402 U.S. 1, 16, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971). Only on the default of the School Committee to proffer an acceptable remedial plan is the district court empowered to fashion a remedy adequate to produce a unitary school system. Id. The threshold question, then, in reviewing the district court’s action, is whether the rejection of the School Committee’s plan of January 27 was proper. In determining the acceptability of a proposed plan, the district court must assess the effectiveness of the plan in achieving desegregation. Green v. County School Board, 391 U.S. 430, 439, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968). The district court, applying this standard, rejected the School Committee plan stating that it “presented no more than a hope for desegregation in Boston”. 401 F.Supp. at 229. On review, we conclude not only that the district court’s assessment of the School Committee plan was proper, but that if the district court had accepted the January 27 plan, we would have been constrained to reverse. See Keyes v. School District No. 1, 521 F.2d 465 (10th Cir. 1975); Jackson v. Marvell School District No. 22, 416 F.2d 380 (5th Cir. 1969). The plan submitted by the School Committee was, in summary, a freedom of choice plan supplemented by magnet schools and third site resource centers. The school assignment process was to be based upon a series of options available to the students and their parents. Starting with the option to remain in the school attended in the previous year if it had been desegregated under Phase I, the student could choose in succeeding options to attend a citywide magnet school, a zonal magnet school, a school in which his race is in the minority, and finally, any school in the zone. At the end of this five step, seven week process a review committee would determine a course of action to deal with over-subscribed schools. 7 In the event that schools remained “racially isolated”, defined by the school department as more than a 15 percent deviation from the racial ratio for that level in the zone, the plan provided for mandatory student participation in resource center activities. The resource center proposal called for integrated educational experiences at a third site, once a week for elementary students, once every two weeks for middle school students and a human relations course for high school students. The entire plan rested on student and parental choice to desegregate the schools. It is well established that freedom of choice plans to desegregate school systems are not per se unconstitutional. Green, supra. In order for such a plan to be constitutionally acceptable, however, it must promise to be as effective in achieving a unitary desegregated system as any alternative and feasible plan. Id. Freedom of choice has a long history of failure in achieving desegregation both in the south, Swann, supra; Monroe v. Board of Commissioners, 391 U.S. 450, 88 S.Ct. 1700, 20 L.Ed.2d 733 (1968); Green, supra, 8 and in other parts *410 of the country, Kelly v. Guinn, 456 F.2d 100 (9th Cir. 1972) (Las Vegas); Spangler v. Pasadena City Board of Education, 375 F.Supp. 1304 (C.D.Cal.1974). To be sure, it may be argued that voluntary plans were less promising in the South than in the North, as segregated dual systems were deeply entrenched there, having had an express legal basis. Yet as the district court’s findings indicate, Boston had gone far in the creation of a de jure dual system. Boston, moreover, for ten years had a policy of open enrollment, followed by a controlled transfer policy riddled with exceptions. In an earlier stage of the present case the district court found that this policy served to increase, rather than decrease, segregation in the school system. Morgan v. Hennigan, supra, 379 F.Supp. at 449-59. The School Committee contends that its plan, although founded on freedom of choice, held promise to achieve desegregation due to the heavy reliance on magnet schools and alternative program schools — some fifty in number. Like freedom of choice, the use of magnet schools to achieve voluntary desegregation has failed elsewhere, Bradley v. Milliken, 484 F.2d 215, 243 (6th Cir. 1974), rev’d on other grounds, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974); Kelly v. Guinn, supra; Spangler v. Pasadena City Board of Education, supra; Dowell v. Board of Education, 338 F.Supp. 1256, 1264 (W.D.Okl.1972). Compare Hart v. Community School Board, 512 F.2d 37, 54 — 55 (2d Cir. 1975). Although Boston has had experience with one school which offered a specialized program and achieved a racially mixed student body, 9 we must agree with the district court that schools offering programmatic alternatives, while a useful supplement to an otherwise adequate desegregation plan, could not realistically sustain the burden of achieving desegregation of the Boston city schools. 10 Finally, the School Committee plan to remedy “racial isolation” with part-time integrated resource centers added nothing to the effectiveness of the overall plan. The objective sought to be achieved in a remedial plan is desegregation, not interracial experience or racial balance. Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974). Accordingly, similar part-time programs have been categorically rejected elsewhere, Keyes v. School District No. 1, supra, 521 F.2d at 477-79; Arvizu v. Waco Independent School District, 495 F.2d 499, 503 (5th Cir. 1974); United States v. Texas Education Agency, 467 F.2d 848, 859 (5th Cir. 1972), and must be rejected here. Against this historical background which promised failure for every feature of the School Committee plan, it is inconceivable that anyone, the School Committee members or the court, could believe that the plan would be effective in eliminating and guarding against officially imposed segregation in Boston. 11 *411 The district court, therefore, was clearly correct in declaring the School Committee in default of its obligation to present a constitutionally adequate plan. It was the district court’s unquestionable duty to utilize all the resources available to it as to fashion expeditiously a remedy that realistically would produce a unitary school system. Swann, supra. 2. The Constitutional Adequacy of the Masters’ Plan and the Applicability of the Equal Educational Opportunities Act of 1974. The Mayor’s principal attack on the court’s plan, essentially joined in by the Association, is that, since the masters’ plan was constitutionally sufficient, the court was without power to require more busing than that contemplated by the masters. The source of this alleged limitation is the Equal Educational Opportunities Act of 1974, 20 U.S.C. §§ 1701-1758. The Act manifests an intention of Congress that mandatory busing not be ordered to a greater extent than is required by the constitution. Section 1712 states that a court shall “impose only such remedies as are essential to correct particular denials of equal educational opportunity or equal protection of the laws.” 12 Section 1713 requires a court to consider, make specific findings about, and adopt a series of less onerous devices before requiring busing. These include school assignments close to home, transfers which would improve racial balance, revision of zones, and the construction of new schools and magnet schools. The Mayor’s argument proceeds simply: the masters’ plan contemplated mandatory busing for 6,100 fewer students of the 84,000 total student population than did the court’s plan; the court did not find that the additional compulsory transportation was required by constitutional necessity; and comparison with other cases demonstrates that in fact the masters’ plan was constitutionally sufficient; therefore, the court exceeded its powers. 13 *412 The masters filed their final recommendations on March 31, 1975. On April 10 the school department furnished new statistics on the size and racial composition of the student body, on the basis of which the court modified the recommended plan. As noted by the district court in its May 28 order denying a stay, “the masters’ report and recommendations contained the key elements and formed the foundation of the plan promulgated by the court.” The most significant change effected by the court was a change in the number of geographic districts from nine to eight. The appellants assert that this change required more busing than the masters’ plan would have required had it been updated to reflect the new figures. While this basic factual promise is open to some doubt, we need not rest our decision on such a necessarily elusive determination. 14 We proceed to examine the district court’s duties and powers, as they are affected by the Equal Educational Opportunities Act of 1974; whether the court purported to be guided by the Act; and whether its findings and conclusions are sufficiently supported. Prior to the passage of the Act it was clear that the mandate governing federal judges was to accomplish “the transition to a unitary, nonracial system of public education” in which “racial discrimination would be eliminated root and branch.” Green, supra, 391 U.S. at 436, 438, 88 S.Ct. at 1694. They were to make “every effort to achieve the greatest possible degree of actual desegregation, taking into account the practicalities of the situation.” Davis v. School Commissioners of Mobile County, 402 U.S. 33, 37, 91 S.Ct. 1289, 1292, 28 L.Ed.2d 577 (1971). And in so doing, “the scope of [their] equitable powers to remedy past wrongs is broad, for breadth and flexibility are inherent in equitable remedies.” Swann, supra, 402 U.S. at 15, 91 S.Ct. at 1276 (1971). The Act disavows any intention to encroach upon the obligation of the courts “to enforce fully” the constitution, § 1702(b). Moreover, it places reliance on specific findings with respect to the efficacy of particular remedies in the “priorities” section, § 1713, and on a more general finding of the inadequacy of other remedies in § 1755. By explicitly leaving the district court the power to determine the adequacy of remedies, the *413 Act necessarily does not restrict the breadth of discretion of that court to determine what scope of remedy is constitutionally required. Thus the Act manifests a purpose not to limit judicial power but to guide and channel its exercise. In a sense it is a statutory “less restrictive means” guideline, endeavoring to ensure that substantial compulsory transportation be used as a last resort. Our reviewing function remains the limited one of scrutinizing the record for sufficient factual support for the scope of the remedy, its reasonableness and its feasibility. See Swann, supra, 402 U.S. at 31, 91 S.Ct. 1267. The Act adds the factor that the district court’s findings must reflect a substantial consideration of the less restrictive means which Congress has required to be explored and used prior to resorting to compulsory transportation of any magnitude. 15 We turn therefore to an analysis of these two areas. Initially, we look to whether the remedy imposed in this case reflected the channelling contemplated by the Act. On this question, there can be little doubt. The district court clearly purported to be guided by the Act’s mandate. In the course of presenting its plan, the court stated: “Assignment of every student to the school closest or next closest to his residence, considering only school capacity, natural physical barriers or both, along with grade level and the type of education provided, cannot achieve substantial desegregation in Boston due to the geography of the city and racial and ethnic distribution in the city. 20 U.S.C. § 1713(a), (b), § 1714. Revision of attendance zones and grade structures, construction of new schools and the closing of old schools, a controlled transfer policy with limited exceptions and the creation of magnet schools have been used in the formulation of the plan here adopted in order to minimize mandatory transportation. 20 U.S.C. § 1713. The court finds, however, that some transportation of students to schools other than those next closest to their residences is required to remedy adequately the denial of plaintiffs’ constitutional rights and to eliminate the vestiges of a dual school system in Boston. 20 U.S.C. § 1702(b), § 1714(a).” 401 F.Supp. at 264. 16 We do not understand the Mayor to contest these findings; nor would there be any basis in the record for such a challenge. Rather, the Mayor’s theory goes less to which type of remedy (e. g., busing, magnet schools, etc.) was used by the court than to whether the remedial plan as a whole effectuated more desegregation (and as a result more busing) than was required. Since this argument concerns the district court’s discretion in determining the adequacy of the remedy, and therefore gains nothing from the Act which leaves that discretion unaffected, it must stand or fall upon those traditional principles of equity which would govern this issue even were the Act not in existence. We turn, therefore, to those principles and our traditional reviewing function. It is important to understand what the court’s plan accomplished. Assuming that it involves the forced busing of up to — but probably substantially fewer than — 6,100 more students than the masters’ plan, what did it get in return? We see the plan as one involving relatively unskewed, contiguous, compact districts without extensive gerrymandering or satellite zoning. It should be borne in mind that compulsory busing *414 occurs only within, not between, districts. Although both the masters’ plan and the court’s plan leave the East Boston schools 95 percent white, the court’s plan eliminated the other virtually one-.race schools (as would have existed under the masters’ plan in West Roxbury, 93 percent white) and many racially identifiable schools (predictable for the masters’ Burke district, 63 percent black). It reduced the racial distortion in the three districts which, under the masters’ plan (as the new figures affected it), would have been markedly dispro-portioned. By reducing racial identifia-bility, the plan did more to avoid the dilemma of either denying a district’s pupils access to citywide schools or allowing the district to lapse back into a one-race status. That the plan was not perfectionist is shown by the parameters of its district ratios, the percentage of whites ranging from a high of 61 percent to a low of 40 percent, in a school population 52 percent white, 36 percent black, and 12 percent other minority. Moreover, a 25 percent deviation was permitted for each district school. This compares with parameters in the masters’ plan, when the newer School Department data are taken into account, which are substantially more extreme at both ends. As for the compulsory transportation, the maximum is a 5 mile, 25 minute trip, the average being 2% miles, 10 to 15 minutes. Could the district court have reasonably found this additional desegregation to be constitutionally required? 17 To put it another way, does the court’s plan go beyond “every effort to achieve the greatest possible degree of actual desegregation, taking into account the practicalities of the situation”? Davis, supra, 402 U.S. at 37, 91 S.Ct. at 1292. The “practicality” of white flight is not a viable basis for declaring the plan invalid. See part A 4, infra. Apart from the practicality of geography which induced the court to exempt East Boston from Phase II, we see no other geographical factor of significant dimensions. To the extent that funding is a problem, we note that virtually the entire expense of any incremental busing is fundable by the state Board of Education, which supports the court’s plan. See Mass.G.L. c. 15, § 1I; c. 71, §§ 7A, 7B, 37I(ii). The amount of additional desegregation which was “purchased” by the court’s plan was neither trivial nor disproportionately burdensome. When we ask ourselves whether a slight increase of maximum percentage of planned white enrollment at some schools, or leaving some elementary schools unaffected, or adding a few more magnet schools would achieve the consti *415 tutional minimum with less compulsory busing, we realize that the concept of the minimum cannot be identified with precision. Whatever prescription may be adopted by a judge, after months and years of consideration, it is doubtless always possible to make a case that something less will do. We have no basis for holding that the court exceeded its obligation to do all that it feasibly could to extirpate the effects of the constitutional violations over the years. 3. Alleged Overbreadth of the Remedy. On January 20, 1975, the Association submitted a desegregation plan that was designed to restore the racial composition of the Boston schools to that which they would have had in the absence of any illegal official action. It contended that the district court was obliged to determine the extent to which the segregation in the Boston schools was attributable to official action and to limit the remedy to eliminating only that segregation. The factual premise of the Association’s plan was that the racial composition of most of Boston’s elementary schools and of two of its high schools, Charlestown and East Boston, is the result of residential patterns in Boston and not of the illegal acts of the School Committee. Its plan, accordingly, provided that these schools should not be affected by the court’s remedy. Since the Association’s plan challenged the remedial guidelines contained in the district court’s order of October 31, 1974, the district court treated the Association’s document as both a motion to modify the remedial guidelines prescribed by that order and as a proposed desegregation plan. The district court held a hearing on the issues presented by the Association on January 23, 1974. In support of its plan, the Association offered to introduce evidence that would establish that population patterns, not illegal state action, caused the existing racial segregation in the schools in question. The district court denied the motion to modify the remedial guidelines, holding that the Association’s remedial theory was inconsistent with the controlling Supreme Court precedents. 18 The court also held that the Association’s desegregation plan was constitutionally inadequate and could not be considered by the masters. Finally, the court refused to admit the Association’s evidence, on the grounds that it was irrelevant since the only question before the court at that time was how to accomplish the greatest amount of actual desegregation consistent with the practicalities of the circumstances and that the motion represented an attempt to reopen the findings of fact made by the district court at the liability stage of the proceedings and affirmed by this court. The Association has appealed, contending that the district court erred, first, in refusing to frame the remedy in terms of the specific consequences of the proven constitutional violations, and, second, in rejecting the evidence that the racial segregation in particular Boston schools was not the result of any state action. The central question on appeal is whether the district court erred in refusing to accept the Association’s remedial theory. In support of its theory that the district court must ascertain the extent to which state action caused the existing segregation in the schools and limit its remedy to . eliminating that segregation, the Association relies upon the language of the Fourteenth Amendment itself. Since the amendment prohibits only state imposed racial segregation, see Civil Rights Cases, 109 U.S. 3, 3 S.Ct. 18, 27 L.Ed. 835 (1883), the Association contends that a district court’s remedial power is limited to remedying the specific effects of the “state action” that formed the basis of the constitutional violation. Although the theory possesses some surface plausibility, the Supreme Court precedents clearly establish that the district court was correct in rejecting the Association’s proposed modification of its remedial guidelines. The remedial principles set forth in Swann, Davis, and *416 Green do not tolerate anything less than ensuring that the effects of constitutional violations are eliminated, and we are certain that the application of the Association’s remedial theories could not eradicate the effects of state imposed segregation. 19 To appreciate why the Association’s remedial theory must be rejected, it will be helpful to consider the nature of the constitutional violation. In the first stage of this case, the district court found that the School Committee’s policies regarding the construction of new facilities, the use of portable classrooms, overcrowding, districting and redistricting, feeder patterns, open enrollment, transfers, and faculty and staff assignments “were all marked by segregative intent” and substantially contributed to the segregated character of dozens of Boston schools. 379 F.Supp. at 426-30, 433-37, 442-45, 455, 459, 466-68, and 472. The latter part of the district court’s , opinion dealt specifically with the School Committee’s defenses that the racial segregation in the Boston schools— which it conceded to exist — resulted from private residential patterns and/or its racially neutral neighborhood school policy. The district court found that Boston had never followed a true neighborhood school policy and rejected the defense, relying in part on its earlier findings. 379 F.Supp. at 469-74. These specific findings, many of which related to elementary schools, were the basis for the determination that Boston’s school system as a whole violated the Fourteenth Amendment. Although the defendants could have limited the geographic scope of the violation by proving that parts of the Boston school system were geographically unrelated to the rest of the system and had not been operated with “segregative intent,” Keyes, supra, 413 U.S. at 203-05 and 210-13, 93 S.Ct. 2686, the School Committee failed to satisfy this burden with respect to any portion of the system. We affirmed the district court’s findings and legal conclusions in their entirety, Morgan v. Kerrigan, supra. The Association, in effect, argues that the trial on liability should be treated as the first of two battles, and that the second battle should involve a more particularized inquiry into the causes of the segregation at the individual schools within the system. Although the defendants failed, at the trial on liability, to persuade the district court that private residential patterns alone caused the segregation in the Boston schools, the Association maintains that the district court must, at the remedy stage, reconsider the effects of non-official action, determine the degree to which private action caused the existing segregation, and fashion a remedy that preserves the segregation that can be separated from governmental causation. This second battle would be considerably more complicated than the first. The logical implication of the Association’s proposal is that it would be proper for any group connected with any school to introduce proof that that school’s racial profile was only partially attributable to official action.' The district court could be faced with the task of making percentage findings as to every school in the district. The short answer to the Association is that its position is squarely contrary to the remedial principles of Swann, Davis, and Green. 20 See also *417 Keyes, supra, 413 U.S. at 200 and 214, 93 S.Ct. 2686. These cases establish that when intentional official action has significantly contributed to segregation in substantial portions of a school system, the individual schools in the system must be subjected to the maximum feasible desegregation if official action “created or maintained” the racial segregation contained therein. Swann, supra, 402 U.S. at 21 and 28, 91 S.Ct. 1267. See Keyes, supra, 413 U.S. at 214, 93 S.Ct. 2686. In the Boston case, the School Committee had the opportunity to prove that official action had not contributed to the segregated character of some of the individual schools in the system, but the district court found that the School Committee had failed to satisfy this burden. 21 Hence, under Swann, the district court was obligated to fashion a remedy that would accomplish the greatest amount of system-wide desegregation taking into account the practicalities of the situation. Swann, supra, 402 U.S. at 15-16, 91 S.Ct. 1267; Davis, supra, 402 U.S. at 37, 91 S.Ct. 1289. The Supreme Court has tacitly recognized the impotence of a remedy designed only to eliminate the demonstrable effects of past official conduct. While de jure segregation may not have been established at each and every school in a system, “common sense”, to use the words of the Court, supports the conclusion that effects of the proven discriminatory actions pervade the school system as a whole. Keyes, supra, at 201. Acts that establish one school as white or as black will have a reciprocal effect on the racial composition of nearby schools. Id. at 202-03; Swann, supra, 402 U.S. at 20-21, 91 S.Ct. 1267. The use of various devices to earmark schools according to their racial compositions may well have had a “profound reciprocal effect on the racial composition of the residential neighborhoods within the [city], thereby causing further racial concentration within the schools.” Keyes, *418 supra, 413 U.S. at 202, 93 S.Ct. at 2694. “People gravitate toward school facilities just as schools are located in response to the needs of the people.” Swann, supra, 402 U.S. at 20, 91 S.Ct. at 1278. 22 From a practical point of view, the problems of determining what the racial composition of neighborhoods would have been, absent the unlawful discrimination, would be especially acute in such a case as this where the school authorities did not follow a genuine neighborhood school policy and where one form of discrimination was the locating of new schools and the overcrowding of existing facilities. 379 F.Supp. at 427-29, 469-74; see Keyes, supra, 413 U.S. at 211-13, 93 S.Ct. 2686. The task of unscrambling cause and effect would be, to understate it, awesome. Even if we assume that the district court could reexamine each school to determine the shares of segregation attributable to public and private action, the application of the Association’s theory would fail to vindicate the constitutional rights of many students presently enrolled in the Boston schools. It, of course, is the rights of the individual students that are in question. Morgan v. Kerrigan, 509 F.2d 599, 600 n. 3 (1st Cir. 1975); see Brown v. Board of Education, 347 U.S. 483, 494, 74 S.Ct. 686, 98 L.Ed. 873 (1954). Even if the court could reliably determine that 40 percent of a school’s segregation was caused by official action and 60 percent by private residential patterns, it could not bifurcate an individual student. The result would inevitably be that some victims of the School Committee’s official policy would be forced to continue a segregated education. Apart from the failure of the Association’s theory to remedy the violations of individual rights per se, its adoption would seem to us to turn the process of desegregation on its head. Unconstitutional segregation is defined not only by percentages but also by community and administrative attitudes, see Keyes, supra, 413 U.S. at 196, 93 S.Ct. 2686, and by psychological effects, see Brown, supra, 347 U.S. at 494, 74 S.Ct. 686. To require a district court to preserve intact every scrap of segregated education that somehow can be separated from governmental causation is to involve the federal courts in planning continued segregation and in perpetuating the community and administrative attitudes and psychological effects which desegregation should assuage. Cf. Wright v. Council of City of Emporia, 407 U.S. 451, 465-66, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972). 23 We add one further observation. The court had found that in the past the School Committee had given racially isolated black schools less support, particularly in the quality of faculties supplied, than other schools. The court had good reason to believe that if black and Hispanic students were not spread, visibly and bodily, in the mainstream among schools throughout the system, the racially identifiable schools in which they remained would continue to invite discriminatorily unequal treatment evading timely and effective remedy. A racially *419 mixed population in each school would, on the other hand, be an insurance policy against any purposefully unequal allocation of resources. While the liability phase of this case primarily involved discriminatory separation, the realistic prospect of discriminatory inequality of support for schools which, though no longer subject to racial districting, transfer, or other such policies, remained predominantly black is a pragmatic factor further supporting the court’s remedial order. It is clear, therefore, that practical considerations as well as established principles mandated the rejection of the Association’s remedial theory. 24 Hence, we hold that the district court did not err in refusing to modify its remedial guidelines. It follows that the district court was correct in refusing to admit the Association’s evidence on the causes of the existing segregation in the Boston schools. To the extent that the evidence was offered to establish the degree to which illegal state action had caused the racial segregation in the Boston schools, the evidence was legally irrelevant at this stage of the proceedings. 25 The only question before the district court was how to accomplish the greatest amount of actual desegregation consistent with the practicalities of the circumstances. To the extent that the evidence was offered to rebut the Keyes presumptions, it came too late and was barred by established principles of preclusion. 26 It is elementary that the district court was not required to reopen factual findings and legal conclusions reached after vigorous litigation — the whole point of which was to determine whether the segregation which eoncededly existed was caused by unconstitutional state action— and affirmed on appeal. See, e. g., White v. Higgins, 116 F.2d 312, 317-18 (1st Cir. 1940); Bee Mach Co. v. Freeman, 131 F.2d 190, 192-93 (1st Cir.), aff’d, 319 U.S. 448, 63 S.Ct. 1146, 87 L.Ed. 1509 (1943); Hodgson v. Brookhaven Gen’l Hospital, 470 F.2d 729 (5th Cir. 1972). 4. The “White Flight” Controversy. The district court ruled that “white flight,” defined as the departure of white children from the Boston city schools to parochial, private, or suburban school systems, is not a practicality for which the plan must make an accommodation. Morgan v. Kerrigan, 401 F.Supp. at 233-34. See Davis, supra, 402 *420 U.S. at 37, 91 S.Ct. 1289. The Mayor and the Association challenge this ruling as an abuse of discretion, claiming that “white flight” alters the effectiveness of a desegregation plan and leads to “reseg-regation” of the schools. 27 White flight is an expression of opposition by individuals in the community to desegregation of the school system. Monroe, supra; Jackson v. Marvell School District No. 22, 416 F.2d 380 (8th Cir. 1969); Lee v. Macon County Board of Education, 448 F.2d 746 (5th Cir. 1971). From the inception of school desegregation litigation, accommodation of opposition to desegregation by failing to implement a constitutionally necessary plan has been impermissible. Brown v. Board of Education, 349 U.S. 294, 75 S.Ct. 753, 99 L.Ed. 1083 (1955) (Brown II). 28 Appellants contend, however, that white flight differs from other forms of opposition, because its effects, the withdrawal of white pupils from the school system, alter the effectiveness of the desegregation plan. The school system, they claim is “resegregated”: the city school system largely black and other minority; the private and suburban systems, largely white. 29 To prevent this *421 result, appellants claim that the district court should consider white flight a “practicality”, and limit the amount of desegregation to that level which would enjoy acceptance in the white community. 30 There are two endemic flaws in this argument. First, in the trial on liability, evidence was presented that feeder patterns, district lines, and open transfer policies were established for the purpose of satisfying purported white community desires. 379 F.Supp. at 438, 449. This evidence resulted in a finding that the Boston schools were administered in violation of the Fourteenth Amendment. Appellants now ask that the district court, in devising a remedy for these violations, respond in the same way as the Boston School Committee did to the same perceived community attitudes: draw district lines, assign pupils to schools, and limit racial mixture to reduce “white flight”. In other words, while appellants dwell upon the unpleasant prospect of an inner city black school system surrounded by suburban white school systems, the prospect contemplated by their approach is that of an inner city segregated system, created unlawfully, but permitted to endure because the apprehension of massive white flight has made legal what had once been in violation of the constitution. Second* appellants’ claim that white flight destroys the effectiveness of the school desegregation plan, because of “resegregation” of the school system, founders on the constitutional definition of unlawful segregation. The Supreme Court has recently reemphasized that the constitutional right is to attend school in a unitary, non-discriminatory, public school system. It is not to attend school *422 in a system which is comprised of students of a racial balance which exists in the general geographical area. Milliken v. Bradley, 418 U.S. 717, 746, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974). Accord Calhoun v. Cook, 522 F.2d 717 at 719 (5th Cir. 1975); Mapp v. Board of Education of Chattanooga, 525 F.2d 169 at 170-171 (6th Cir. 1975). What the layman calls “resegregation” is not constitutionally recognized segregation. It is racial isolation imposed by historic school district boundaries, Milliken v. Bradley, supra, or by individual choices to attend private institutions. Compare McCrary v. Runyon, 515 F.2d 1082 (4th Cir.), cert. granted, 423 U.S. 945, 96 S.Ct. 354, 46 L.Ed.2d 276 (1975). This racial isolation becomes constitutionally significant only when the district boundaries are drawn with segregative intent, Evans v. Buchanan, 393 F.Supp. 428, 445-46 (D.C. Del.), aff’d 423 U.S. 963, 96 S.Ct. 381, 46 L.Ed.2d 293 (1975); United States v. Missouri, 515 F.2d 1365 (8th Cir.), cert. denied 423 U.S. 951, 96 S.Ct. 374, 46 L.Ed.2d 288 (1975); see United States v. Scotland Neck Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75 (1972), or when the state participates in the private institutions. Norwood v. Harrison, 413 U.S. 455, 463-65, 93 S.Ct. 2804, 37 L.Ed.2d 723 (1973). The constitution cannot solve all problems. On the contrary, to the extent that it demands that rights which have previously been overridden be enforced, it creates social problems. It inconveniences, sometimes substantially, law enforcement officers, prison wardens, university administrators, and government bureacrats. And, when it allows tasteless books to be sold or movies shown, many are offended. But ex-pectable individual, official or group reaction does not outweigh constitutional rights. We therefore must agree with another court which said, “concern over ‘white flight’ . . . cannot become the higher value at the expense of rendering equal protection of the laws the lower value.” Mapp v. Board of Education of Chattanooga, supra, quoting 366 F.Supp. 1257, 1260 (E.D.Tenn.1973). The bright note in this otherwise somber picture is the care and imagination that the district court has displayed in structuring a diversified educational system offering superior opportunities for children, both white and black. The plan is not a mechanical device to ensure that the races share equally, but serves its constitutional goals within a framework offering educational hope for the children of the city. Nevertheless, federal courts have a limited jurisdiction and competence. To the extent that reorienting the Boston school system involves social expenses, it must be paid for in coin less dear than the constitutional rights of the city’s citizens. Here as elsewhere, the Boston community must look to other institutions, city, state, federal and private, to contribute to an effort to vindicate the constitutional rights of its citizens at a minimum of social cost. B. Detailed Challenges to the Court’s Plan. The issues we have discussed above address the constitutional and statutory limitations on the court’s power to issue a plan at variance with the Committee’s or the masters’ plan (Parts A 1 and 2) and to refuse, in tailoring its plan to consider evidence of the precise impact of prior official segregative actions and of the extent to which its plan might cause the departure of white students from the school system (Parts A 3 and 4). We now consider a range of more specific objections to particular features of the plan and its method of implementation. 1. Ratios and the Examination Schools. Appellants School Committee, Mayor and Association challenge the district court’s use of racial percentages and quotas in several contexts: the guidelines for the composition of the schools in community school districts, the guidelines for the composition of citywide magnet schools, and the minimum percentage of black and Hispanic students mandated to be admitted to the elite, examination schools. Appellants claim *423 that the district court’s use of percentages was in violation of the Supreme Court’s disapproval of fixed racial quotas in Swann, supra. We find that the district court resorted to percentages in an appropriate manner throughout. The district court first ordered the overall racial composition to be used as a starting point in designing a school desegregation plan. This approach is specifically approved in Swann. In devising the court plan, community school districts were drawn to provide for a rough equality of racial composition among the districts to be desegregated. 31 Within each community school district, the court ordered that students should be assigned to particular schools so that each school’s population approximates the composition of the community district. A deviation margin of ± 25 percent of the racial percentage figures was allowed to provide flexibility in planning. The desegregated schools in Boston, therefore, range between 30-70 percent white and 30-70 percent black and other minority under the plan. 32 This use of statistical ranges is consistent with other desegregation cases, United States v. School District of Omaha, 521 F.2d 530, 546-547 (8th Cir. 1975) (citywide school racial composition, 80% W— 20% B; schools to be 0-35% B, 65-100% W) Yarborough v. Hulbert-West Memphis School District No. 4, 457 F.2d 333 (8th Cir. 1972) (citywide elementary schools 47% W-53% B; schools to be 30-70% W, 30-70 B), and does not establish racial quotas in contravention of Swann. A more significant challenge if made to the guidelines for the composition of the citywide magnet schools. Each school is to limit its enrollment to fall within 5 percentage points of the citywide racial composition. 33 The district court found that a narrow range of enrollment ratios was permissible for schools in the citywide district because the practicalities to be accommodated, primarily geographic elsewhere in the system, were significantly smaller in the citywide district schools. There are, moreover, substantial positive reasons for enforcing a narrow range of permissible enrollments. The magnet schools were designed to maximize voluntary desegregation. Significant departures from the overall racial composition could cause the magnet schools to hinder rather than help the process of desegregation. First, attendance at the schools is voluntary. Few individual students will choose to attend a school which is predominantly of the other race. Second, the students attending the magnet schools come from the community school districts. If disproportionate numbers of one race transfer to the citywide school district, the racial composition of the community districts suffer. Fundamentally, the magnet schools, in order to prove of value to the desegregation plan, had to be carefully circumscribed to ensure that they would not serve as a haven for those seeking to attend a school predominantly composed of those of their own race. Finally, appellants challenge the district court’s order that at least 35 per *424 cent of the incoming class at the elite, examination schools be black or Hispanic as a racial quota. At face value, this directive does appear to establish a fixed racial balance for those classes at Boston Latin School, Boston Latin Academy, and Technical High School. But, once the schools are viewed as magnet schools, and as part of the citywide school district, the objection disappears. True, these schools are treated differently from the other magnet schools in that only the entering classes are desegregated, rather than the entire student body. This is to accommodate the cumulative nature of the instruction offered at these schools. For the entering classes, however, the district court’s order provides that the racial composition shall be similar to that of the other magnet schools: at least 41% black and other minority. 34 It is, therefore, no more an impermissible racial quota than are the guidelines established for magnet schools generally. There are, however, other, more fundamental challenges to the district court’s treatment of the examination schools. The Association protests the schools’ inclusion in the remedial plan since no specific segregative acts were proven in their administration. It also claims that a racial preference for admission is unconstitutional discrimination on the basis of race. The Mayor challenges the rejection of several alterna-five plans for the desegregation of the examination schools which would have imposed specific admissions criteria. We start with the proposition that it is not unconstitutional per se for a city school system to operate an elite school even though low income or minority children may be under-represented in the student body, Berkelman v. San Francisco Unified School District, 501 F.2d 1264, 1267 (9th Cir. 1974). The examination schools in Boston, however, are an integral part of a school system which has been found to be administered in an unconstitutional manner. They are presumed to be unlawfully segregated. Keyes, supra. As such, the examination schools must be part of the remedial plan. See Part A 3, supra, 35 Several other plans for desegregation of the examination schools were presented to the district court. All relied on the Secondary School Admission Test (SSAT) to establish admissions criteria. 36 The district court found that the SSAT, although apparently of some predictive accuracy, had not been validated as a means of identifying students who can benefit from the examination schools’ curricula. Cf. Castro v. Beecher, 459 F.2d 725, 732, 735-36 (1st Cir. 1972). Because of the limited amount of statistical data available, moreover, there was no assurance that any of these plans *425 would admit a significant number of minority students. Given these factors, we find that the court acted within its discretion in rejecting the alternative plans. The appellants claim that the district court’s order will destroy the examination schools as elite academic institutions. If the order were inflexibly to require, for some years to come, the admission of blacks and Hispanics despite demonstrable underqualification by validated selection processes, we would hesitate to affirm. But the order is a temporary expedient, designed to ensure that the examination schools participate in the desegregation of Boston schools, pending development of racially neutral admissions criteria and the desegregation of the elementary schools. Advanced work classes at the elementary level which successfully feed students into the examination schools are being desegregated. Therefore, there is promise that more minority students will become eligible for admission to the examination schools under any admissions standard. 37 The parties have been specifically invited by the court to develop admissions criteria which can be shown to identify accurately students who can benefit from the examination schools’ programs. 401 F.Supp. at 244. See Smuck v. Hobson, 132 U.S.App.D.C. 372, 408 F.2d 175, 187-90 (D.C.Cir. 1969). We are convinced that the district court will monitor the progress of desegregation in the examination schools and will adjust its future orders to ensure the continued vitality of these schools as elite institutions. For the present, however, we affirm the court’s treatment of the schools as within its discretion. 2. Challenges to Composition of Masters’ Panel Expert Dentler and to the Masters’ Compensation. After the various desegregation plans had been submitted, the district court appointed two experts to assist it in evaluating the plans and also appointed a panel of four masters to consider the desegregation plans, hold evidentiary hearings, report to the court and recommend a desegregation plan. It also directed the city defendants; the School Committee, and the Superintendent of Schools, to pay the masters compensation of $200 per diem plus expenses. The School Committee objected to the appointment of three of the masters— Francis Keppel, Charles Willie, and Edward McCormack — and of one of the experts — Robert Dentler — on the grounds that their associations with the plaintiffs rendered them unqualified to serve. 38 These objections were overruled, and the masters and experts performed their assigned tasks. On May 2, 1975, the district court ordered the city defendants to pay the masters compensation of $21,-906.13. The School Committee objected, 39 but its objection was overruled. The School Committee appeals the district court’s rulings on the qualifications of the masters and expert and on the masters’ compensation. There is no merit to the School Committee’s contention that the district court’s failure to disqualify the masters and expert constituted reversible error. We observe, preliminarily, that the chance of error, if any, inflicting prejudice was remote. We have held that the masters’ and district court’s rejection of the School Committee’s desegregation plan was constitutionally required. The *426 masters’ plan, while-relied upon by the district court, was itself strengthened by the district court and thereby made less acceptable to the School Committee. Moreover, the impartiality of the masters is suggested by their criticism and rejection of the plaintiffs’ plan. Thus, any error would seem to us to have been rendered harmless. See Swann v. Charlotte-Mecklenburg Board of Education, 431 F.2d 138, 148 (4th Cir. 1970), rev’d on other grounds, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971). We need not rest our decision on this basis since we agree that there were no grounds for disqualifying any of the masters or experts. The basis for the School Committee’s objections to Masters Keppel and Willie is that they have been associated with the Harvard University Graduate School of Education and thus are indirectly related to the Harvard Center for Law and Education, three of whose staff attorneys have helped represent the plaintiffs in this case. The ground for the objection to Master McCormack and Expert Dentler is that they have supported the NAACP which, although not a party, supports the plaintiffs’ suit, has advanced funds to cover disbursements, and has permitted its General Counsel to serve as one of plaintiffs’ counsel. The relationships of these individuals to the plaintiffs’ counsel are attenuated. Master Keppel was Dean of the Graduate School of Education until 1962 and Master Willie has been a member of the faculty there since 1974. 40 The Harvard Center for Law and Education, which was founded in 1969, is completely independent of the Graduate School of Education. Although it is housed in the Graduate School, it is funded by the federal government and pays the Graduate School for space and services. We failed to see how a relationship with the Graduate School could give rise to a reasonable inference of possible bias in favor of the plaintiffs with respect to the scope of the desegregation remedy in this case. We reach a similar conclusion with respect to the other master and expert. Expert Dent-ler has been a member of the NAACP in the past, but could not remember whether he paid his dues for 1975. Master McCormack became a life member in 1960. Although these relationships suggest general agreement with the goals of the NAACP, they hardly provide a reasonable basis for concluding that these individuals would be biased in the plaintiffs’ favor with respect to the breadth of the desegregation remedy. We note, moreover, that several of these individuals have relationships with the School Committee which might, to an equal extent, suggest bias in the School Committee’s favor. 41 Since masters and experts are subject to the control of the court and since there is a need to hire individuals with expertise in particular subject matters, masters and experts have not been held to the strict standards of impartiality that are applied to judges. See United States v. Certain Parcels of Land, 384 F.2d 677, 681 (4th Cir. 1967); Scott v. Spanjer Bros., Inc., 298 F.2d 928, 931 (2d Cir. 1962). Cf. 28 U.S.C. §§ 144, 455. Here, we doubt that, even under the strict standards applied to judges, the district court erred. Motions to disqualify judges have frequently been denied even in cases far stronger than this case. See, e. g., McNeil Bros. Co. v. Cohen, 264 F.2d 186, 188-89 (1st Cir. 1959); Eisler v. United States, 83 U.S.App.D.C. 315, 170 F.2d 273, 278 (D.C.Cir. 1948); Darlington v. Studebaker-Packard Corp., 261 F.2d 903, 906 (7th Cir. 1959); Weiss v. Hunna, 312 F.2d 711, 714 (2d Cir. 1963). We *427 conclude that these individuals were qualified under the more relaxed standards that are applied to masters and experts. The School Committee’s challenge to the order that it pay the masters’ compensation is also without merit. The district court has broad discretion in fixing the amount of such compensation and in determining which of the parties to charge. F.R.Civ.P. 53(a). Here there is no basis for concluding that the district court abused its discretion. The grounds for the School Committee’s objection is its belief that the reference to the masters was unnecessary and unwise and that the masters expended a great deal of time and energy fruitlessly. Assuming arguendo that these allegations would constitute grounds for setting aside the district court’s order, we note that the district court reasonably reached contrary conclusions regarding the necessity and utility of the masters’ work. We hold that there was no abuse of discretion. 3. Alleged Encroachment on School Committee Functions. To effectuate its plan in this ease, the district court adopted what it called a “multiplicity of measures” beyond the mechanical redistribution of students. The broad purpose of these measures, as articulated in the opinion below, was effective implementation of equal educational opportunity through banning active discrimination, meeting the persisting effects of past discrimination, and alleviating the difficulties of transition to a desegregated situation. The School Committee concedes these are proper goals, but objects to some specific measures as encroachments on its traditional powers and functions. Although we discuss the challenged provisions individually below, the greatest justification of each, in our view, is common to all. The overriding fact of the matter is that the district court in this case has had to deal with an intransigent and obstructionist School Committee majority. These elected officials engaged in a pattern of resistance, defiance and delay. See generally Part II of the opinion below; on previous Committee resistance to state desegregative efforts, see 379 F.Supp. at 418-21, 429-32, 438-41, 451-56, and 477. The most visible recent example of such conduct was their refusal to authorize submission to the court of the desegregation plan drawn up by the Committee staff, an action which provided the basis for subsequent contempt convictions; upon the rather lenient purging of those convictions, they then submitted a freedom of choice plan which they knew was inadequate. And they did not hesitate to manifest directly to the court their unwillingness to cooperate with the effort to desegregate the Boston schools. The court addressed interrogatories to the Committee members late in December, 1974, relative to their willingness to promote the peaceful implementation of the Phase I interim plan in effect during 1974-75. . The then Chairman responded: “I will continue to obey lawful orders of the Court, but I will take no initiative or affirmative action to advocate or supplement this plan . .” Similar responses were made with respect to the future promulgation of a court-developed plan. In short, the obstruction by the School Committee was substantial and the district court had every reason to believe it would continue. This crucial fact justifies, in our opinion, a number of extraordinary measures which might otherwise be open to question. a. Specifying magnet programs. The School Committee does not object to the concept of magnet schools; indeed, Boston has had a number of such schools for many years. The Committee’s plan retained these, and proposed a total of about 50 magnet programs, more than twice as many as the court’s plan provided for. 42 The narrow objection under this heading is to the court’s specifi *428 cation of the particular programs to be developed at several of those schools. 43 The magnet schools play a central part in the court’s desegregation scheme. The citywide school district contains an unusually high number of these facilities meant to attract students from all over the city to specific distinctive programs that appeal to them. If successful, this desegregative tool will stimulate substantial voluntary student transportation, reducing forced busing; and 20 U.S.C. § 1713(f) required the court to look to magnet schools before resorting to mandatory transportation beyond the next nearest school. Implicit in the power to use magnet schools, at least upon the default of the School Committee, is the power to specify programs essential to make them magnetic. In this case, that power was more crucial because of the importance of the magnet schools to the plan being implemented. We do not find in the record below any School Committee objection to the court’s power to order programs, or to the specific programs required. Nor do we see any indication that the district court would be unreceptive to any equally attractive programs which the School Committee might desire to substitute for those provided for in the court’s plan. Good faith cooperation by the Committee would appear to be all that is required. b. The contracts with colleges and universities. As part of its effort to provide magnet programs that would be distinctive and attractive enough to work, and more generally to equalize educational opportunity, the court enlisted the aid of a number of area colleges and universities, each to participate in the development of programs at specific schools. The court ordered the school department to “use its best efforts to negotiate a contract pertaining to each paired school acceptable to both the School Committee, and the contracting institution of higher learning.” 401 F.Supp. at 247. The court refrained from specifying any of the terms of these contracts, except to note that the institutions should not “usurp or replace the proper role of the School Department or any of its employees . .” The state defendants, which have advanced $900,000 and are likely to absorb much of the future cost of his project, see Mass.G.L. c. 71, § 371; e. 15, § II, and 401 F.Supp. at 246, and the Boston Teachers Union support this aspect of the court’s order; but the School Committee charges that it enters the area of quality of education, and is thereby outside the court’s authority. “In default by the school authorities of their obligation to proffer acceptable remedies, a district court has a broad power to fashion a remedy that will assure a unitary school system.” Swann, supra, 402 U.S. at 16, 91 S.Ct. at 1276. The magnet schools were a central part of the court’s remedy, but to accomplish their purpose they had to be both well conceived and implemented in good faith. The School Committee takes out of context remarks of the court that assistance of the colleges would improve the quality of education in the system and “wonders how the issue of quality of education arose”. This argument — that the court has no legitimate function to improve quality — entirely misses the point that if magnet schools are to act as lodestones, drawing students voluntarily to their programs, quality is a key to this aspect of a plan of desegregation. And the court had been led in the clearest terms to understand that the Committee would do only what it was ordered to do. Reliance on the Committee to create imaginative programs of utility and attractiveness would not only have been ill advised, but the supervision of compliance in this area, as opposed to student assignment for example, would have been extraordinarily complex and might well have drawn the court into *429 purely educational decisions. Paradoxically, therefore, the resort to outside contracting may have an effect precisely the opposite to that alleged by the appellants. In ordering the contract talks, the court removed itself from the content of educational decisions as much as possible by mandating best efforts negotiation rather than any specific provisions. In the context of this case, that order amounted to the most reliable and least intrusive method of ensuring that the School Committee would act to implement the magnet school provisions, which in turn were important to make the remedial plan work. This provision of the order, being innovative, is without precise precedent in other cases. But in light of the background of the case and the particular objective being served, we hold the best efforts provisions to be within the equitable discretion of the court. c. The citywide and community councils. The Citywide Coordinating Council and Community District Advisory Councils were established to monitor implementation on behalf of the court. The CCC was also charged with trying to identify and resolve problems, but was explicitly not authorized to “co-manage or make policies for the Boston schools.” 44 Citizen participation in the desegregation process was embodied in the Community District Advisory Councils, drawn largely from the unobjectedto racial-ethnic parents’ and students’ councils established the previous year. Such groups have been apprtíved in many cases, e. g., Dowell v. Board of Education, 465 F.2d 1012, 1015-16 (10th Cir. 1972); Singleton v. Jackson Municipal Separate School District, 426 F.2d 1364, 1370 (5th Cir. 1970). The challenge to these provisions asserts that they involve more than mere monitoring. The Committee objects that the CCC’s duties of attempting to resolve problems and of offering advice to the school department infringe on the Committee’s power to have general charge of the schools. As we view the mandate, however, no substantive power was accorded to the CCC and CDACs. Apart from the invaluable function of clustering groups of different generations and races in an attempt to promote citizen understanding and support, the monitoring process is a basic responsibility of the court. To the extent that the myriad of minor problems which will arise can be resolved without the necessity of resorting to the district judge, the process of implementation will be facilitated. The Committee objects that the CCC was instructed “to support efforts to improve the quality of education.” Attention to educational programs from the perspectives of banning active discrimination, assuring equal educational opportunity, meeting the effects of past discrimination, and alleviating the transition to desegregation is obviously a proper goal needing no justification. While better quality education as a general goal is beyond the proper concern of a desegregation court, we do not view the court’s instructions to be divorced from its efforts to devise an effective plan of desegregation. d. Additional supervisory personnel. Since the court’s plan established three more districts than had previously existed, it required three more district superintendents to be hired. This requirement is contested, but not vigorously. The more major objection is to the court’s order that each elementary school be headed by an administrator of the rank of headmaster or principal. If these provisions were reasonable, both were within the power of the district court. The equalization of educational opportunity through the implementation of the desegregation remedy at the schools requires effective administration. Similar orders have been upheld routinely. See Davis v. School District of City of Pontiac, 487 F.2d 890 (6th Cir. 1973); Plaquemines Parish School Board v. United States, 415 F.2d 817, 821 n. 2 (5th Cir. 1969). But the *430 Committee (joined in this point by the Mayor), argues that the order as to elementary school administration was arbitrary and unreasonable, and therefore an abuse of discretion. The elementary schools previously were joined in multi-school districts headed by district principals. With the dissolution of these mul-ti-school units, a number of which the district court had found to be segregated within themselves, 379 F.Supp. at 437, over 100 elementary schools exist under the court’s plan, including six with a capacity between 100 and 180 students. The School Committee charges that the court’s'order would require 80 new principals; the Mayor, perhaps partly because he concedes the need for principals in each elementary school housing more than 1,000 pupils, places the figure at 53. This argument is not addressed to the proposition that each school should be under the supervisory responsibility of one person but to the implication of the court’s order, which refers to “rank of principal or headmaster”, that the smallest of schools shall be headed by a person commanding the same grade level and pay as a person who supervises the largest. Subsequent to the court’s order, we are informed, the court stated that its intent was simply that there be a “person in charge” of each facility; the level of compensation was left to the School Department and Committee. This seems to us the sensible approach. The district court having clarified the meaning of its order, we see no reason at this juncture to set it aside, but leave the matter to be worked out by the school authorities to the satisfaction of the district court. e. The power given to the experts. The Committee objects to three orders of the court with respect to its experts. The May 10 order promulgating the court-developed desegregation plan provided that the assignment of students be carried out “under the supervision of representatives of the court”. The memorandum of decision on June 5 stated: “The nature of instruction given in the schools must also receive the attention of the court and its representatives. Instruction must be non-discriminatory and avoid racial stereotyping.” 401 F.Supp. at 234. We see nothing unusual in these orders. Experts are commonly used to assist the court in planning desegregation, see generally Hart v. Community School Board of Brooklyn, 383 F.Supp. 699, 764-67 (E.D.N.Y.1974), aff’d, 512 F.2d 37 (2d Cir. 1975), and under the circumstances of this case were justifiably used to assure implementation as well. It is regrettable that there was a need to closely monitor the assignment process, but it is the Committee’s own doing. To have neglected it, or to have failed to assure that instruction was nondiscriminatory would have been irresponsible of the court. The June 20 order was a response to a delayed School Committee proposal about certain of the guidelines which needed to be settled before the student assignment process could begin. The proposal conflicted substantively with the court’s previous orders in several respects, and advanced positions which (also contrary to court orders) had not been discussed with the court’s experts. Operating under serious time pressure to effectuate the desegregation plan smoothly for the fall, the court “authorized the court-appointed experts to resolve forthwith the remaining issues with respect to facilities utilization, program allocation and enrollment units. The court will as soon as feasible review with the court-appointed experts their determination of these remaining issues . .” While this order perhaps gave the experts an unusual, if brief, amount of power, it was justified by the School Committee’s actual violations of the court’s substantive and procedural orders, and its unwarranted delay in the face of the urgent necessity of finalizing these decisions. The court specifically found that absent this order “the student assignment process would be stalled and the implementation of the May 10 plan be jeopardized.” Along with the provisions for immediate court review, this order was proper under the circumstances. Moreover, all of the duties specifically provided for in these three orders having *431 now been substantially performed, the only active controversy remaining under this heading probably is the continued existence of the experts themselves. The Committee urges that their function is completed, and that they “should now return to their university.” We agree that this will be appropriate once a unitary school system has been established, and perhaps even earlier in the court’s discretion. The speed with which this goal is accomplished, however, rests in large measure in the hands of the Committee itself. We note that a new Committee was elected in November; it is our hope that the new majority will be more constructive than the old. If so, the court measures discussed in this section, many of them undoubtedly aggravating, may cease to be necessary. $ * * . * * $ We therefore affirm the District Court’s plan and implementation order for Phase II. Footnotes [1] The court specifically found that segregative policies had operated in the following respects: in the utilization of facilities and planning of new structures; in drawing and redrawing school district lines; in developing feeder patterns determining enrollments at specific high schools; in the open enrollment policy, the subsequent controlled transfer policy and the exceptions thereto; in the hiring, promotion, and assignment of black faculty and staff. Because of the operation of the second presumption in Keyes v. School District No. 1, 413 U.S. 189, 208, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973), which it found had not been rebutted, the court also held that official intentional segregation infected the elite, citywide examination schools and vocational schools and programs. [2] The original parties, in addition to the plaintiffs and the Boston defendants, included the State Board of Education (nominal defendants, although supporting the court’s order and ap•pellee here). Subsequently allowed to intervene were the Boston Teachers Union, the Boston Association of School Administrators and Headmasters, the Boston Home and School Association, and El Comite De Padres Pro Defensa De La Education Bilingüe. The Commissioners of the Public Facilities Commission, the Director of the Public Facilities Department, and the Mayor of Boston (hereinafter, collectively, the Mayor) were joined as parties defendant. [3] Following the determination on liability in 1974, the district court ordered the implementation of a plan devised by the state for the 1974-1975 school year. This plan was commonly referred to as Phase I. See 379 F.Supp. at 483. [4] The order provided that the “starting point” for its desegregation decree would be that “the racial composition of the student body of every school should generally reflect the ratios of white and black students enrolled at that grade level of schools, elementary, intermediate, and secondary, throughout the system.” [5] This action by the School Committee led to contempt proceedings and a finding by the district court that three members were in continuing contempt of the order of October 31, 1974. This court denied a stay pending appeal of the civil contempt order. Morgan v. Kerrigan, 509 F.2d 618 (1st Cir. 1975). Subsequently, the district court found that they had purged themselves, and, ultimately, on January 27, 1975, the School Committee did submit a plan, different from the staff proposal filed by counsel. [6] An objection was made by the School Committee to the naming of two of the masters on the ground that their association with the Harvard University Graduate School of Education constituted a disqualifying interest in the Harvard Center for Law and Education, which presently or formerly employed three attorneys representing plaintiffs. The School Committee also objected to the appointments of a third master and one expert because each had in the past supported the NAACP, which had provided financial assistance and counsel to the plaintiffs. The issue arising from the overruling of these objections will be dealt with later in this opinion. So also will be a further objection of the School Committee to the court’s order awarding compensation to the masters. [7] According to this plan, various options were to be considered to deal with oversubscribed schools. These included the use of temporary classrooms, and extension of the school year to increase capacity. The use of temporary ' classrooms to enable white schools to operate in excess of their capacity was found to be one of the bases of liability in this action. Morgan v. Hennigan, supra, 379 F.Supp. at 427-28. [8] Accord, Bivins v. Bibb County Board of Education, 424 F.2d 97 (5th Cir. 1970); United States v. Board of Education of Baldwin County, 423 F.2d 1013 (5th Cir. 1970); United States v. Hinds County School Board, 417 F.2d 852 (5th Cir. 1969); United States v. Jefferson County Board of Education, 417 F.2d 834 (5th Cir. 1969); Hall v. St. Helena Parish School *410 Board, 417 F.2d 801 (5th Cir. 1969); Jackson v. Marvell School District No. 22, 416 F.2d 380 (8th Cir. 1969); United States v. Lovett, 416 F.2d 386 (8th Cir. 1969); Anthony v. Marshall County Board of Education, 409 F.2d 1287 (5th Cir. 1969); United States v. Greenwood Municipal Separate School District, 406 F.2d 1086 (5th Cir. 1969). [9] The Trotter School was designed to offer cultural enrichment programs for children from AFDC families. White students residing in AFDC homes from outside the city of Boston made up part of the student body. 379 F.Supp. at 430. [10] The masters found: “The magnet concept as devised by the Committee is unrealistic and unworkable. Magnet programs could assist desegregation if they satisfied certain conditions. They must be limited in number, and they must be carefully placed, so that the effect of the applications they attract is to promote desegregation. The Committee Plan, which proposes the introduction of many magnet programs, satisfies none of these conditions, and could not possibly be put into operation by September, 1975.” [11] Comments of a member, and of the then Chairman, are revealing: “The plan is ‘pie in the sky.’ It is a contradiction, it is impossible. All of us would love to see a voluntary desegregation system put into effect. It is not a practical reality. Of course I will vote for this.” “I would never vote for a plan that involved the busing of school children. It is unfortunate that is the way our society exists . but the only way you are going to desegregate city schools is through forced busing.” *411 The only member of the School Committee to refuse to vote for the plan stated: “I agree that it would be an ideal solution and therefore will not vote against it, but I will vote ‘present.’ I believe we should have our votes consistent with what we feel is a reality.” Remarks of Member Sullivan, Emergency Meeting of the Boston School Comm., January 7, 1975. [12] One provision, § 1714, facially proscribes any court ordered transportation of a student to a school other than one “closest or next closest to his place of residence” of the appropriate grade level and type of education. The Mayor properly concedes that this section must be read in conjunction with § 1702(b) which states that provisions of the Act “are not intended to modify or diminish the authority of the courts ... to enforce fully . the Constitution. . . . ” [13] The Association also argues that the adoption of the masters’ plan was required by F.R.Civ.P. 53(e)(2), on the theory that the “delineation of the districts was a factual determination” which could not be changed by the district court unless found to be clearly erroneous. This argument misconceives the type of reference made in this case. Rule 53(c) provides that “[t]he order of reference to the master may specify or limit his powers”, and that was done in this case. The Order of Appointment and Reference to Masters of February 7, 1975, specified that the masters were “to conduct hearings and make recommendations for a desegregation plan for Boston public schools together with the reasons for recommending that plan, including discussion of the key issues.” This clearly described a mission not involving findings of fact meriting deferential treatment under Rule 53(e)(2). That rule’s mandate simply “does not . . . apply where the master is directed only to report the evidence [and] to make recommendations . . .. ” 5A J. Moore, Federal Practice ¶ 53.12[2], at 3002; Matter of Van Swearingen Corp., 180 F.2d 119 (6th Cir. 1950). See Hart v. Community School Board, 383 F.Supp. 699 (E.D.N.Y. 1974), aff'd, 512 F.2d 37 (2d Cir. 1975). This reference was not a substitute for trial, where the master to a large extent takes over the fact-finding process; the district judge could not delegate his duty to evaluate for himself what actions had to be undertaken in order to remedy past failure to comply with the constitution. Indeed, we doubt that the determination of district boundaries is accurately described as purely factual in nature, given the nature of this case. It is uncontested, moreover, that the figures on which the masters relied in drawing their districts were outdated and invalid in light of new figures submitted by the School Department shortly after the masters’ final report was filed. Both as to racial composition and numbers of students in the districts, the new figures rendered the factual assumptions un- *412 derlying the masters’ determination clearly erroneous. [14] The estimate of 6,100 additional students who would be bused under the court’s plan (21,000-court minus 14,900-masters) ignores the impact of the new data, even had the masters’ districts remained essentially intact. For example, the Burke district, eliminated by the court, had a capacity of 8,250 and an enrollment projected by the masters of 7,590. But the new statistics revealed it as containing 11,-620 students. Even after assuming that some of the overflow would be absorbed by citywide magnet schools, there would remain perhaps 2,000 excess students. The Mayor does not show that shifts in district lines to accommodate these students either by merging contiguous neighborhoods and their schools into Burke or by removing fringe neighborhoods from it would not encounter the difficulty of increasing Burke’s percentage of black students, already a high 63%. Although the amount of additional busing that would have been needed to take care of this situation is a matter of conjecture, it would appear to be substantial, if increased segregation were to be avoided. Not only student population but racial concentration estimates of the masters were at odds with the realities of the new figures. The following discrepancies were found to exist: Master's New figures applied by court to Possible result of allowed deviation of + 10 percentage final masters' points within report districts each school West Roxbury 80% W 93% W 100% W South Boston 60% W 67% W 77% W Burke 50% B 63% B 73% B The result, particularly when the ± 10 point deviation is considered, represented substantial changes, unanticipated by the masters. While the masters gave greater weight to neighborhood identity than did the court, it is far from clear that they would have tolerated the higher concentrations shown by the new data. The Mayor’s assumption that the court’s plan would involve anything close to the mandatory busing of 6,100 more students is therefore significantly vulnerable. [15] As the court said in Brinkman v. Gilligan, 518 F.2d 853, 856 (6th Cir. 1975), “We construe the 1974 Act, read as a whole, as not limiting either the nature or the scope of the remedy for constitutional violations in the instant case.” [16] Additional comments in the same vein are to be found in various parts of the court’s opinion filed subsequent to the plan itself: “The plan that the court has ordered into effect reflects the court’s continuing efforts to hold compulsory busing to a minimum.” “The districts in this plan and the guidelines for assigning students have been drawn to minimize required transportation as much as possible consistently with desegregating the city’s schools.” “The plan adopted by the court attempts to minimize forced busing.” [17] The Mayor would have us answer this question in the negative by referring to other cases, affirmed by courts of appeals, where more schools were allowed to remain all black or more dominantly black than the masters’ plan contemplated. Apart from the Mayor’s assumption that the masters’ plan contemplated no school more than about 53% black — an estimate which, in the light of the new data, should be closer to 73% — we reject such a simplistic color matching test to determine the constitutional sufficiency of any plan. The cases cited by the Mayor reveal the hazard of picking as a control any one statistic of residual racial imbalance which another court may have approved. In three cases, Mims v. Duval County School Board, 329 F.Supp. 123 (M.D.Fla.), aff’d, 447 F.2d 1330 (5th Cir. 1971); Goss v. Board of Education, 482 F.2d 1044 (6th Cir. 1973) (en banc); and Northcross v. Board of Education, 489 F.2d 15 (6th Cir. 1973), geographic factors and, at least in Mims and Northcross, the desire to avoid extensive long distance busing were relied upon to justify the lesser degree of desegregation attempted. In two cases, Pate v. Dade County School Board, 434 F.2d 1151 (5th Cir. 1970) and Ross v. Eckels, 434 F.2d 1140 (5th Cir. 1970), while more all black schools were permitted to remain, the courts of appeals had insisted on reducing the number of pupils attending such schools to around 5% of the total school population. Were we to take this statistic as a control, we would have to declare the court’s plan, which leaves over 7% of the total student population attending the East Boston white schools, constitutionally inadequate. The remaining case, Carr v. Montgomery Board of Education, 377 F.Supp. 1123 (M.D.Ala.1974), aff’d per curiam, 511 F.2d 1374 (5th Cir. 1975), concededly insisted on a less stringent plan. Whether the desire to avoid cross-city busing would be, in our minds, sufficient to justify the lower objective, we have no reason to decide. The exercise of one district court’s discretion, in a particular case, though affirmed on appeal, cannot establish constitutional limits for other courts facing other circumstances. [18] The district court also ruled that modification was not required by the Equal Educational Opportunities Act of 1974. See note 24 infra. [19] A remedy may sometimes properly “exceed the violation” in that it may do more than eradicate the constitutional wrong. To the extent that “overbreadth” in the remedy is necessary to ensure that the constitutional violation is corrected, it is not at all unusual. There are many instances in the law in which remedial law places greater restrictions on primary activity than did the substantive law that had been violated. For example, a company that is found to have violated the Sherman Act will have its future operations governed by a much more restrictive standard than that imposed by the Sherman Act itself. See United States v. United Shoe Machinery Corp., 391 U.S. 244, 88 S.Ct. 1496, 20 L.Ed.2d 562 (1968). [20] The Association recognizes that Swann, Davis, and Green provide that, during the remedial phase of a school desegregation case, the district court must order the maximum practicable desegregation regardless of the degree to which the actual segregation in the schools is demonstrably the result of unlawful state action. The Association attempts to distinguish *417 these cases on the ground that each involved school systems that had been segregated by statute for years. In such school systems, the Association contends that the application of these remedial principles was warranted because all segregation could be presumed to be the result of illegal official action. The Association’s attempt to limit the applicability of these remedial principals to cases in which there had previously been a statutory dual school system fails. In Swann and Davis, the Court clearly did not proceed on the assumption that the application of its remedial principles would operate only to eliminate the segregation which was directly attributable to illegal official action. The Court recognized that the segregation in those systems that remained after the school authorities abolished the statutory dual system and adopted a “neighborhood” school policy was, to some extent a consequence of private residential patterns. See Swann, supra, 402 U.S. at 25-26, 91 S.Ct. 1267; Davis, supra, 402 U.S. at 36, 91 S.Ct. 1289. The Court clearly provided that although all the remaining segregation in certain schools may not be attributable to illegal state action, the schools must be subjected to the maximum practicable desegregation. Swann, supra, 402 U.S. at 21 and 28, 91 S.Ct. 1267. The lower federal courts have consistently rejected desegregation plans that attempted to justify the failure to desegregate certain schools on the ground that the racial composition of those schools results from housing patterns. Maximum feasible desegregation is required unless the school authorities can demonstrate that their actions in no way contributed to the segregated character of the individual schools. See, e. g., Lee v. Macon County Board of Education, 448 F.2d 746 (5th Cir. 1971); Goss v. Board of Education of Knoxville, 443 F.2d 632 (6th Cir. 1971); Clark v. Board of Education of Little Rock, 465 F.2d 1044 (8th Cir. 1972); Brewer v. Board of Education of Norfolk, 397 F.2d 37 (4th Cir. 1960). [21] The Association places extensive reliance on the language in Swann that provides that school authorities may, at the remedy stage, attempt to limit the geographic scope of the remedy by proving that the racial composition of the schools in certain areas in the system is in no way the result of present or past discriminatory action on their part. Swann, supra, 402 U.S. at 26, 91 S.Ct. 1267. Swann, however, cannot be read as supporting the proposition that causation is at issue during the remedial phase of a school segregation case when there has been a trial on liability. The burden this language describes is essentially identical to that which the presence of substantial intentional segregation shifts to the School Committee during the trial on liability. See Keys, supra, 413 U.S. at 203-05, and 210-11, 93 S.Ct. 2686. In Swann, the school authorities were afforded the opportunity to satisfy this burden at the remedy stage of the proceedings because here had been no trial on liability. Here, of course, the school authorities have had their chance. [22] For example, although a dominantly white school in a white neighborhood may appear to be wholly the result of the ethnic patterns of the neighborhood, when the School Committee has taken steps to incorporate the residential segregation into the schools, to earmark that school as a white school, and to permit white students from other parts of the city to attend it, that school has played a major role in skewing the racial profiles of the other schools in the system. [23] The Association places considerable reliance on Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974), and, in particular, on its statement that “the remedy [in a school desegregation case] is necessarily designed, as are all remedies, to restore the victims of discriminatory conduct to the position they would have occupied in the absence of such conduct.” Id. at 746, 94 S.Ct. at 3128. Although we understand why the Association believes this language supports its position, we read this language as entirely consistent with the continued application of the remedial principles of Swann and other such cases. As we have demonstrated, restoring the victims of unconstitutional segregation requires far more than eliminating the specific, demonstrable effects of the proven discriminatory acts. Restoration is necessarily a complex and widespread process. [24] The Association also contends that the Equal Educational Opportunities Act of 1974 required the district court to adopt its remedial theory. In particular, the Association points to § 213 of the Act, which provides that “in formulating a remedy for ... a denial of equal protection of the laws, a court . shall seek or impose only such remedies as are essential to correct particular denials of equal protection of the laws.” 20 U.S.C. § 1712. This language, like the statement from Milliken v. Bradley, 418 U.S. 717, 746, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974), discussed in note 23 supra, is entirely consistent with the continued application of the remedial principles of Swann, Davis, and Green. Because of the pervasive effects of the School Committee’s actions, maximum practicable desegregation is “essential to correct [the] particular denials of equal protection [that occurred].” See note 23 supra. [25] At oral argument, the Association argued that the Tenth Circuit’s recent decision on remand in Keyes v. School District No. 1, 521 F.2d 465 (10th Cir. 1975) (Keys II), supports its contention that, even after a trial on liability, the district court is under an obligation to receive evidence on the discriminatory effects of the proven violations as part of its task in fashioning a remedy. Keyes II in no way supports this proposition. The evidence the district court admitted in Keyes II pertained to the question of liability, not to remedy. Id. at 471-73. [26] The Association suggests that it should be permitted to attempt to rebut the Keyes presumptions now because the defendant School Committee was not afforded an opportunity to do so at the liability stage of proceedings since Keyes had not been decided at that time. We find this suggestion utterly without merit. The effect of Keyes was before the district court during the trial on liability, it having been decided on June 21, 1973, a time when that trial had been reopened. The School Committee did not seek a reopening on the issues presented by Keyes; it described Keyes as a restatement of the principles of the earlier Supreme Court decisions. In any event, the district court found that the record was sufficiently complete to apply Keyes, 379 F.Supp. at 479, and that finding was not challenged on appeal. [27] In the alternative, it is claimed that the district court should have taken white flight into account in modifying the masters’ plan. See Part A 2, supra. The masters, however, rejected considerations of white flight in drafting their plan. [28] “But it should go without saying that the vitality of these constitutional principles cannot be allowed to yield simply because of disagreement with them.” 349 U.S. at 300, 75 S.Ct. at 756. [29] Subsequent to the district court decision, we allowed to be filed, subject to a determination of relevance, voluminous affidavits and other materials by social scientists on the subject of white flight. The admissibility of these submissions have been attacked as being outside the record, F.R. A.P. 10(a), and as hearsay. The materials have been defended as sociological data in the nature of “legislative facts” relevant to a determination of the law governing the district court in this matter. For reasons we discuss in the text, we reject all these materials as irrelevant to the issues before us on this appeal. We include a brief synopsis of these materials to illustrate the difficulty in evaluating white flight if it were relevant. The data submitted initially by the Association consists of an affidavit and report prepared by James S. Coleman, Professor of Sociology at the University of Chicago. Dr. Coleman states that his recent study shows that while there is decreasing segregation within school districts, segregation between school districts in the same metropolitan area is increasing; that rapid increase in loss of white children from central city schools follows immediately after school desegregation; and that as a consequence, desegregation has not significantly raised the levels of academic achievement of blacks. Plaintiffs counter with the transcript of testimony given by Jane R. Mercer, Associate Professor of Sociology, University of California at Riverside, in a case involving Indianapolis. Her testimony shows that she studied desegregation in school districts throughout the state of California; that white exit to private schools is a short term phenomenon; and that declines in white population in the cities are part of a long term demographic trend independent of desegregation. The plaintiffs also submit a paper written by Meyer Weinberg, Editor of Integrated Education magazine, which summarizes other studies showing white flight to be an avoidable phenomenon, not an inevitable consequence of mandatory desegregation. Weinberg’s paper and another paper written by Professors Green of Michigan State University and Pettigrew of Harvard University, criticize Dr. Coleman’s methodology, claiming that the source of his raw data is unknown; that failure to evaluate large cities which have been subject to massive desegregation orders separately from large cities which have not been subject to court orders undermines the study’s relevancy; and that failure to control for other variables, which may be correlated with white flight, jeopardizes the validity of his conclusions. Dr. Coleman replies in an affidavit filed with a copy of a working paper, “Trends in School Segregation, 1968-73”. In his affidavit, Dr. Coleman states that the report previously filed with the court was prepared for oral delivery and was based on the attached working paper. He defends his methodology, states that his conclusions are consistent with the findings of the studies cited by Weinberg, and the studies conducted by Dr. Mercer. Dr. Coleman states that his study shows that massive white flight will occur when there is a significant decrease in segregation in a city where there is a high proportion of blacks in the central city and suburbs of a significantly different racial composition. Plaintiffs, in rebuttal, file another study prepared by Christine H. Rossell, of Boston University. The Rossell study, prepared from data on 86 northern school districts subject to *421 court ordered or legislatively enacted school desegregation, suggests that white flight is minimal and a temporary reaction to school desegregation. Plaintiffs claim that Dr. Ros-sell’s study differs from Dr. Coleman’s in that Dr. Rossell deals only with northern school districts subject to desegregation plans while Dr. Coleman does not distinguish between forms of desegregation. In the final submission, Dr. Coleman defends his study against the Rossell findings, suggesting that her analysis is inadequate to examine the effects of desegregation on a core city school system. He further claims that his model has proved accurately predictive of the Boston experience. Throughout this series of submissions this court has been burdened with reports written for sociologists by sociologists utilizing sophisticated statistical and mathematical techniques. We lack the expertise to evaluate these studies on their merits. We do come to one conclusion, however. The relationship between white flight and court ordered desegregation is a matter of heated debate among experts in sociology, and a firm professional consensus has not yet emerged. Appellants have also filed with this court copies of the Boston School Department’s current census of students according to race and minority group. The figures facially suggest loss of a significant number of white enrollees. We note, however, that the district court is currently studying the accuracy of past enrollment data. We decline to make any conclusions concerning the existence or nonexistence of white flight on the sparse figures available to us. [30] Appellants suggested approach would necessarily involve the district court in something like the following analysis: (1) take evidence concerning the prospects of white flight under the various plans proposed; (2) exclude the causes of such flight attributable to any historic trend, or such factors as overcrowding, transportation difficulties, deteriorated housing, taxes, crime, pollution, industrial migration, etc.; (3) make a judgment as to the effect which different levels of desegregation would have on white flight; and (4) select or devise that plan which will incorporate enough desegregation, to bring about the maximum amount of inter-racial contact in the schools after taking account of the white flight such desegregation would be expected to induce. The experts have difficulty in attempting to justify conclusions as to the effect of past desegregation plans on white flight, see note 29 supra; the task of making estimates of expected exodus of whites attributable to varying future desegregation plans would seem to be more difficult. Conceivably, public attitude sampling could be undertaken, using various hypotheses. This might involve questioning parents in a particular section or school district whether they would be likely to place their child elsewhere, or move, if the child were to attend a school which was x, y, or z percent black. The possibility is a real one that surveys would indicate that the prospect of any substantial amount of desegregation or busing would provoke sufficient expressions of intent to flee as to negate any desegregation plan. Alternatively, if expressed intentions were to be heavily discounted, their utility would accordingly diminish. [31] Rough equality between districts was desirable to allow students from all the districts equal access to the citywide magnet schools without adversely affecting desegregation of the districts. The district court divided the school system into eight community districts and a ninth citywide district. The overall racial composition of the system for the 1975-76 school year was projected at 51% white (W), 49% black and other minority (B&OM). The projected enrollments in the seven community districts which would be desegregated ranged in racial composition from 61% W — 39% B&OM to 40% W — 60% B&OM. The eighth district, East Boston, was left predominantly white because of its isolated geographic location. [32] In district 4, with a projected racial composition of 61% W — 39% B&OM, the permitted variance of the composition of the schools is 70% W — 30% B&OM; 50% W — 50% B&OM. Because the percentage of black and other minority students cannot fall below 30% (39%— 25% (39) = 30%), the percentage of white students cannot rise above 70%. The same observation, in reverse, pertains to district 7, which is 40% W — 60% B&OM. [33] Since citywide project composition was 51% W — 49% B&OM, magnet schools could permissibly enroll 56-46% W, 44-54% B&OM. The bilingual, bi-cultural magnet, Hernandez, was excepted from this requirement. [34] The entering classes normally enroll 6-8% Asian-Americans. When these numbers are added to the black and Hispanic students, the percentage of the entering class which is characterized as black and other minority is in the range of 41 — 43%. As school officials testified that it would be difficult to find sufficient black and Hispanic students to satisfy the minimum figures, the district court apparently did not consider it necessary to place an upper limit on minority enrollment. [35] This fact also disposes of the Association claim that imposing a racial preference at the examination schools is unconstitutional for those reasons expressed in Justice Douglas’ dissent in DeFunis v. Odegaard, 416 U.S. 312, 320, 94 S.Ct. 1704, 40 L.Ed.2d 164 (1974). Whatever the constitutionality of racial preferences in the absence of past unlawful discrimination, they are a basic tool in remedying constitutional violations. See Castro v. Beecher, 459 F.2d 725, 737 (1st Cir. 1972). [36] The old method of selecting admittees to the schools was mandated by a consent decree between the School Committee and the Massachusetts Commission Against Discrimination (MCAD). This decree introduced the SSAT and admission was based solely on the scores of that test. As the plaintiffs were not parties to the MCAD litigation, the district court was not bound by the consent decree. An alternative plan, proposed by the alumni associations of the affected schools, would have had 65% of admission based on the SSAT scores alone, the remaining 35% to be chosen in racial proportions but setting a score in the 50th percentile on the SSAT as a floor for admissions. The masters’ plan allowed for grade point averages as well as SSAT scores to be used, but adopted the use of a SSAT score floor. [37] Appellants claim that only 25% of admit-tees come from advanced work classes. While it is obvious that the examination schools cannot be completely desegregated by the advanced work classes, desegregation of the classes is a positive step. [38] The district court rested its decision on the qualifications of Master McCormack and Expert Dentler on the alternative ground that the School Committee’s objections were not timely filed. The objection to the qualifications of these individuals took the form of a motion to expunge filed on March 4, 1975, more than a week after the masters’ hearings concluded and almost a month after the February 6 deadline for filing objections. Since we hold that the district court did not err in its ruling on the merits, we need not reach this point. [39] We need not express an opinion on the timeliness of this objection in the light of our holding on the merits. [40] On appeal, the School Committee suggests that it was improper for the district court, in considering the School Committee’s objections, to rely upon interviews with Masters Keppel and Willie and upon representations of plaintiffs’ counsel. It is clear from the record, however, that the School Committee acquiesced in this method of investigation. [41] The Harvard Graduate School of Education has, on several occasions, been hired by the School Committee to conduct surveys. Expert Dentler, in his capacity of Dean of Boston University’s School of Education, has doné work for the Committee. Thus, three of the four individuals challenged have benefitted financially from their relationships with the School Committee. [42] The experts and masters felt this number would so saturate the city as to deny the magnet schools a desegregative effect, their raison d’etre. See note 10, supra. [43] The court’s plan left intact the programs of those citywide schools which had established themselves over the years as magnets, and it apparently adopted a number of the magnet programs as proposed in the Committee plan. The Committee’s appeal is addressed only to five or six instances where it is felt that the court’s plan is inconsistent with its own proposals. [44] The city defendants, charged with paying the costs, have not objected to these councils.