These appeals are from orders entered in 1980 in the continuing Boston school desegregation case. For prior aspects of the case, see Morgan v. Kerrigan, 530 F.2d 401 (1st Cir.), cert. denied, 426 U.S. 935, 96 S. Ct. 2648, 49 L. Ed. 2d 386 (1976); Morgan v. Kerrigan, 509 F.2d 580 (1st Cir. 1974), cert. denied, 421 U.S. 963, 95 S. Ct. 1950, 44 L. Ed. 2d 449 (1975).

In appeals numbered 80-1271, 80-1272 and 80-1296 certain defendants1 and an intervenor-Boston Home and School Association-challenge orders entered in March and April of 1980 by the district court rejecting parts of a "Unified Facilities Plan" developed at the court's request by the Boston School Department and others. The challenged orders directed the closing of two elementary schools in addition to the ten recommended by the defendants. The court also conditionally rejected defendants' "moderate" student assignment plan for the 1980-1981 school term. Finally, it refused to accept two novel proposals-"linkage" and "beacon schools"-said to ameliorate the destabilizing effects of school closings.

In appeal numbered 80-1288 the intervenor Boston Home and School Association appeals from the district court's April 25, 1980 denial of its motion that the court altogether terminate its jurisdiction over student assignments.

Underlying these appeals are the district court's efforts between 1974 and 1980 to eliminate surplus seats in the Boston public schools by compelling the closing of unneeded school buildings. In its original desegregation opinion, rendered in 1974, the district court found that policies regarding the building of new schools and the closing of old ones had contributed to the intentional segregation of the Boston school system. The court stated,

Plaintiffs have proved that the defendants intentionally segregated schools at all levels, ...; built new schools for a decade with sizes and locations designed to promote segregation; maintained patterns of overcrowding and underutilization which promoted segregation at 26 schools; and expanded capacity of approximately 40 schools by means of portables and additions when students should have been assigned to other schools with the effect of reducing racial imbalance.

Morgan v. Hennigan, 379 F. Supp. 410, 481 (D.Mass.1974), aff'd sub nom. Morgan v. Kerrigan, 509 F.2d 580 (1st Cir.), cert. denied, 421 U.S. 963, 95 S. Ct. 1950, 44 L. Ed. 2d 449 (1975).

In 1975, when outlining remedial steps to desegregate the Boston schools, the district court, pursuant to recommendations by court-appointed experts, indicated that some 20 schools should be closed. It said,

A major reason for closing schools is that desegregation is more easily and economically achieved through the consolidation of student bodies. Many of the city's elementary schools in black areas have in the past been overcrowded; many elementary schools in white areas have been underutilized, e.g., when a new school was constructed to replace an old one in a predominantly white neighborhood, the school committee accommodated parents protesting the closing of the old one by keeping them both open. Should school facilities be uniformly used to capacity, an excess of several thousand available seats at the elementary school level would remain. Thus a number of the older elementary schools can be closed, with accompanying savings of the costs of operating and heating those schools. Elementary schools will be kept open whose locations enable busing to be minimized overall, and which permit the more efficient assignment of students by geocodes, accomplishing desegregation and minimizing the need to split geocodes. Uniform utilization of facilities throughout the city will also tend to equalize the availability of the system's resources to all students.

Morgan v. Kerrigan, 401 F. Supp. 216, 245 (D.Mass.1975), aff'd, 530 F.2d 401 (1st Cir.), cert. denied, 426 U.S. 935, 96 S. Ct. 2648, 49 L. Ed. 2d 386 (1976).2

Describing the "development of a long-range plan for the construction, repair, and replacement of school facilities" as "one precondition to disengagement," the court in 1977 directed the School Committee, the City of Boston and the State Board of Education (referred to collectively as the "Joint Planners") to submit a "Unified Facilities Plan (UFP) ... (to) include a schedule for closings, construction and renovation, replacements, as well as repair and refurbishing for all nine school districts, for the years 1977 through 1986." The Joint Planners were told to take account of "the likely effect upon desegregation of any developments in implementing the UFP."

Six months after this order, on November 25, 1977, the Joint Planners submitted a lengthy "Unified Facilities Plan" with the requested projections relating to school construction and utilization. The 1977 UFP, however, generated considerable community debate and intense opposition. One community group, the Citywide Parents Advisory Council (CPAC), which had not been afforded an opportunity to have input into the Plan, thought it

to be based upon inaccurate and insufficient data; to ignore educational, desegregation and community factors in its decision; to burden minority children unfairly in its pattern of school closings; to fail to prove its economic arguments for closing; to overlook the long-term implications for Boston of its recommended actions.

Apparently because of such opposition and because the plan lacked sufficient community input, the district court, in modifying orders on March 21, 1978, implicitly rejected the proposed UFP. It called for "a Unified Facilities Plan which will meet future needs and gain the understanding and support of parents and community groups." The court later urged the Joint Planners to begin by agreeing on the basic criteria for a UFP. In response, the Joint Planners developed a "Manual for District Planning Activities," which they submitted to the court on April 23, 1979. This Manual contained a six-step planning process and included criteria for determining which schools would be closed in future years. Since the closing decision was linked to the number of anticipated students, a formula for determining the excess capacity in the system was also developed and included in the Manual.

The court enthusiastically approved the Manual on August 15, 1979, noting, however, that portions of the Manual "impl(y) erroneously that safeguarding continuity and stability of the education process is of equal importance to achieving desegregation and equal education opportunity." The court stated firmly that "(d)esegregation, in accordance with the flexible guidelines promulgated in 1975, must be given first priority." The court further ordered that the new UFP

shall eliminate by July 1, 1980 no fewer than half of all excess seats (as of April 15, 1979) in elementary schools; and shall identify by name elementary school facilities to which students will not be assigned for the 1980-81 school year.

The court ordered the Joint Planners to file the new UFP no later than December 1, 1979.

On October 3 the Joint Planners published a draft UFP, which came to be known as the "Green Book." The Green Book called, among other things, for the immediate closing of 16 elementary schools,3 the designation of the remaining nearly 80 elementary schools as either "base" schools (indicating they would remain in long-term service) or "support" schools (indicating they were subject to closing based on enrollment trends), and the conversion of Roxbury High School to a middle school. Like the 1977 UFP, the 1979 Green Book generated intense community debate, particularly regarding the schools slated for immediate or eventual closing. In response to these and other concerns, Boston School Superintendent Wood sent a memorandum on November 29, 1979 (two days prior to the December 1 UFP deadline set by the court) to the School Committee which recommended substantial revisions to the UFP.4 Superintendent Wood's memo became the next incarnation of the UFP-what the appellees call the "Wood Plan."

The Wood Plan reduced the number of schools scheduled for immediate closing from 16 to 10. Instead of designating certain schools as "support" schools, the Wood Plan proposed two new categories of facilities-"linkage" schools and "beacon" schools. The linkage and beacon proposals were staff-generated ideas5 meant to ameliorate the negative community reaction to marking certain schools for closing by identifying them as "support" schools.

Under the linkage proposal, contiguous community district elementary schools (some schools were apparently as close as half a mile from their linked partner school) were to be paired for student assignment purposes. A child assigned to a linked pair would be assured that at least one of the linked schools would remain open for a minimum of seven years. When the combined enrollment for two linked schools could be accommodated in 85 percent of the available capacity in one of the schools, one or the other school would have to be closed. The decision as to which of the two to close would be made in consultation with the district superintendent, parents, and other affected parties. The idea behind the plan was to reassure parents that children in schools slated for closing would face only minimal dislocation. The plan also provided a built-in incentive for parents to keep enrollment high in both schools, and its "automatic" closing mechanism reduced some of the political tension surrounding school closings.

Under the beacon plan, students could apply for a limited number of seats in selected elementary schools (to be designated "beacon" schools) in their district as an alternative to attending their regularly assigned district school.6 Such an assignment would only be allowed, however, if it enhanced desegregation at the beacon school and did not send the student's former school out of compliance with court-ordered racial quotas. Some 1,600 beacon "seats" at selected community schools were identified for this program.

Neither the linkage nor the beacon school concept had been the subject of public meetings or hearings, nor had the other Joint Planners approved the ideas in advance. Nevertheless, on November 30, the Boston School Committee voted to support the filing of the Wood Plan as the UFP "without approving or disapproving said Plan."7 The Public Facilities Department of the City of Boston, one of the Joint Planners, likewise joined the filing on December 3 but stated that it did "not unequivocally approve or disapprove the contents of the Superintendent's Plan." The last Joint Planner, the Massachusetts Board of Education, initially declined to endorse the Plan, but finally did so on December 17 with specific objections to various aspects of the proposal.

Court hearings were held in the following weeks at which the school closings, the linkage and beacon concepts, and the student assignment proposal were explored by all parties. Extensive testimony was also heard regarding how many seats were truly excess in the Boston system. On January 16, 1980, the Joint Planners filed a consolidated statement in which they collectively endorsed the UFP's list of ten schools proposed for immediate closing, the linkage and beacon concepts, and the "moderate" student assignment approach. They said that they viewed the components of their plan as an "interdependent package" and warned that "(s)ignificant departure from the essential components of the plan presented will surely threaten and may well end the joint nature of the plan."

For its part, the court had already expressed its dissatisfaction with the UFP at a court hearing on January 11, noting that the "proliferation of documents since December 1st ... falls far short of a facility plan ...."8 The court had earlier noted that the revised UFP was also more than a mere plan for building and closing schools, amounting instead to "a modification of the desegregation plan ...." The court specifically stated on January 11, 1980 that it would "not entertain these substantial changes in the extant Court orders (i.e., the original desegregation plan) with respect to proposals like beacon and linkage schools unless there is a motion filed specifying the modification that is sought to be made in the existing Court orders." The court called on January 11 for further filings revising the method for determining excess capacity. It finally issued its orders and a memorandum of decision on March 31, 1980, and April 2, 1980. The court accepted the Joint Planners' recommendation to close ten named elementary schools. However, it added two others, Conley and Richards, stating that closing these would more closely approximate the court's August 15, 1979 requirement that "half" of excess capacity be eliminated. In addition, it refused to allow the Planners to convert Roxbury High School to a middle school, rejected the linkage and beacon school concepts, ordered further student assignment modifications-including possible "splitting" of geocodes in order to enhance achievement of the court's racial ratios for the 1980-81 school year-and bid the Joint Planners resume unified facilities planning. It is from these orders that the City of Boston, the Boston School Committee, and the Boston Home and School Association appeal in cases numbered 80-1271, 80-1272, and 80-1296.

Appellants contend that under the principles enunciated in Milliken v. Bradley, 433 U.S. 267, 280-81, 97 S. Ct. 2749, 2757, 53 L. Ed. 2d 745 (1977),9 the district court lacked remedial authority to reject the proffered UFP "package" and to substitute its own judgment on the matters at issue here-specific school closings, student assignments, and the beacon/linkage proposals.10 They assert that the Joint Planners' proposals-as the good faith efforts of the official agencies charged with running the schools-deserved deference rather than rejection by the court and that the court exceeded its legitimate role as the enforcer of a desegregation remedy and strayed into the realm of general educational policy.

We turn first to the assertion by appellees that many of the issues presented have been mooted by actions taken since these appeals were filed. Appellees point out that on May 2, 1980, the district court stayed its school closing order at the request of all parties.11 Thereafter, on May 11, 1981, the court approved a proposal by the Joint Planners to close some 26 elementary schools-including Richards.12 In addition, in 1981 the court approved the Joint Planners' earlier proposal to convert Roxbury High School to a middle school for the 1981-82 school year. Appellees argue that there is thus no "live" controversy regarding the court's orders to close the Conley and Richards schools and to keep Roxbury High School open. North Carolina v. Rice, 404 U.S. 244, 246, 92 S. Ct. 402, 404, 30 L. Ed. 2d 413 (1971). Similarly, with respect to the student assignment orders of April 2 and April 24, 1979, appellees point out that these orders were "tentative" at best and related only to the 1980-81 school year. Finally, they point out that the Joint Planners have failed to reintroduce the "beacon" school proposal in subsequent filings (though the "linkage" concept has been reasserted). With respect to school closings, student assignments, and at least the beacon concept, therefore, appellees argue that the case is moot.

We agree with appellees that these appeals are moot with respect to the district court's 1980 orders closing Conley and Richards, and retaining Roxbury High. Appellants themselves have since requested the closing of Richards; they do not now contest that Conley will remain open; and they have already obtained from the district court the relief they wish respecting its earlier decision about Roxbury High. Because of the 1981 closings, the orders no longer present live controversies which this court could resolve by the granting of any specific relief. Nor are they, in these particular circumstances, "capable of repetition, yet evading review." Weinstein v. Bradford, 423 U.S. 147, 149, 96 S. Ct. 347, 348, 46 L. Ed. 2d 350 (1975).

As to the student assignment orders13 and the proposals regarding beacon and linkage schools, however, we believe there remains a case or controversy. Student assignments are annual events which would evade review ad infinitum if treated as moot simply because the relevant year expires before an appeal is decided.14 Weinstein v. Bradford, 423 U.S. at 149, 96 S.Ct. at 348. The linkage and beacon school concepts, meanwhile, either have been or could be15 reasserted as part of the UFP. In such a situation, appellants are entitled to review of the district court's rejection of these proposals. The ongoing character of the UFP process guarantees that the appellants have the requisite "stake in the outcome" to present a justiciable case. Deposit Guaranty National Bank v. Roper, 445 U.S. 326, 335, 100 S. Ct. 1166, 1172, 63 L. Ed. 2d 427 (1980). See also United States Parole Commission v. Geraghty, 445 U.S. 388, 395-96, 100 S. Ct. 1202, 1208, 63 L. Ed. 2d 479 (1980).

Appellants challenge both the district court's April 2, 1980 order rejecting the UFP's "moderate" assignments approach (and ordering possible geocode "splitting" to enhance racial balance in the school system) and its April 24 conditional order pending the appeal of its April 2 school closing decision. Neither order was ever implemented, as the court ultimately elected to reinstitute the assignment plan for 1979-80. See note 11, supra. After a careful review of the record, we cannot say that the district court abused its authority in issuing the orders it did in lieu of accepting defendants' proposed moderate plan. For reasons amplified below when we reach the Boston Home and School Association's bald assertion that the court should at that time have relinquished all jurisdiction over student assignments, the district court was not required to view its role with respect to student assignments as then at an end.

The original desegregation plan, which this court upheld on appeal, required each of Boston's schools to reflect the racial mix of the district within which the school was located-with, however, considerable leeway (25 percent) for deviation.16 By 1980, these goals had still not been reached at 16 of the schools, although the parties dispute how meaningful and serious the disparities were. While desegregation may not be reduced to a matter of ratios, and while population shifts occurring after the termination of willful segregative conduct may constitutionally result in racially imbalanced schools, see Morgan v. Kerrigan, 530 F.2d at 421-22, the use of racial quotas during the remedial phase of a desegregation case to achieve and measure progress towards a unitary system has been endorsed by the Supreme Court. Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. at 25, 91 S.Ct. at 1280. In 1980 the remedial phase was still in progress in Boston. The lower court had discretion, paying due regard to all other relevant factors, to seek to bring the Boston schools into a closer correspondence with the flexible ratios it had earlier prescribed.

That the case remained in the remedial stage in 1980 distinguishes it from Pasadena City Board of Education v. Spangler, 427 U.S. 424, 96 S. Ct. 2697, 49 L. Ed. 2d 599 (1976). There the district court, having accepted and supervised the successful implementation of a locally generated desegregation plan, construed the plan, questionably, to require continued annual changes in student assignments to ensure that no school would ever have a "majority of any minority." The Supreme Court held that the district judge lacked the authority to impose such a requirement because he already had accomplished his constitutionally mandated objective: successful implementation of a racially neutral attendance pattern. 427 U.S. at 436-37, 96 S.Ct. at 2704-05. Subsequent changes in attendance resulting from population shifts rather than segregationist policies were not the court's to correct.

The history in this case has been quite different. In the 1970's, Boston officials vigorously fought desegregation, forcing the court to create and enforce its own desegregation plan, which included the racial ratios in question. Appellants had not by 1980 taken adequate steps either to amend the desegregation plan or to terminate the district court's involvement in the remedial phase of the case. See section III, infra. The court was thus entitled to regard its desegregation plan, which this court had earlier endorsed, as a still viable blueprint. At such a time, the court could insist on a student assignment program more rigorously geared toward achieving the express goals of its desegregation plan than was a rival program offered by the school authorities. In so saying, we have no occasion to consider, and do not consider, whether and to what degree a court might impose some kind of racial quotas, on a continuing basis, after termination of the remedial phase. The orders under review were all issued prior to such termination-at a time when, notwithstanding considerable improvement, the goal of a unitary system had yet to be achieved.

Spangler also differs in the rigidity of the court's "no majority of any minority" requirement as contrasted with the flexibility of the racial quotas built into the Boston court's desegregation plan. The latter were tailored to particular schools' own districts and allowed a substantial leeway for deviation. See note 16, supra. Moreover, in pursuing these goals, the district court did not in 1980 blindly disregard all other considerations: it retreated from its initial attempt to impose a more radical assignment program and acted, insofar as we can see, in a responsible and flexible manner when it finally continued the previous year's system. Thus, we do not see Spangler as requiring disapproval of the 1980-81 student assignments.

In approving the court's action we also recognize the need to defer to the district court's on-the-spot knowledge of local conditions at the time it acted. Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. at 28, 91 S.Ct. at 1282. Throughout much of the 1970's there had been overt resistance to desegregation. While matters were much improved by 1980, we cannot say at this distance, that the court's insistence upon an assignment plan more consistent with the goals stated in its 1975 plan-in lieu of the School Committee's so-called moderate plan-was an abuse of its broad discretion in a matter so central to any desegregation remedy.17

Having said this, we nevertheless emphasize what Spangler and other decisions of the Supreme Court make clear: "desegregation" is not to be equated simplistically with racial balance per se. The goal is a "unitary" system free from all invidious racial discrimination, not maintenance of particular ratios.

Furthermore, it must be recognized that aims such as the stabilization of school assignments and the encouragement of educational quality, while in no way substitutes for a non-discriminatory system or excuses for a discriminatory one, are vital concerns to students and parents of all races. These are matters on which the policies of local officials must be accommodated unless they jeopardize the court's mission to bring the system into compliance with the constitutional standard. As the remedial phase winds down, we expect the district court to recognize fully the right of local school officials to pursue educational goals and policies of their own choosing.

With these qualifications, then, we affirm the court's power to have issued in 1980 the student assignment orders it did.

The beacon and linkage programs present different problems which are less easily tied to familiar remedial moorings in a desegregation case than are student assignments. Appellants assert that neither of these rejected proposals would have enhanced segregation or was likely to make schools more racially identifiable. Indeed, the beacon concept was specifically put forward as a method of enhancing de segregation while improving the quality of selected community district schools. The desegregation goal was safeguarded, appellants claim, by the provision that no transfer could place the "sending" school outside the racial ratios set by the court, while the effect of the transfer at the beacon school had to be an enhancement of desegregation. Appellants also argue that the beacon system would enhance the prestige of selected community district schools and thereby "narrow the perceived gap" between the district schools (now apparently viewed as relatively undesirable) and citywide magnet schools (which draw students from all the city's school districts).18

The district court rejected the beacon concept, finding that it "would more likely create educational gaps and inequities within a community district and citywide district elementary schools." By favoring only selected community district schools with "beacon" status, the court felt, though it pointed to no specific evidence on this question, "a possible revival of discriminatory practices which led to the court's 1974 finding of liability" could occur. Secondly, the court stated that "the beacon program would be top-heavy with administrators and awash in paper work, and would accomplish very little not presently attainable under extant court orders." Finally, the court was simply skeptical that the proposal would succeed given the "formidable" desegregation goals contained in the plan, the fact that no final list of beacon schools had been developed, and the absence of public discussion of the concept.

Though the question is a close one, we think the record supports the district court's rejection of this proposal at the time it was presented. The beacon concept was a last-minute addition to what the court expected to be primarily a school closing plan. The record indicates that the beacon proposal was incomplete on many of its key points (including which schools were to be "beacons"). It could reasonably have been viewed by the court as a distraction from the school closing plan, which the court had believed was an essential part of the desegregation remedy.

To be sure, the court's stated reasons for rejecting the program are also open to the interpretation that the court simply disliked the beacon schools program on educational grounds. If this were all, the court would have erred, since under Milliken v. Bradley, 433 U.S. 267, 97 S. Ct. 2749, 53 L. Ed. 2d 745 (1977), courts must narrowly tailor their remedial orders to the unconstitutional conditions which gave rise to the need for court intervention. In so doing, courts should defer whenever possible to the reasonable proposals of the local officials charged with administering the school system. Id. at 280-81, 97 S.Ct. at 2757. A court has no constitutional mandate to dismiss a program merely because it believes the program would be "awash in paper work" or "top-heavy with administrators." Nor may it reject a program on the ground that all schools must be of equal quality (or mediocrity). Desegregation is not a mandate to equalize schools except insofar as inequality reflects racial bias.

But viewing the court's reasons as a whole, and in context, we do not perceive them simply as a pretext for an ill-advised effort by the court to substitute its educational views for those of school officials. Rather, given the timing of the beacon schools proposal, and the risk that it would draw time, effort and attention away from other issues more central to desegregation, we think the court was exercising its valid authority to keep its desegregation remedy on target. Had the plan been accepted, it would have required much reworking,19 consuming scarce administrative resources for a purpose the court considered not directly related to the main object of the UFP. (The court had already had several years of experience with its similar "magnet" program.)

These factors, combined with the program's tangential relevance to the facilities problem and several statements by defendants' experts that the proposal could "possibly" lead to resegregation,20 provided sufficient support for the district court's rejection of the beacon school plan in the particular form and at the particular time it was proposed. We accordingly hold that the district court did not abuse its discretion when in 1980 it rejected this aspect of the UFP. We so hold, however, only in that context. Nothing herein is to be taken as a decision by this court that on another occasion this plan or one like it could not or should not appropriately be adopted.

The linkage proposal presented the district court with different problems than did the beacon concept in that "linkage" was not a student transfer program-nor even, in a narrow sense, an assignment proposal. Rather, it was-and is-a proposal for targeting certain schools which may some day be closed.

According to appellants, the program was merely a reasonable administrative way of defusing the opposition which must come whenever schools are scheduled for closing-whether or not those schools are part of a segregated or "unitary" system. By assigning students to a linked pair of community district elementary schools-only one of which would be closed-rather than to only one which certainly would close, the system could assure some continuity in each student's elementary education. The plan was meant to depoliticize the school closing process by making closing "automatic," and to provide parents-of all races-with greater certainty regarding the elementary schools their children will attend.

Underlying the district court's analysis of linkage was its apparent perception that the proposal was really a way of delaying the "inevitable" job of closing schools. The court apparently felt it was better policy simply to announce which schools would close and let parents make their choices based upon this completely "known" situation. The district court also objected to linkage because it upset, by implication, several outstanding desegregation orders respecting grade structures, assignments of principals, and geocodes. Finally, the court was concerned, though here again it failed to ground its concerns on any specific record findings, that the linkage system would introduce an element of rigidity into student assignments which would make it more difficult "in future years ... to carry out the court's desegregation plan ...." The court was particularly worried about this added rigidity in light of factors such as the "decline of the birthrate" and the "proliferation of special programs" which have impinged upon the desegregation plan. As with the beacon idea, the court also felt that linkage would not work. Instead of giving parents the assurances claimed by the Joint Planners, the linkage concept threatened to be, in the words of the court, "divisive, confusing and destructive of the very stability which this court has sought to achieve in numerous previous orders."

In reviewing the court's rejection of the linkage proposal, we are faced with a confusing record. Appellants assert with some force that many of the district court's fears concerning linkage are simply not supported. The linkage proposal, for example, explicitly provides for a principal at each linked school, as required in the desegregation plan. Uniform grade structures, meanwhile, are preserved within each linked pair of schools, as are geocodes since the geocodes for each school are simply combined to achieve the geocode for each linked pair of schools. Factors such as a declining birthrate and the growth of special programs, meanwhile, will affect any proposal to close schools-whether it is characterized as a "linked" proposal or as a "support" school plan. Finally, appellants assert, it is not for the district court to substitute its judgment for the Joint Planners' on whether linkage would help with matters such as student recruitment or stabilization. Under Milliken, the court's sole concern must be desegregation. There was no substantial showing that the linkage program would adversely affect desegregation in "future years," claim appellants, while testimony was heard that desegregation would be enhanced.21

We think the appellants, on this record, have raised a substantial question respecting the linkage proposal. School closings are painful, but they are required by the desegregation plan formulated by the district court and approved by this court.22 The particular method or formula used to accomplish such closings, however, if developed in good faith23 and in the absence of record evidence that desegregation would be impaired, may be an occasion for deferring to the local authorities' interest "in managing their own affairs." Milliken v. Bradley, 433 U.S. at 280-81, 97 S. Ct. at 2757. Certainly, the district court should not overturn a good faith proposal merely on the speculation that it might cause problems due to a "decline in the birthrate" and "proliferation of special programs." These are factors beyond anyone's ability to control and do not appear to be of constitutional moment. Whether the proposal would lock the defendants into a system which could hinder desegregation in "future years," meanwhile, seems a matter of conjecture on this record.

The linkage proposal, however, made its appearance very late in the day. It was conceived just before the UFP deadline. It might have appeared to the court to have the potential-in unknown amount-of complicating or disrupting previously developed plans related to the difficult issue of school closings. The list of which schools were to be linked was incomplete, and the program had not been exposed to any public debate. We also think it important that the appellants did not comply with repeated district court requests that the linkage proposal be accompanied by specific motions to modify the 1975 desegregation plan where appropriate. The court was entitled to the benefit of such motions as part of its review of the proposal. Without explicit motions to this effect, it would be difficult, if not impossible, for the district court to manage the process of desegregation and for this court to make an informed review of district court orders.

Nonetheless, we cannot say on this record that linkage is so threatening to desegregation as to warrant its permanent rejection by a desegregation court. We are informed by the parties that the linkage proposal has been resubmitted with the 1981 UFP for district court consideration. This proposal is still pending before the district court, which deferred action on it in light of this appeal. It is not clear from the record whether this 1981 filing is the same as the one under review here or whether substantial changes have been made. We note, however, that substantial changes have occurred in the circumstances surrounding the proposal. The proposed closing of some 26 schools as part of the 1981 submission, in particular, radically alters the context of the proposal. We expect, therefore, that the district court will re-examine the linkage proposal in the light of these changed circumstances and any new filings which it has before it. If the linkage plan still contains elements which would conflict with outstanding court orders, defendants should bring these aspects of the plan to the attention of the court and move for specific relief from these orders.

In sum, in appeals numbered 80-1271, 80-1272 and 80-1296, we affirm the district court's orders respecting the UFP but with directions that it consider the beacon and linkage proposals in a manner consistent with this opinion should they be reintroduced by the defendants. Any further rejection of these ideas by the court should be accompanied by full record findings detailing why and how the programs frustrate desegregation.

In appeal numbered 80-1288 the Boston Home and School Association (BHSA) appeals the district court's April 25, 1980 denial of its motion to terminate the court's jurisdiction over student assignments. We affirm the district court's denial of this motion.

As we said earlier this year, "(a)bsent proceedings and proof showing that Boston's dual school system has been abolished, the district court retained power to make orders premised on the unconstitutional conditions it found to exist." Morgan v. O'Bryant, 671 F.2d 23, 26 (1st Cir.), cert. denied, ---- U.S. ----, 103 S. Ct. 62, 74 L. Ed. 2d 64 (1982). On this appeal we have no more of a formal record that desegregation has been achieved than we did in February. Indeed, the BHSA motion which is the subject of this appeal is but a two paragraph recitation of the major desegregation orders of this case unaccompanied by any stated grounds for relief24 or request for a hearing. The motion simply asks that the court relinquish jurisdiction.25 In such a posture it is difficult to do otherwise than affirm the district court's ruling.

The record on this appeal suggests, however, that much progress has been made toward the goal of desegregation and that jurisdiction over student assignments may now be coming to a close. The district court has observed the changing attitudes of the school defendants in carrying out its desegregation orders. On January 13, 1981, in ruling on a motion regarding staff reductions, the district court said,

The school defendants have done too much over the last few years toward implementing affirmative action and toward endeavoring to comply with a multitude of court orders for them to have to come in and establish their good faith each time there is an issue raised about something that they have done.

At this stage of the case, the good faith of the school defendants by and large is taken for granted by the Court. It is assumed by the Court.

This statement would seem to be a milestone in a case which has been marked by a lack of cooperation on the part of the defendants. Good faith in the operation of a school system and in the development and presentation of facilities and other plans is not alone enough to require automatic acceptance by the district court of the ideas and suggestions put forward by the school defendants. Dayton Board of Education v. Brinkman, 443 U.S. 526, 538, 99 S. Ct. 2971, 2979, 61 L. Ed. 2d 720 (1979); Green v. County School Board, 391 U.S. 430, 439, 88 S. Ct. 1689, 1694, 20 L. Ed. 2d 716 (1968). But good faith is not irrelevant in measuring the appropriate limits of district court discretion in instances where the court is asked to take account of "the interests of state and local authorities in managing their own affairs...." Milliken v. Bradley, 433 U.S. at 280-81, 97 S. Ct. at 2757. In affirming the original remedial orders in this case, for example, we specifically noted that the obstructionism exhibited by the School Committee in 1975 "justifies, in our opinion, a number of extraordinary measures which might otherwise be open to question." Morgan v. Kerrigan, 530 F.2d at 427. (Emphasis added.) By the same token, the ending of obstructionism plainly signals a return to greater local control.

In a late phase of the case, then, where good faith compliance with court orders is "assumed," groups such as the BHSA are entitled to press for specific and detailed findings on issues such as whether or not good reason remains for the court's continued jurisdiction over assignments.26 When hearings are held on that issue, the court should determine whether defendants are acting in good faith respecting assignments and whether compliance by defendants with the various court-mandated assignment remedies in this case has accomplished "maximum practicable desegregation," Morgan v. Kerrigan, 530 F.2d at 417 n.20, and brought an end to state-imposed segregation. See Spangler v. Pasadena City Board of Education, 611 F.2d 1239, 1244 (9th Cir. 1979) (terminating jurisdiction after substantial compliance achieved with court-approved plan and school board passed resolution promising to act in good faith to prevent a return to segregation).

In sum, with the reservations herein expressed, we affirm the challenged orders of the district court entered in 1980. It is obvious that progress has been made toward the goal of a unitary system for Boston's schools since 1974. At this late stage of the case, we urge the parties and the district court to renew their efforts toward the twin goals now sought-the achievement of a unitary system and court disengagement.

Affirmed.