I. DAYTON MOTION
On June 29,1978,1 directed the parties to submit proposed findings to the court pursuant to Dayton Board of Education v. Brinkman, 433 U.S. 406, 97 S.Ct. 2766, 53 L.Ed.2d 851 (1977), setting forth “the extent of the incremental effects resulting from the acts found in this court’s April 30, 1976 decision to have been constitutional violations.” Since the Board’s submission in response to this order was inadequate, on September 5, 1978, I directed the Board to prepare a supplementary response. All papers have been submitted, and oral argument on the question was held on November 17, 1978. I am now prepared to rule on the question.
In determining whether the constitutional violations had a systemwide impact, I am not reconsidering questions of liability; these questions were determined in the April 30, 1976 liability decision, Arthur v. Nyquist, 415 F.Supp. 904 (W.D.N.Y.1976), reconsidered in the decision of March 1, 1977, Arthur v. Nyquist, 429 F.Supp. 206 (W.D.N.Y.1977), and affirmed by the Second Circuit on March 8, 1978, Arthur v. Nyquist, 573 F.2d 134 (2d Cir. 1978). Certiorari was denied by the Supreme Court on October 3, 1978, Manch v. Arthur, 439 U.S. 860, 99 S.Ct. 179, 58 L.Ed.2d 169 (1978).' Nor am I addressing whether actions of the defendants taking place after the initial finding of liability constitute additional constitutional violations. Dayton only requires the court to decide whether the segregative acts, as already found by the court, had a sufficiently widespread impact on the school system to justify the imposition of a systemwide remedy.
Dayton does not require a finding that each and every school or student in the system was directly affected by the defendants’ actions as a prerequisite to ordering a systemwide remedy.1 The test is whether there has been a “systemwide impact.” Thus, in Reed v. Rhodes, 455 F.Supp. 546 (N.D.Ohio 1978), the court found that over 200 separate instances of segregative acts affecting at least 60% of the schools and all of the staff in the Cleveland system had a systemwide impact under Dayton.
Dayton reaffirmed the well-established equitable principle requiring court-imposed remedies to be tailored to the scope of the violation. If the defendants’ acts have had only isolated effects on discrete portions of the school system, then the remedy must be designed to correct the isolated effects only. By the same token, if the incremental segregative effect of the violations is system-wide, the court must fashion a systemwide remedy. In either case, the remedy must be “designed as nearly as possible ‘to restore the victims of discriminatory conduct to the position they would have occupied in the absence of such conduct.’ ” Milliken v. Bradley, 433 U.S. 267, 280, 97 S.Ct. 2749, 2757, 53 L.Ed.2d 745 (1977) (Milliken II), quoting Milliken v. Bradley, 418 U.S. 717, 746, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974) (Milliken I).
The Buffalo public schools currently have in place a partial remedy, consisting of two components: QIE and magnet schools. *833Apart from deficiencies in these programs (see the court’s order of June 29, 1978 and the discussion, infra), the present remedy leaves untouched many all-minority schools. In the context of this case, resolution of the Dayton question determines whether additional desegregation is required in order to complete the remedy begun in 1976.
I have reviewed the record in this case, including the stipulation of facts, the liability decision, the Second Circuit’s decision, and the legal briefs submitted on this question to the court. For the reasons stated below, I find that the impact of the constitutional violations was systemwide under the standards announced in Dayton.
The liability decision of April 30,1976 and the reconsideration decision of March 1, 1977 recite in great detail the evidence of deliberate segregation relied upon by the court in finding constitutional violations, and it would serve no useful purpose to reiterate those findings here. As the Second Circuit stated in affirming these findings, “[w]e are unable to imagine a set of facts, short of a public admission of wrongdoing, which would be more suggestive of intentional discrimination.” 573 F.2d at 145.
Unlike the circumstances in Dayton, this is not a case which involved only a few isolated acts of discrimination or only a few schools or areas in the City. The majority of schools in the Buffalo system were directly affected by the defendants’ acts and policies. Even those schools which were not specifically discussed in the liability decision were indirectly affected by the broadranging and longstanding discriminatory policies of the defendants, because these policies contributed to the public’s tendency to identify certain schools and certain neighborhoods as minority and other schools and neighborhoods as non-minority. In addition, many of the discriminatory policies, such as staff recruiting and assignment, were necessarily applied on a systemwide basis. Even those which were confined to a limited number of schools had indirect effects on other schools because changes in the composition in part of the system necessarily affected the rest of the system.
In light of the Second Circuit’s decision, the court’s extensive findings in prior decisions, and the position taken by the Board at oral argument, the court need not set forth in detail the reasons for the court’s findings of systemwide impact. For purposes of completeness, however, a brief summary follows.
The first constitutional violation discussed in the April 30,1976 liability decision involved East High School. By redistricting the attendance zone on a number of occasions beginning as early as 1954, the defendants increased and maintained racial segregation at East. 415 F.Supp. at 922-24. In addition, by permitting large numbers of white students to transfer out of the East district to take foreign languages only offered at other high schools, the defendants encouraged concentrations of minorities in one out of seven academic high schools. Id. at 926-30. This manipulation of the racial composition at East had an impact on the predominately majority high schools by preserving and in some cases increasing their majority enrollments.
A second major violation involved transfer policies. Over a period of years, the Board permitted numerous transfers of white students out of predominately black schools for specious or blatantly discriminatory reasons. In addition, certain principals unofficially allowed white students to transfer out of black school districts. The opinion cites transfers involving schools 4, 16, 30, 34, and 38 merely to illustrate the impact that the policy had on particular schools, but the transfers occurred throughout the system. The combined effect of official and unofficial transfers, including language transfers at East, was an annual total of 2,000-4,000 white students avoiding their neighborhood schools and thereby contributing to segregation. To make matters worse, transfers in the primary grades had potentially double and triple segregative ramifications because of feeder patterns to junior and senior high schools. Id. at 936--39.
*834Another policy which was found to be a constitutional violation was the use of optional attendance zones, which enabled white students to avoid schools they normally would have attended in favor of predominantly white schools. The liability decision specifically singles out five optional districts involving at least ten schools. But the policy was also applied throughout the system because all of the district lines were optional areas. Id. at 939-41. The Board’s transfer and optional areas policies “were substantial contributing factors to the segregation at all levels of the BPSS, and . this segregative effect was clearly foreseeable by the Board.” Id. at 941.
The districting of Woodlawn Junior High School was another constitutional violation, which the court found was done so as to guarantee that Woodlawn would open as an all-minority school. This not only affected students in the Woodlawn district but also had an impact on certain majority students who were able to avoid Woodlawn by virtue of its districting, optional zones, and transfers. Id. at 930-36.
Another constitutional violation involved discriminatory admission policies in four of the City’s six vocational-technical high schools. Id. at 941-43. Recognizing that “the Board does not act in a vacuum, and the actions it takes with respect to admissions at one vocational school will necessarily be felt at another,” the court found that these policies “caused segregated conditions to exist at a significant number” of the vocational schools. Id. at 943.
An additional violation involved discrimination by the Board against minorities in the hiring and assignment of school staff, both instructional and noninstructional. Id. at 943-48. This finding clearly has system-wide implications because the policies affected the minority staff composition at every public school in Buffalo.
Finally, constitutional violations were found in the general pattern of obstruction which characterized the actions of the Board and the Common Council over a number of years. These actions also had a systemwide impact because they pertained not just to individual schools but to the school system as a whole.
In summary, the record demonstrates the systemwide impact of the violations. Specific violations directly affected virtually all of the academic and vocational high schools, the junior high and middle schools, a substantial number of the elementary schools, and all of the school staff. A number of the violations concerned systemwide policies. Those which did not indirectly affected many of the other schools in the system by causing certain schools to be identified as black and others as white.
Although the defendants’ supplemental papers argue that the impact of the violations was not systemwide, at oral argument defendants’ counsel took the position that the question of systemwide impact had been resolved by the Second Circuit and that the court should now develop a final remedy order. Defendants’ counsel stated:
We have said that we believe that the Second Circuit indeed has said to this Court that you are not going to have difficulty finding systemwide impact and we have already appealed to the United States Supreme Court, or requested a writ which has been denied, and so it serves no purpose for anyone here to belabor the kinds of things that plaintiffs went into for an hour or so, as we see it.
Transcript of Nov. 17, 1978, at 71. See also pp. 56-59. It is clear that the defendants do not seriously challenge the systemwide impact of the violations in this case. The court therefore has not undertaken a more detailed analysis of the Dayton question.
Since the impact of the constitutional violations was systemwide, a systemwide remedy must be fashioned which to the greatest extent possible restores the victims of the discrimination to the position they would have had in the absence of such violations. Milliken I, at 746, 94 S.Ct. 3112, supra. The ramifications of this finding for the Buffalo schools are the subject of the next section of this decision.
*835II. ADDITIONAL DESEGREGATION: ELEMENTARY SCHOOLS
The remedy currently in place is the “Buffalo Plan” (Ex. 181), filed by the defendants on January 5,1977 and adopted by the court on May 4, 1977 after extensive court hearings. The central features of the plan were the establishment of the magnet schools and the expansion of the QIE program. Although it made substantial changes in the composition of the schools, the Buffalo Plan is not a systemwide remedy because it leaves intact at least fifteen all-minority schools.2 I find that additional steps must be taken in order to desegregate these fifteen schools to the greatest extent practicable. The legal principles and other considérations which lead me to this conclusion are set forth below.
Throughout the remedy phase of this lawsuit, the court has stated that the Buffalo Plan was incomplete and has urged the Board to reduce the number of all-minority schools.3 The defendants maintain that the amount of desegregation already achieved is sufficient, given the practicalities involved, to pass constitutional muster. In support of their position, they cite enrollment statistics showing a significant decline in the number of racially isolated elementary students in the BPSS since implementation of the remedy. In 1975-76, 26,173 elementary students attended racially isolated schools. Two years later, in 1977-78, this number by their calculations dropped to 7,845 students. (Ex. 276 at 53.) In 1978-79, it dropped to 6,530 students. (Ex. 356.) They therefore argue that the Buffalo Plan has in large degree succeeded and that further planning should be left to the Board.
But these statistics fail to reveal that the reduction in students attending one-race schools is largely due to the elimination of all-majority schools. According to the plaintiffs’ figures, in 1975-76, 10,131 minority students attended all-minority elementary schools. In 1977-78, this number dropped to 8,500 minority students. (Ex. 279 at 5-6.) This figure is larger than defendants’ figures because it includes several all-minority schools deleted by defendants.4 Even in the desegregated schools, a large number of students remain in segregated grades or classes. (Ex. 279 at 6-8). Fifteen elementary schools remain segregated with no showing by the Board on a school-by-school basis that it would be impracticable to desegregate them.
In Swann v. Board of Education, 402 U.S. 1, 26, 91 S.Ct. 1267, 1281, 28 L.Ed.2d 554 (1971), the Supreme Court stated that a remedial plan “should make every effort to achieve the greatest possible degree of actual desegregation and will thus necessarily be concerned with the elimination of one-race schools.” Although the Court refused to adopt a per se rule against one-race schools, it did establish a presumption against schools substantially disproportionate in their racial composition and directed school authorities to justify the continued existence of any one-race schools in any remedial plan. The Court stated:
Where the school authority’s proposed plan for conversion from a dual to a unitary system contemplates the continued existence of some schools that are all or predominately of one race, they have the burden of showing that such school assignments are genuinely nondiscrimina*836tory. The court should scrutinize such schools, and the burden upon the school authorities will be to satisfy the court that their racial composition is not the result of present or past discriminatory action on their part.
Id.
Contrary to the Board’s suggestion, under Swann it is not enough simply to eliminate identifiably white schools. For instance, in Bradley v. Milliken, 540 F.2d 229, 236-39 (6th Cir. 1976), aff’d on other grounds, 433 U.S. 267, 97 S.Ct. 2749, 53 L.Ed.2d 745 (1977), the district court’s remedy order left untouched three of the eight regions in the Detroit school system, which contained approximately 83,000 students, over 90% of whom were black. The district court found that desegregation of the all-white schools was sufficient under the circumstances. The Sixth Circuit reversed because the lower court “did not subject the exclusion of these three regions to the close scrutiny required by Swann ” but merely concluded that any attempt to desegregate these regions would be futile in light of the fact that the Detroit schools were over 75% minority. 540 F.2d at 238. The court stated:
This perfunctory treatment of the inner city falls far short of the “root and branch” requirements of Green v. County School Board, 391 U.S. 430, 437-38, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968), and the “all-out desegregation” requirements of Keyes v. School District, 413 U.S. 189, 214, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973).
Id.
One-race schools can be retained under Swann only if the Board can show that physical barriers, insuperable distances, demographic obstacles or other practicalities prevent further desegregation and that the present remedy represents the maximum desegregation practically achievable. United States v. DeSoto Parish School Board, 574 F.2d 804, 815 (5th Cir. 1978); Bradley v. Milliken, 540 F.2d at 237; Allen v. Board of Public Instruction, 432 F.2d 362, 367 (5th Cir. 1970). It is not enough to generally allege these barriers. The board assumes a heavy burden of justifying in detail the retention of any one-race schools. Thus, in Milliken v. Bradley, supra, the court recognized that some one-race schools were unavoidable because of the overwhelming number (75%) of black students in Detroit but nevertheless found that the district court did not provide adequate justification for excluding the three regions.
In United States v. DeSoto Parish School Board, supra, the Fifth Circuit rejected a desegregation plan which left four of the five black schools 100% black because the board had not shown that no reasonable alternatives existed. In that case, there was no entrenched residential segregation and several of the all-black schools were located in close proximity to the all-white schools. In United States v. Seminole County School District, 553 F.2d 992 (5th Cir. 1977), the court refused to authorize the continued existence of only one all-black school after the five remaining all-black schools had been successfully integrated. In Ellis v. Board of Public Instruction, 465 F.2d 878, 880 (5th Cir. 1972), cert. denied, 410 U.S. 966, 93 S.Ct. 1438, 35 L.Ed.2d 700 (1973), the court found that three one-race schools in Orange County, Florida had to be desegregated even though integration had already taken place in the rest of a large system, involving 98 separate schools and over 85,000 students. In each of these cases, the board had failed to show that practical considerations prevented desegregation. See also Lemon v. Bossier Parish School Board, 566 F.2d 985, 988 (5th Cir. 1978); Keyes v. School District No. 1, 521 F.2d 465, 480 (10th Cir. 1975); Boykins v. Fairfield Board of Education, 457 F.2d 1091, 1095 (5th Cir. 1972).
In cases in which desegregation plans have retained one-race schools, evidence was introduced and specific findings were made as to practical barriers to further desegregation. For instance, in Stout v. Jefferson County Board of Education, 537 F.2d 800 (5th Cir. 1976), a plan was upheld which retained two all-black and one all-white schools, serving a total of 1100 stu*837dents, in a school district containing many thousands of pupils. The court found that transportation between the schools would require crossing a substantial chain of small mountains on steep and heavily traveled roads and that desegregation of these schools would jeopardize the racial balance already achieved at others. Another example is Carr v. Montgomery County Board of Education, 377 F.Supp. 1123 (M.D.Ala.1974), aff’d, 511 F.2d 1374 (5th Cir.), cert. denied, 423 U.S. 986, 96 S.Ct. 394, 46 L.Ed.2d 303 (1975), in which a plan was approved retaining ten all-black elementary schools located in the heart of a black residential area because the court found that cross-city bussing would involve too great of a distance to be practicable.
In this case, evidence has been received in prior proceedings about the practicalities of further integrating the all-minority elementary schools. This evidence was received in the spring of 1977 when the court was considering whether to implement the plaintiffs’ or the Board’s proposed plan and before the decision of the Supreme Court in Dayton. It was from that hearing that the QIE-Magnet concept was approved. Perhaps much of that evidence remains relevant to the question of how much further the all-minority schools should be integrated. On the other hand, there have been changes in the residential patterns and the school system itself since that time which may make some of this evidence inappropriate to consider at the present time. In view of the Dayton decision and my finding that a systemwide remedy is appropriate for the Buffalo Public School System, the record must be updated. The cases I have just discussed require the Board to show what practicalities prevent further integration of the all-minority elementary schools.
I am therefore directing the Board to draft a new remedy plan which to the greatest extent practicable will desegregate the all-minority schools. In doing this, the Board should consider the cases discussed in this section. If the Board’s position is that further steps to integrate the all-minority schools ought not to be taken, the burden is on the Board to show why these steps cannot be taken and it must compile evidence showing clearly the physical or other barriers which it claims prevent further integration.
III. PRESENT REMEDY: QIE
Before setting forth general guidelines to be followed in designing a new remedy plan, the QIE program must be examined to determine the extent of its success and whether it may continue to play a role in the new desegregation plan. A brief evaluation of the magnet schools will follow in a subsequent order.
When the court allowed the defendants to proceed with plans for expanding QIE two years ago, it did not appear, that QIE was an ideal desegregation device but I wished to give the Board an opportunity to design and implement its own desegregation programs with minimal judicial interference. The Board represented that the Buffalo Plan would eliminate segregated education by guaranteeing minority enrollments of 20% or more to the previously all-white schools and by using the ten magnet schools to attract majority students to inner-city schools. The plan was approved by the court because it promised to make significant progress toward desegregating the schools.
The expanded QIE program has now been in place for two years. Its contribution to desegregation has been significant: approximately three thousand minority children attend integrated schools as a result of QIE. However, for a number of reasons QIE does not provide a workable long-term remedy. As discussed in the previous section, QIE by design leaves a number of all-minority elementary schools intact. Other shortcomings of QIE can be summarized as follows. QIE places a heavy burden on minority children, which under applicable legal standards is neither desirable nor permissible. Its description as a “vol*838untary” desegregation program is highly misleading. Its complexity, both by design and as implemented, renders it unmanageable and inefficient and precludes intelligent planning. Because of the very nature of QIE, its complexity increases every year. Finally, QIE increases the underutilization of inner-city schools and fails to attract students in the lower grades.
Many of these shortcomings are matters of implementation which could not have been anticipated two years ago when the plan was approved. Others were inherent in its structure from the beginning. But regardless of the source of these difficulties, they are so numerous and serious that the QIE program must be discontinued. What follows is a description of QIE and a discussion of each of these additional shortcomings.
A. DESCRIPTION OF QIE
QIE is described as a “voluntary” transportation program for inner-city minority students in the elementary grades who desire to attend predominately majority elementary schools on the periphery of the City. Approximately 3,000 students in grades K-8 were enrolled in the program during the 1978-79 school year. (Ex. 356.) They came from eighteen sending schools 5 and were transported by bus to one of thirty-six receiving schools.6 37% of the students participating in QIE were in grades K-4, 21% in grades 5-6, 41% in grades 7-8, and 1% in grade 9. The QIE feeder pattern for 1978-79, as ordered by the court on August 8, 1978, is set forth below in Table I:
TABLE I
1978-79 QIE FEEDER PATTERN RECEIVER SENDING SCHOOLS
3 48, 53, 74, 39,12 (Metro)
(no new applicants)
11 23, 59, 62, 90, 39
18 43
(no new applicants)
19 8, 48, 23, 61, 74
21 23, 61, 59, 62, 74
26 31, 37, 39, 62, 48, 6, 74
27 39, 48, 62, 90, 59, 57, 37, 53,
61
28 62, 75, 4, 23, 61
29 31, 59, 62, 90, 12, 57
33 37, 48, 90, 62, 39
38 8, 37, 48, 53, 62
42 " 23, 61, 74, 53
43 31, 59, 39, 23, 90
44 12, 31, 39, 75, 90, 57, 62
45 8, 53, 74, 48, 59
51 23, 61, 59, 62, 74
56 45, 59, 61
(no new applicants)
60 23, 61, 62, 74, 59
63 61, 23, 8, 53, 48
64 23, 59, 61, 62, 74
65 23, 59, 61, 62, 74
66 23, 53, 59, 61, 62, 74
67 31, 39, 53, 62, 90, 75, 37
68 23, 61, 62, 74, 59
69 62, 39, 23, 37, 90, 9, 31, 75,12, 59
70 6, 59, 62, 75, 90
71 9, 23, 61, 62, 37, 74
72 23, 53, 39, 59, 62, 9, 8, 90, 75, 31
77 8, 37, 39, 48, 74, 61
80 61, 78, 82, 23, 74 (no new applicants)
81 23, 74, 59, 61, 82, 53, 39, 62
86 8, 23, 61, 59, 74
West Hertel 23, 39, 61, 62, 74, 59, 12
*839Students are recruited for QIE in the spring of each year by Board staff. Visits are made to each sending school during which the program is explained to the students and they are given brochures and application forms to take home to their parents. Interested parents then file applications and their children are enrolled in one of the receiving schools.
The number of QIE students and minority percentage at each QIE receiver, as well . as the number of sender schools, is set forth below in Table II, taken from Table I, the Board’s QIE Registration as of October 19, 1978, and the 1978-79 BEDS data:
TABLE II
NUMBER OF QIE STUDENTS AND PERCENTAGE MINORITY ENROLLMENT IN QIE RECEIVER _SCHOOLS IN 1978-79_
RECEIVING SCHOOL NUMBER OF SENDING SCHOOLS QIE STUDENTS MINORITY ENROLLMENT
3 5 10 60.9%
11 5 40 48.0%
18 1 1 39.1%
19 5 53 36.4%
21 5 119 45.4%
22 10 54.5%
26 7 119 29.6%
27 9 182 28.6%
28 5 64 29.9%
29 6 83 23.2%
33 5 21 32.5%
38 5 79 32.8%
42 4 28 49.8%
43 5 65 25.1%
44 7 46 38.5%
45 ’ 5 23 32.8%
51 5 94 31.5%
52 99 34.8%
56 3 3 56.0%
60 5 111 30.1%
63 5 58 46.8%
64 5 113 44.1%
65 5 100 28.5%
66 6 176 37.4%
67 7 132 31.6%
68 5 78 54.7%
69 10 136 36.3%
70 5 71 22.6%
71 6 155 42.4%
72 10 162 22.7%
77 6 24 43.0%
78 3 80.5%
80 5 11 50.1%
81 8 177 45.0%
86 5 63 33.7%
West Hertel 7 301 44.9%
B. BURDEN ON MINORITIES
One of the most serious deficiencies in QIE is that it places a heavy burden of desegregating the neighborhood elementary schools on minority students. In the QIE program, it is the minority students who *840leave their neighborhood school districts and are transported across the city to unfamiliar schools in unfamiliar neighborhoods. Majority students who do not attend magnet schools are only required to walk to their neighborhood schools where desegregated education is brought to them.
It is true that some majority students participate in the desegregation program by opting to attend one of the City’s ten magnet schools. But the magnet schools have special educational programs which are not available to the remaining elementary schools, whether they be QIE receivers or all-minority senders. As far as desegregating the regular neighborhood schools, the entire burden falls on QIE students. Moreover, majority students receive an integrated education whether they remain in their neighborhood school or enroll in a magnet school. In contrast, minority youngsters who wish to obtain an integrated education must travel to either a magnet or a QIE receiving school. It therefore is clear that majority students do not share equally in the burden of desegregating the elementary schools.
A number of courts have rejected proposed remedial plans calling for one-way bussing of minority students because such plans place a disproportionate burden on minority students. NAACP v. Lansing Board of Education, 559 F.2d 1042, 1052 (6th Cir. 1977); Higgins v. Board of Education, 508 F.2d 779, 793 (6th Cir. 1974); Brice v. Landis, 314 F.Supp. 974, 978 (N.D.Cal. 1969). Although mathematical equality is not required, the school authorities must make all reasonable attempts to equalize the inconveniences. Hart v. Community School Board of Education, 512 F.2d 37, 53-54 (2d Cir. 1975). The plan must be one which means “more than merely allowing black children into hitherto closed white schools.” Id.
In light of these legal principles, I find that QIE places an undue burden on minority students and therefore is of questionable validity as a long-term desegregation device. This consideration, when coupled with numerous other problems with the program discussed below, has prompted my decision to dismantle the QIE program.
C. INVOLUNTARY NATURE OF QIE
Although the Board continues to describe QIE as a “voluntary” transfer program which gives minority students an option to attend either their all-minority neighborhood schools or a QIE receiving school, this description is deceptive. As found by the court in its June 29, 1978 order, over one-third of the QIE students come from four school districts (# # 23, 61, 62, and Fillmore) which have no neighborhood school option after grades 5 or 6 because their school districts were eliminated when the Campus East and Build magnets were established at Fillmore and Genesee-Humboldt. (Ex. 279 at 5.) In addition, students from six additional minority school districts whose neighborhood schools do not continue through grade 8 (# # 12, 39, 48, 59, 75, and 90) are likewise forced to participate in QIE. (See court’s order of June 29,1978, at 26-28.) Thus, over 40% of the students enrolled in QIE have no neighborhood schools and therefore are required to enter the program. For these displaced students, QIE is in no sense a “voluntary” transfer program. To characterize it as such when almost half of its participants are not volunteers is misleading at best.7
The characterization of QIE as a voluntary transfer program is inaccurate in another respect. • In applying for QIE for the 1978-79 school year, few of the QIE parents actually chose the receiving school to which the child was sent, despite Board representations to the contrary.8 The application *841form for 1979-80 has been amended to provide a space for parents to indicate their preference of receiving schools. But it is too early to determine whether this has changed prior practice because the application process is not yet complete. The QIE brochure used for recruiting for next year continues to provide that “QIE students will be assigned to a QIE school by the Department of School Integration Staff according to space available and court ordered transportation patterns.” Moreover, it is simply unrealistic to expect primary students to convey to their parents sufficient information about the program to enable their parents to make informed choices among the many unfamiliar receiving schools.
D. COMPLEXITY OF QIE AND RELATED PROBLEMS
The complexity of the QIE feeder pattern and of its resulting transportation system creates insurmountable practical problems in administration and numerous inefficiencies in financing. During the 1978-79 school year, eighteen sending schools fed thirty-six receiving schools. Subject to some restrictions imposed by the court in an effort to simplify the program, students were allowed to attend any one of the receiving schools, regardless of the schools attended by other QIE students from the same sending school. Assignments were made by the central staff on a case-by-case basis with little regard for transportation burdens. Thus, during 1977-78, in many instances a receiving school would receive one or two students from a large majority of the sending schools, with an individual QIE student frequently being the only child in a given neighborhood to travel to a particular school. (Ex. 276.)9 Bus routes then had to be devised which would pick up students from many school districts for each receiving school. The result was a patchwork of criss-crossing and overlapping bus routes of a complexity difficult to imagine.10 This in turn has caused inordinately long bus rides. QIE students had one-way rides of over one hour on seven of the bus routes to receiving schools, over 45 minutes on 27 routes, and over 30 minutes on 47 of the routes. These riding times were frequently extended further during poor driving conditions due to inclement weather. (Ex. 303 at 235-242.)
In addition to long bus rides, this scheme is extraordinarily inefficient. Although precise figures on the cost of transporting QIE students are not readily available, it is safe to assume that bus routes of this complexity are much more costly than simpler, shorter transportation schemes. At a time in which the City of Buffalo and the Board of Education face severe budgetary deficits, such a system simply does not make sense.
In the order of June 29, 1978, the court criticized the program for its inefficiency and lengthy bus rides. In order to decrease these difficulties, the court ordered the Board to limit the number of sending schools feeding each receiver to five or less except where specifically authorized by the court.
Yet the program remains unmanageable in its second year. So many exceptions exist to the five sending school limitation that the exception has virtually swallowed the rule. Twelve of the receiving schools have more than five senders and two have as many as ten. (Table II.)
When QIE is examined from the perspective of the number of receiving schools serving each sending school, its complexity *842becomes even more apparent. Table III lists the receiving schools and grade levels for each sending school used by QIE recruiters this spring in informing parents which schools their children could attend if they opted for QIE. (See Ex. 363.)
TABLE III
SENDING SCHOOLS NUMBER OF RECEIVING SCHOOLS RECEIVING SCHOOLS AND GRADE LEVELS
# 8 4 19, 88, 72 (K-8 Schools), 86 (K-6 School)
69, 71, 72 (K-8 Schools) #9 3 #12 8 29, 69 (K-8 Schools), West Her-tel (7-9)
#23 15 19, 21, 28, 43, 66, 69, 71, 72, 81 (K-8 Schools) 42, 51, 60, 65, 86 (K-6 Schools), West Hertel (7-9)
#81 6 26, 29, 43, 67, 69, 72 (K-8 Schools)
#37 7 26, 27, 83, 38, 67, 69, 71 (K-8 Schools)
MLK 9 26, 27, 33, 43, 67, 69, 72, 81 (K-8 Schools), West Hertel (7-9)
#48 5 19, 26, 27, 33, 38 (K-8 Schools)
#53 7 27, 38, 66, 67, 72, 81 (K-8 Schools); 42 (K-6)
#59 15 21, 27, 29, 43, 64, 66, 69, 72, 81 (K-8 Schools) 51, 60, 65, 86 (K-6), 70 (K-7), West Hertel (7-9)
#61 14 19, 21, 27, 28, 64, 66, 71, 81 (K-8 Schools) 42, 51, 60, 65, 86 (K-6 Schools),
#62 19 West Hertel (7-9) 21, 26, 27, 28, 29, 33, 38, 64, 66, 67, 69, 71, 72, 81 (K-8 Schools) 51, 60, 65 (K-6 Schools) 70 (K-7 School), West Hertel (7-9)
#74 13 19, 21, 26, 64, 66, 71, 81 (K-8 Schools) 42, 51, 60, 65, 86 (K-6 Schools), West Hertel (7-9)
#75 5 28, 67, 69, 72 (K-8 Schools), 70 (K-7 School)
#82 1 #81 (K-8 School)
#90 8 27, 29, 33, 43, 67, 69, 72 (K-8 Schools), 70 (K-7) School #52 (Mini Magnet) Accepts Q.I.E. Students City
Wide
Thus schools 23, 59, 61, 62, and 74 all send QIE students to over ten receiving schools, with # 62 sending students to as many as nineteen.
The feeder pattern is actually even more complex that it appears on Table III because a number of hardship transfers were authorized during 1978-79 for parents who *843did not want their children to change schools as a result of the new feeder pattern. (Ex. 358.) In addition, the court suspects that a number of unauthorized departures were made from the court ordered feeder pattern during 1978-79, complicating the system even further.11
It is apparent that the complexity of QIE was not significantly reduced in 1978-79. The bus rides remained too long. (Ex. 357. See also Ex. 364.) In addition, the planning and execution of such circuitous bus routes continued to carry a disproportionate price tag considering the amount of desegregation actually achieved by QIE.
The complexity of the QIE program has additional ramifications for the high schools. QIE feeder patterns cross high school district lines, and QIE students in the eighth grade are given a choice of attend- • ing either the high school fed by their QIE receiving school or the high school fed by their residential or home district. For example, QIE students from the sending school district of school 59 may be assigned, inter alia, to receiving schools 21, 27, 29, 45, or West Hertel. These receiving schools feed Bennett, Southside, South Park, Lafayette, and the Riverside districts, respectively. The students may elect to attend one of these high schools or Kensington, the high school fed by the home district of school 59. These options make it possible for a student from school 59 to attend any one of the six academic high schools in the city.
As a result of the choices given eighth graders in QIE, it is impossible to predict the anticipated racial breakdown of the high schools by examining the eighth grade enrollments at each high school’s feeder schools. This in turn unduly complicates and impedes the process of designing racially balanced attendance zones for the high schools.
Another equally serious difficulty caused by the complexity of QIE has not been discussed in previous orders but has persistently recurred. This is the question of court supervision. Certainty and manageability are important ingredients of desegregation decrees. Cf. Evans v. Buchanan, 435 F.Supp. 832, 841 (D.Del.1977). The remedy imposed must be supervised by the court to determine whether or not it is being successfully and properly implemented. This necessitates the existence of objective standards for evaluating the proposed remedy. It also requires a remedy which is not so complex that compliance cannot as a practical matter be determined.
As Tables I, II, and III suggest, QIE by design alone borders on being too complex to administer. Very few individuals on the Board’s staff are conversant with the details of the system. QIE records are in most cases kept by hand and are not readily accessible. Thus, when a question arises as to the current design of the program, information is very difficult to obtain and usually is provided, if at all, many months after it is requested.12 Moreover, the factual detail is so overwhelming that it is exceedingly difficult to determine whether the program is being carried out as promised. See, e. g., the problems discussed in text accompanying notes 14-15, infra. Finally, because new students are recruited each year, the composition of the program is in a perpetual state of flux, necessitating painstaking yearly evaluations. School officials *844cannot predict from year to year how many students will enroll, what schools they will attend, and the minority-majority enrollments of sending and receiving schools. Due to changes in residential housing patterns, school officials cannot even predict which schools should be sending schools and which should be receiving schools.13 Since QIE applications are accepted on an ongoing basis after spring recruitment, no firm estimates-of enrollment figures are available until a month or so before schools open. By that time, students have already been promised they could attend a particular school and most of the logistical arrangements have already been made. As a practical matter, the court finds itself in many instances locked into a feeder pattern which it neither approves nor understands. Intelligent planning on a long-term basis is virtually impossible.
The complexity of QIE’s design is compounded by its implementation. On countless occasions in the past, express directions of the court regarding QIE have not been carried out. This casts substantial doubt over the Board’s control over the program.
One illustration of this difficulty has already been outlined by the court in its order of June 29, 1978, at 31-32 and need not be repeated. Several others shall suffice. In its order of August 8, 1978, the court specified the feeder pattern to be used for QIE in 1978-79 and directed the Board to follow the pattern “without exception.” It also made clear that it rejected the Board’s proposal to expand the QIE program in succeeding years. Despite these directions, the staff sent QIE students to a number of receivers not authorized by the court. For instance, in the court-ordered feeder pattern for 1978-79 (see Table II), school 9 was to send students to schools 69, 71, and 72. It in fact sent students to twenty-one receiving schools.14 For 1979-80, it has been proposed that school 9 send students to eight receiving schools, which only includes one of the three receiving schools ordered by the court on August 8, 1978. (Ex. 360.) Similarly, school 78 was not listed as a sending school in the August 8 order,15 yet in 1978-79 it sent nine students to eight different receiving schools. It simultaneously received three QIE students in the eighth grade from school 8, thereby acting as both a sending and a receiving school. For 1979-80, it has been proposed that school 78 send students to twelve receiving schools. (Ex. 360.)
Although some of these deviations from the court-ordered feeder pattern may be explained by the 119 hardship transfers (see Ex. 358), certainly all are not. Moreover, it is clear from the proposals for next year that the QIE staff has no intent of simplifying QIE feeder patterns in accordance with the general directives of this court.
A second illustration of failure to obey express court orders as well as the Board’s promises to the court is more blatant and inexcusable. According to the only information available to the court on QIE enrollment for this year (Ex. 359), over forty-five white students are listed as participating in QIE. Yet QIE, according to the Board’s own proposal, is offered only to minority students.
A final illustration will suffice. As part of the Buffalo Plan, the Board promised not to assign new QIE students to naturally integrating elementary schools. The Board *845also promised not to assign QIE students in such a way as to adversely affect the racial balance of the high schools fed by the receiving schools. (See Exs. 126 and 181 at 26.) In particular, the Board represented that it would not assign QIE students to the elementary schools feeding Bennett and Kensington or to naturally integrating schools 9, 68, 71, 78, and 80.
Despite these representations, large numbers of new QIE assignments were made during 1977-78 to these schools. In almost all cases, further new QIE assignments were made to these schools in 1978-79. Table IV lists the schools identified by the Board which would not receive additional QIE students and the number of students they actually received.
TABLE IV
QIE RECEIVER HIGH SCHOOL DISTRICT NO. OF NEW QIE STUDENTS 1977-78 1978-79
11 Kensington 27 15
21 Bennett 61 45
22 Bennett 7 0
48 Kensington 16 45
44 Kensington 27 11
68 Bennett 62 0
66 Bennett 88 84
68 Kensington & Bennett 86 0
69 Kensington 107 77
71 Kensington 65 73
78 Kensington 1 0
80 Kensington 16 0
81 Bennett _68 64
TOTAL 625 414
SOURCES: Plaintiffs’ Motion of March 5,1979 Regarding High Schools, at 10. Ex. 356.
These assignments were in direct violation of the Board’s representations in the Buffalo Plan and contributed to racial imbalance at Bennett and Kensington and at certain elementary grade levels in the receiving schools.
These examples are cited merely as illustrations of the difficulty the court has experienced in obtaining compliance with its orders regarding QIE. Undoubtedly other departures have occurred without the court’s knowledge and which because of the complexity of the feeder pattern are difficult to uncover. Although the court has in the past tolerated such violations as unintentional errors or misunderstandings, a pattern has now clearly emerged of failure to implement QIE in accordance with the court’s instructions.
E. ADDITIONAL SHORTCOMINGS
Implicit in the program is the promise that the education provided at the end of the bus ride will somehow be superior to the education offered in the neighborhood schools. This is suggested in the name of the program, “Quality Integrated Education,” and in the promotional literature handed out to students by QIE recruiters. This approach raises serious problems of equity between QIE receiving and QIE sending schools. The tendency on the part of the Board is to concentrate its efforts on improving the receiving school programs, leaving at least the impression that the QIE sending schools do not receive the same attention as the receivers or the magnets receive. Furthermore, as Table V demonstrates, many of the all-minority neighbor*846hood schools in the inner city have large capacities but are enrolling fewer and fewer students, partly as a result of the drain placed on the enrollments by QIE.
TABLE V
ENROLLMENTS AND CAPACITIES OF FIFTEEN 80-100% _MINORITY ELEMENTARY SCHOOLS._
ENROLLMENT
SCHOOL 1975-76 1977-78 1978-79 CAPACITY
12 298 258 209 434
23 727 556 470 783
31 706 491 452 1,107
37 629 471 444 1,215
39 767 621 609 1,296
48 324 176 150 675
53 840 1,031 971 1,188
59 546 330 225 474
61 607 480 383 621
62 739 591 477 918
74 612 515 502 783
75 314 188 136 648
76 633 618 610 864
82 580 593 490 837
90 548 473 402 648
TOTAL 8,870 7,392 6,530 12,491
SOURCE: Exs. 321 at 9 and 362.
This severe underutilization in turn makes minority schools prime candidates for school closings.16 At the same time that underutilization of minority schools is increasing, many QIE receiving schools are filled to and beyond capacity. This is not sound use of existing resources. In addition, it increases the disproportionate burden on minorities by contributing to school closings in minority neighborhoods. Cf. United States v. Texas Education Agency, 467 F.2d 848, 871-72 (5th Cir. 1972) (en banc).
A final limitation of QIE warrants brief discussion. As pointed out in the court’s order of June 29, 1978, an additional shortcoming of QIE is the failure to recruit students in the lower grades. During the 1977-78 school year, nearly 41% of the QIÉ students were in grades 7 and 8. (This is caused, at least in part, by the mandatory enrollment in QIE of students from schools 12, 23, 39, 49, 54, 61, 62, 75, and 90, whose neighborhood schools terminate in grades five and six.) Despite promises by the Board to recruit more students from the lower grades, no improvement took place in 1978-79, with the percentage figure remaining the same. (See Plaintiffs’ Motion of March 5, 1979, Ex. F.) As a result of this structure, a number of receiving schools are integrated only in the higher grades, and QIE students are perceived as intruders. It is generally recognized that desegregation is most successful when it occurs in the lower grades before racial attitudes are strongly formed. I realize that it is much more difficult to recruit QIE students in the lower grades. But this fact does not justify the resulting top heavy structure of QIE; it merely highlights one of the inherent limitations of the program.
For all of these reasons, I find that QIE does not provide a workable desegregation remedy for the Buffalo public schools. It *847therefore must be discontinued as soon as possible with an alternative desegregation program put in its place.
IV. REMEDY GUIDELINES
Having found that steps must be taken to further desegregate the all-minority schools in Buffalo without relying on QIE, I now turn to the task of setting forth guidelines for the development of a new remedy plan which meets these objectives. In the Phase I order of July 9, 1976, I outlined in open court the relevant legal principles governing remedy orders in school desegregation cases. These principles remain applicable today and bear repeating:
In 1954 the Supreme Court in Brown held that segregated educational facilities are inherently unequal. In my decision of April 30th, I found that the Buffalo Public School System was operated contrary to law as interpreted in Brown and in subsequent Supreme Court decisions.
Therefore, it must be the goal of the Court, and of the parties, to attempt to fashion a desegregation plan which eliminates black schools and white schools. This does not mean that all schools in the system must show the same or nearly the same ethnic composition as the system does on a whole, but every effort must be made to eliminate assignment patterns that leave certain schools so out of proportion in their ethnic makeup that they are in effect black or white schools. A desegregation plan to be acceptable must, to as great an extent as possible, eliminate racially identifiable schools. This rule applies to all schools in the district, elementary, junior high school, high school and vocational school. Theoretically, the ideal plan would be one in which each school in the district would reflect the racial composition of the pupils in the district as a whole. However, it is evident that such a plan would not take into account the needs of the community, the distances and difficulties of travel, the resources of different schools, the lack of physical plant in others, the age and abilities of the children attending schools and many other factors. Such factors may be considered by the defendants, for, as the Supreme Court has stated, it is their duty to, “make every effort to achieve the greatest possible degree of actual desegregation, taking into account the practicalities of the situation”, a quote from David [Davis] v. Board of School Commissioners, 402 U.S. 33 [91 S.Ct. 1289, 28 L.Ed. 577] (1971). However, any practicalities that impede the desegregation effort must be fully explained by the school board and the other defendants. Vague conclusory and unsupported assertions are not sufficient.
. Federal courts may adopt desegregation remedies requiring bus transport only as a last resort, 20 U.S.C. Sections 1713 and 1755. These sections do not prohibit busing but only require the Court to consider all other alternatives first. It appears from the nature of the residential areas in the City of Buffalo that any effective desegregation plan will have to include some transport of students. Indeed, many students use bus transport in the City now and have used it for many years in the past. As the Supreme Court made clear in the Swann case, in school districts with segregated neighborhoods, desegregation plans cannot be limited to the walk-in school. Furthermore, busing is a normal and accepted tool of educational policy and cannot be rejected without a showing that the time or distance of travel is so great as to risk either the health of the children or significantly impinge on the educational process. That is a quote from Swann against Charlotte-Mecklenburg, 402 U.S. 1 [91 S.Ct. 1267, 28 L.Ed.2d 554] (1971).
Any plan that places great reliance upon open enrollment or parental choice cannot be constitutionally adopted. Green against School Board, 391 U.S. 430 [88 S.Ct. 1689, 20 L.Ed.2d 716]. Furthermore, freedom of choice plans have had a long history of failing when they were attempted to be used.
*848To sum up this brief review of the law, if the plan proposed does not fully desegregate the school system, there must be an explanation based upon hard evidence explaining why full desegregation cannot be effected. The appropriate plan must carefully balance the costs of desegregation against the possible results to be achieved. Where the benefits to be gained are negligible, the steps to be taken must be carefully weighed. An effective and feasible plan should prevent re-segregation if possible. It should have permanence if possible. This does not mean that the Court may consider the threat of white flight, but if there is sufficient believable evidence in the record that resegregation will occur, then this evidence may be considered as part of the practicalities of administering the plan.
Transcript of July 9, 1976, at 16-21.
To this general discussion I need only add that under Dayton, supra, a systemwide remedy is appropriate for the Buffalo schools.
The defendants are directed to draft a new remedy plan in accordance with these principles and the following guidelines.
1. The Board has an affirmative duty to take whatever steps are necessary to eliminate segregated schools in Buffalo. It must devise a realistic and meaningful plan which integrates the remaining one-race schools “at the earliest practicable date.” Green v. County School Board, supra.
2. The goal of the plan is for each school in the system to reflect the racial composition of the pupils in the district as a whole. According to current enrollment figures, the Buffalo school system is approximately 50% majority and 50% minority. I recognize that due to practical considerations a mathematically perfect mix would be difficult to achieve in every elementary school. As a rule of thumb the court will consider a racial composition between 25% and 65% minority to be a desegregated school. I believe that under all of the circumstances shown by the evidence in this case, this standard is a reasonable one.
Any deviations from this guideline will be allowed only after a specific and detailed showing on a school-by-school basis that practicalities such as distance and difficulties of travel or physical barriers preclude further desegregation. See Davis v. Board of Commissioners, 402 U.S. 33, 91 S.Ct. 1289, 28 L.Ed.2d 577 (1971). I realize that much of this information has been provided to the court in prior proceedings in one form or another. But in view of this order and the direction of the Supreme Court in Dayton, it is necessary that this analysis be updated. It may be that it will be impossible to rid the system of every all-minority school, but an attempt must be made and, if this objective cannot be reached, good reason must be set forth in the record.
3. The plan wherever possible shall avoid reassigning students already attending integrated schools. This recognizes the desirability of maintaining integrated schools in integrated neighborhoods. See Swann v. Charlotte-Mecklenburg Board of Education, supra, 402 U.S. at 28, 91 S.Ct. 1267. For these purposes, a school shall be considered integrated if the minority enrollment is between 25% and 65%.
4. However, the discontinuance of QIE necessarily entails reassigning many students currently enrolled in QIE. Although the disruption that this will involve is regrettable, it is necessary in order to avoid prolonging the inefficiencies and inequities of QIE and to enable intelligent planning to proceed. As part of its plan, the Board may wish to adopt pairing, clustering or other similar arrangements between QIE sending and receiving schools in order to minimize the' relocation of QIE students. But in general, QIE as it is currently structured must be dismantled as soon as possible. I note that once the remaining all-minority schools are desegregated, as a practical matter there will be no need to rely on QIE to bring partial desegregation to these schools.
*8495. As I have already noted in this decision, the Board by the use of QIE and the magnet program has eliminated virtually every all-majority school in the system. The desegregation of previously all-majority schools is certainly a worthwhile and commendable accomplishment and ought to be continued in the new plan.
6. The desegregation techniques to be used by the Board in meeting the goals of this order shall consist of those which have been approved by the United States Supreme Court, including pairing or clustering of schools, redrawing of geographic attendance zones, and closing schools, with students reassigned so as to enhance racial balance. The plan may contain voluntary components (such as the magnet schools) so long as such components promise speedy and effective means of desegregation and do not place unequal burdens on any racial group. Green v. County School Board, 391 U.S. 430, 439-41, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968). However, any voluntary components shall be carefully scrutinized by the court to make sure constitutional standards are met.
7. As I noted in my original remedy order of July, 1976, it appears from the nature of residential patterns in Buffalo that any effective desegregation plan will include transportation of some students. But this will not necessarily require a sharp departure from prior practice. According to the Board’s figures (Ex. 356), almost half of the students in the system are already being transported by bus, the majority of them for reasons unrelated to desegregation.17 ' Furthermore, in order to minimize the inconvenience and burden of lengthy bus rides, the defendants are directed to devise transportation routes which to the greatest possible extent do not require any child to ride a bus for more than 35 minutes each way. In any case in which the projected transportation time will exceed 45 minutes, the Board shall make specific application to this court for approval.
8. The Board’s School Utilization Committee and New York State Education Department have filed reports finding that substantial savings could be achieved by school closings. The State has recommended that as many as 15 elementary schools be closed for next year. (Ex. 361.) Given the fiscal difficulties now facing the City and the Board, as well as the sharp decline in school enrollments in the past decade, it is clear that the new remedy plan must include plans to close a substantial number of schools.
9. The status of the academic and vocational high schools is currently under discussion by the parties, and various plans are being considered to improve the racial balances at these schools for next year. The court shall continue to monitor the parties’ negotiations on the high schools and, if necessary, shall issue a subsequent order. In devising a new remedy plan, the Board shall take into account the ramifications that changes in the elementary schools will have for the high schools and design a comprehensive plan encompassing all of the public schools.
Because substantial changes in the desegregation program are required by this order, I am encouraging but not requiring the defendants to devise a comprehensive plan for the 1979-80 school year. The planning and implementation may be done in phases, with the final deadline for full implementation being the fall of 1980.
*850As the first phase of the new plan, the Board ought to close a substantial number of schools for next year. School closings provide an excellent means not only of reducing costs but also of achieving additional desegregation. The Board already has under consideration a number of recommendations for school closings. These deliberations should be completed as soon as possible, and the Board’s proposals shall be submitted to the court by June 22, 1979. Although specific closings must be approved by the court, it is the Board’s responsibility to make the initial determinations of which schools to close and where to send relocated students.18
On June 22, 1979, the Board shall also explain in writing what additional measures could be taken for the 1979-80 school year to achieve further desegregation. The comprehensive remedy plan shall be filed with the court no later than November 15, 1979.
The plaintiffs have moved to have a consultant appointed by the court to prepare a final remedy plan. As previously discussed in the order of June 29, 1978, a consultant could provide valuable assistance to the court and the community in preparing a comprehensive, equitable, and affordable remedy plan. The presence of a consultant would also enable the Superintendent and his staff to devote their full attention to the difficult task of day-to-day administration of the schools.
After careful reflection, I have concluded that the best approach for all concerned is to direct the Board of Education to select its own consultant, subject to court approval. I recognize that the primary responsibility for devising a desegregation plan rests on school authorities. Green v. County Board of Education, supra, 391 U.S. at 439, 88 S.Ct. 1689. It is their duty to run the school system, and they are in the best position to make the difficult decisions regarding the use of resources and the structure of the school system which inevitably arise in devising a desegregation remedy. In addition, although I have pointed out various deficiencies in the present remedy plan, the Board and its staff have at all times cooperated with the court in attempting to devise and implement desegregation programs since the liability decision was issued in 1976. There has been no finding that the shortcomings of the current plan were the result of discriminatory intent. Rather, it appears that the deficiencies were caused by the design of the plan, lack of communication, and inadvertence. I am confident that the Board will continue to cooperate in bringing to the students of Buffalo an integrated as well as a sound education. For these reasons, I am giving the Board an opportunity to select a consultant in whom it has confidence.
This consultant must have established expertise in the areas of education and desegregation remedies. He or she also must be willing to devote as much time as is necessary to develop a suitable, comprehensive plan by November 15, 1979.
The Board is directed to submit its recommendation along with a description of the consultant’s powers and duties by July 13, 1979.
Summarizing the preceding order, I find that since the constitutional violations had a systemwide impact under Dayton Board of Education v. Brinkman, 433 U.S. 406, 97 S.Ct. 2766, 53 L.Ed.2d 851 (1977), the scope of the desegregation remedy in this case must be systemwide. The current remedy is not systemwide because it leaves intact a number of all-minority elementary schools. It also relies heavily on the QIE program, which for a number of reasons outlined in detail in this order does not provide a satisfactory long-term remedy. I am therefore directing the Board of Education to restructure its remedy plan in accordance with the specific guidelines set forth in this order. In devising a comprehensive, equitable, and economical remedy plan, the Board is directed to use an outside consultant chosen by the Board and approved by the court.
So ordered.