United States of America, Cross-Appellee v. Board of Education of the City of Chicago, Cross-Appellant

FLAUM, Circuit Judge. For the third time in the nearly six years since its inception, the consent decree entered into by the Chicago Board of Education and the United States in order to settle a federal desegregation suit comes before this court fraught with problems of interpretation and implementation. The history of this decree is notable not for its success in raising an allegedly segregated school system to a constitutionally acceptable level, but for what it has illuminated about the character and limitations of the executive, municipal, and judicial institutions involved. The present state of the record is the product of the parties’ recalcitrance in fulfilling their obligations in a spirit of cooperation similar to the one in which they entered into the settlement, 1 and the seemingly insurmountable difficulties faced by the court system in trying to fill in the interstices of a multi-million dollar desegregation plan. In the latest phase of this litigation, the district court interpreted this court’s 1984 opinion and the decree’s funding provisions as requiring that the Board be treated as a priority, see infra note 10, with respect to all federal funds for which a program contained in the Board’s plan meets the federal statutory criteria. Pursuant to this interpretation, the district court determined the exact amounts of the funds in various programs that the Board is entitled to under the consent decree. While a colorable construction of the decree, this result does not comport with the approach we adopted in our previous opinions. The aim of those opinions was to avoid complete judicial usurpation of the parties’ consensual relationship. Only as a last resort, after the parties refuse or are unable to settle their differences, should the court take steps to determine, on its own accord, the proper resolution of this dispute. This litigation has now reached a point where we must abandon any perception that the case is proceeding along on a proper course. In order to correct the situation, we will now provide more exacting guidelines to advance further judicial action. Accordingly, the order of the district court is vacated and the case is remanded for further proceedings consistent with this opinion. I. The origins of this dispute are well-documented in previous district court and appellate opinions and need not be elaborated here. See United States v. Board of Education of the City of Chicago, 744 F.2d 1300 (7th Cir.1984) (vacating the district court’s order appearing at 567 F.Supp. 272 (N.D.Ill.1983)); United States v. Board of *284 Education of the City of Chicago, 717 F.2d 378 (7th Cir.1983) (vacating the district court’s order appearing at 554 F.Supp. 912 (N.D.Ill.1983)); United States v. Board of Education of the City of Chicago, 621 F.Supp. 1296 (N.D.Ill.1985) (the district court order being reviewed in the present appeal). In its simplest form, this case revolves around the nature of the United States’ obligation to provide funds under ¶ 15.1 of a consent decree entered into on September 24, 1980. This provision does not propose a detailed plan of financing, but merely provides that: Each party is obligated to make every good faith effort to find and provide every available form of financial resources adequate for the implementation of the desegregation plan. The remand instructions of the second appellate opinion established that defining the parameters of ¶ 15.1 was the central issue in the case. 744 F.2d at 1304-07. In our first opinion, we held that U 15.1 imposed a “substantial obligation” on the United States Department of Education. The government’s position that its duties were limited to assisting the Board in the grant application process was rejected. 717 F.2d at 382-83. The obligations inherent in the funding provision were left unarticulated in the hope that the parties, prodded by the court, would provide the meaning of the decree they drafted. “[T]he process of dispute resolution ... failed remarkably in this case” 2 and the dispute came before this court again for review of a district court order which, among other things, interpreted ¶ 15.1 as imposing on the government an obligation to provide all the funds required by the desegregation plan that the Board was unable to obtain from other sources. Having declined in the 1983 appeal to adopt the government’s exceedingly narrow construction of ¶ 15.1, this court in 1984 rejected the opposite extreme as embodied in the district court’s order. In reaching this conclusion, we found that there was nothing in the language of ¶ 15.1, or the extrinsic evidence regarding its adoption, that required the Department of Education to provide the Board with all its needs regardless of the availability of funds for which the plan qualified or to seek legislation providing funds for the Board. 744 F.2d at 1306. Our opinion went a step further, based upon representations of the government’s counsel at oral argument, and instructed the district court that if the Board was receiving on a priority basis the maximum level of funding available from existing desegregation programs under established program criteria, the government then would be in compliance with ¶ 15.1. 744 F.2d at 1305-06. Following the second remand, the basic relationship, or lack thereof, between the parties remained the same while their exact positions shifted slightly. The government took the position that the second appellate opinion constituted an adoption of its November 10, 1983 plan as described by the Assistant Attorney General during oral argument. 3 Under the government’s plan, as *285 it emerged from oral argument during the 1984 appeal, the Board would be given “top of the list” priority with respect to funding programs specifically designated for desegregation purposes — namely, the grants available pursuant to Title IV of the Civil Rights Act of 1964 and monies not otherwise obligated in the Secretary of Education’s “Discretionary Fund.” In fiscal year 1984, the government therefore attempted to fund the Board pursuant to this interpretation of ¶ 15.1. The nature and level of funding provided by the government to the Board were the subject of the district court’s latest opinion. The district court rejected the position that the second appellate opinion had adopted the November 10, 1983 plan and thus had resolved the dispute over the meaning of ¶ 15.1. The district court took the position that the two appellate opinions had established broad limits concerning the scope of ¶ 15.1, but that the definitive meaning of the provision still awaited judicial resolution. The instructions on remand did not define the terms “priority” or “available desegregation funds” and thus left unresolved, according to the district court, substantial interpretive issues. In defining the meaning of ¶ 15.1 and the remand instructions, the district court adopted a tripartite analysis. The issues in the case were divided into: (1) scope — this concerns questions of which funds within the Department of Education budget are “available” for desegregation purposes and thus subject to the priority under U 15.1; (2) pipeline — this concerns at what stage along the appropriation process, from Congress to eventual disbursement to earmarked grantees, must the Secretary consider the Board’s 1115.1 priority; (3) share —this is the derivative issue of what amount of funds the Board is entitled to given the determinations made by the court on the scope and pipeline issues. With respect to the scope issue, the district court adopted an expansive interpretation, rejecting the government’s position that the “existing school desegregation programs” language in the second appellate opinion limited the availability to Title IV and the Discretionary Fund. The court ruled that “[t]he 1115.1 priority extends to any statutory program which could materially further the Board’s desegregation plan, so long as a project in that Plan may qualify for funding under relevant statutory criteria.” 621 F.Supp. at 1310. This conclusion was based on the language of U 15.1, the history of the decree’s negotiations, and the nature of the desegregation plan contemplated by the decree. The court noted that H 15.1 contains no restrictions as to what funds are within its scope; thus a review of the extrinsic evidence was necessary to determine the intent of the parties. With respect to the intent of the government, the district court relied on statements made by the government during the course of negotiations to the effect that funding for the decree was contemplated from diverse programs within the Department of Education as well as other federal *286 agencies. In the view of the district court, these manifestations of intent were not surprising given the unique nature of the desegregation plan. Unlike past programs, the one contemplated by the decree focused on encouraging voluntary desegregation through new educational programs and a “compensatory” educational program for those schools that, due to the demographics of Chicago, must remain segregated. The district court reasoned that it could not have been the intent of the parties to approve sweeping educational reform encompassing a vast number of potential programs while at the same time limiting federal cooperation in this expensive program to those funds specifically appropriated with traditional desegregation models in mind. The district court found that the second appellate opinion failed to resolve the pipeline issue just as it left open the scope of 1115.1. Thus, the district court treated the pipeline issue essentially as one of first impression. Clearly, this court had established that 1115.1 contemplated some kind of priority and that the priority did not extend to forcing the executive branch to lobby Congress for appropriations. Outside these broad parameters, however, the district court believed it was unencumbered by precedent. The district court thus rejected the government’s position that its November 1983 plan with its “top of the list” priority was embraced by the Seventh Circuit in the 1984 opinion. This court had clearly stated that the decree could not be read to require lobbying or serve as a basis for disregarding congressional restrictions on appropriations. The second appellate opinion, however, in the view of the district court, was less clear about the effect of 1115.1 on administrative regulations that restricted the availability of funds for which the Board could qualify under the broader congressional standards. Had this court believed that the Department of Education was free to create program criteria and administrative priorities without consideration of 1115.1, thus leaving the Board to priority treatment with respect to what remained available, the second appellate opinion, according to the district court, would have so stated and unequivocably adopted the government’s position. The district court proceeded to reject the government’s position on its merits, finding that the “top of the list” priority would leave the Board at the mercy of the government since the government would have complete discretion to determine the scope of its performance. This dependent status arises under the government’s theory because the priority only applies to available funds and the “availability” of funds is determined without regard for the government’s obligations under the decree. Thus, according to the district court, the government under its proposed interpretation would possess complete discretion, through the administrative balancing of competing policies, to decide how much money, if any, the Board receives. The district court found this to be untenable and determined that H 15.1 requires that the Department consider the Board’s priority at the time it receives the funds from Congress rather than after it administratively determines how to allocate the appropriated funds among the various programs. The government now characterizes this as an “off the top” priority that, when combined with the holding on the scope issue, allows the Board to receive some unspecified share of the monies in any fund under which the statutory criteria are satisfied, regardless of other priorities or the needs of competing programs or grantees. This is not, however, an accurate statement of the district court’s holding. That court scrupulously precluded such a reading of its holding by stressing that the Board’s priority is not absolute and must be balanced along with the other competing priorities faced by the Department. Under this formulation, the government still retains discretion to organize its priorities to maximize the dollars provided to the Board without seriously damaging other programs. Thus, the Board “gets priority consideration as to all funds, but it actually *287 receives nothing close to all the funds.” 621 F.Supp. at 1320 (emphasis in original). The final, and in many ways the most difficult, issue is the question of how to apply the scope and pipeline holdings to determine judicially, as opposed to administratively, the Board’s “share.” The district court made extensive findings of fact concerning the government’s attempts to finance the desegregation plan. From these facts, the district court determined that the government was not in compliance with 1115.1 as it was construed in that court’s opinion. Significantly, the court refrained from describing the government’s conduct as constituting bad faith. 4 Rather, the focus of the opinion was on the government’s treatment of the 1115.1 priority with respect to specified sources, namely Title IV, the Discretionary Fund, the Block Grant and Follow-Through programs, and certain excess funds within the Department of Education’s budget, as well as the government’s attempt to search through the budget for other sources of funds. In none of these instances did the district court find that the amount received by the Board was equivalent to its “equitable fair share” as determined through the application of the pipeline and scope holdings. As that court stated, in order to achieve this level of funding: [The Board] must receive its ‘equitable fair share,’ an amount which accounts for its enormous needs under the Consent Decree against the needs of other grantees and other interests. The outcome of this balance must yield ‘maximum funding,’ that is, the most funding possible without unduly harming other grantees or crippling the Secretary’s discretion. Where statutory criteria set a maximum for any new grant ... the Board should receive this ‘maximum’ level of available funding. Where there is no statutory maximum, our resolution of the pipeline issue demands that the Board be considered from beginning of the funding process as a significant priority among others. 621 F.Supp. at 1436. Finding that the government had repeatedly violated the consent decree in its funding of the Board for fiscal year 1984, the district court took it upon itself to determine the Board’s fair share under the formula laid out in the 1984 opinion and express that formulation in a remedial order. The district court then went through the various programs and made rough approximations of the Board’s fair share that would be finalized after the parties had an opportunity to respond to the opinion. These determinations were an attempt to maximize the amount provided to the Board while minimizing the impact, viewed by the percentage of available funds affected, on other grantees. Pursuant to this analysis, the district court determined that the Board should receive: (1) $2.5-$3.0 million of available Title IV funds (10-12% of all available Title IV funds) as opposed to the $428,573 earmarked for the Board in the block grants to the Illinois State Board of *288 Education and various Desegregation Assistance Centers; 5 (2) $1.5 million of the monies in the Secretary’s Discretionary Fund, an amount representing less than 10% of the entire Fund and more than 40% of the residual amount remaining following the funding of other priorities (this was used for the Unsolicited Grants Competition under which the average grant was $100,000); (3) $1 million of Follow Through funds of which the Board received only $273,541 in fiscal year 1984; this represented a 5-10% reduction in the amounts received by other grantees; (4) $4,092,935 in Excess Funds as to which the Board has eligible unfunded programs; this represents 100% of all such funds; (5) any amount of excess funds for which a Board program meets the statutory criteria but not the nonbinding line item appropriation that can be successfully re-programmed through consultation with Congress. 6 In addition the district court ordered that the government take further steps to find other sources of funds within the executive branch and also to compel the State of Illinois to provide the Board with a greater percentage of the state’s Chapter 2 block grant funds. Because of the government’s ongoing violations of the decree, the district court ruled that the obligations imposed by If 15.1 would be viewed as commencing during the 1985-86 school year and continue for five years. Both sides appeal from this decision. The United States claims that the district court misconstrued the decree, this court’s opinions, and the extent and nature of the limits placed on the Secretary by Congress. The Board does not dispute the district court’s methodology; its claims are limited to contesting as inadequate the dollar amounts that the court determined the Board should receive through the various programs. II. The story of school desegregation in Chicago can best be described as ironic: litigation concerning the meaning of a consent decree that was designed to avoid protracted litigation continues six years after the decree was approved by the district court. This court twice has vacated attempts by the district court to enforce “definitive” resolutions of 1Í 15.1, but these appellate opinions have done little to move the parties toward any sort of consensus. Both parties maintain rigid positions, based on either the most or least restrictive interpretations of 1115.1 and the appellate opinions, in the hopes of achieving “total victory.” The district court has been receptive to these tactics by accepting the premise that federal-municipal dialogue is not productive and therefore that the courts must assume control of the relationship established under the consent decree. The present situation, while possibly inevitable in light of the shifting ideological winds in the years since the decree was signed, finds no support in the opinions of this court. The conduct of the parties stands in derogation of our concept of the meaning of the decree and has caused an inappropriate judicial reaction to this litigation. Paragraph 15.1 cannot be read to be *289 any more exact than to encompass a broad range of possible funding arrangements and amounts, any of which could be arrived at by the government’s exercise of good faith in the process of considering its obligation to the Board vis-a-vis other funding priorities. Given the inherently nebulous nature of this provision, the parties have a vital, and in fact preeminent, role in constructing a specific type of performance that complies with the general guidelines arising from a good faith standard. The corresponding role of the district court is to set forth the parameters of compliance and give the parties a chance to adjust their conduct accordingly. The district court maintained the position that, on its own, it was able to establish substantial and specific remedial relief out of the vagueness inherent in 1115.1. Therefore, we must do here what we have been reluctant to do in the past: set forth the parameters of compliance and thus inform the district court and the litigants, all of whom bear certain responsibilities for the creation and implementation of the decree, what their respective roles and obligations are. A. At the time it was agreed to, 1115.1 may have appeared to be an appropriate way to settle a critical long term dispute. Regardless of one’s views on the wisdom of that initial decision, hindsight has made it relatively clear that the decree has become an inadequate solution to a serious problem. Nevertheless, 1115.1 carries for the time being, see infra Part IV, the full weight of law, and we thus are obligated to provide a framework for resolving the disputes arising from this litigation. The first step is to delineate the role of the district court in interpreting 1115.1 under our two previous opinions. Both parties and the district court read the second appellate opinion as supporting the tripartite “scope-pipeline-share” analysis adopted below. The dispute between the parties arises because the government takes the position that this court not only inferentially established the mode of analysis, but also decided the issue of interpretation. This is based primarily on our finding that “by guaranteeing that the Board will be funded on a priority basis under existing school desegregation programs, the amount of which funding is determined by program criteria and is subject to the review of the district court, the government would comply with our interpretation of 1115.1.” 744 F.2d at 1305-06. A view that this language was intended to serve as the basis of a mode of interpretation, let alone a final resolution of the issue, seriously misconstrues our finding by ignoring the central themes underlying the two appellate opinions. In the first appeal, we held that the government violated the decree based on a finding that the failure to provide available funds constituted bad faith. Significantly, the first appeal emphasized that this breach did not put the case in a remedial posture. Rather, the government had to be given a sufficient chance to bring itself into compliance with 1115.1, in part because of the lack of any direction as to the meaning of the provision and in part because of the inherent problems with ordering remedial relief against the government. The opinion did not define the parameters of compliance: while the government’s conduct constituted bad faith under 1115.1, what constituted good faith was not readily apparent. The issue was left open based on the assumption that the district judge, as the judicial officer who approved the decree, should make the initial attempt at outlining what compliance requires. The district court on remand chose not to follow the compliance/remedial distinction, finding a bad faith violation of 1115.1 and imposing the identical remedy presented to this court in the first appeal. The second appellate opinion then rejected this conception of the case, vacating both the finding of bad faith and the proposed remedy, 7 since the case still was in the compliance *290 stage. We noted in our second opinion that the government had made some limited steps toward honoring 1115.1 through its “top of the list” priority and that the government’s program could constitute good faith performance if it represented the government’s best efforts to supply the Board with the maximum amount of funding available. The case was remanded for a second time in order to have the government’s new proposals-scrutinized under the good faith standard of U 15.1. Thus, the litigation left this court in 1984 in a compliance, as opposed to remedial, posture; the dominant issue at this point being the nature of the government’s obligation, rather than the remedial powers of the district court in the event those obligations are not honored. The court below appears to have followed the spirit of this opinion by sketching out in detail the requirements of compliance under H 15.1. The district court, however, imposed a remedy in accord with the parties’ wishes that the court take primary responsibility for resolving all differences. The district court described its view of what compliance entailed and then found that 1115.1 had been violated by the government without providing any time for the adjustment of conduct in light of the opinion. Significantly absent was a finding of bad faith, a concept that had been vital to the imposition of any remedial relief in the previous opinions of this court. Throughout this litigation, the attitude of the parties as reflected in the record has been that it is the responsibility of the court to determine the correct interpretation of a decree drafted by the parties, albeit with judicial approval. Thus, the district court has assumed the role of an arbitrator in this dispute, alleviating the parties of the primary responsibility of conciliation and allowing them to vigorously and rigidly argue for their own “legal” positions. Such a division of responsibilities has proven inappropriate for the ultimate resolution of this case. A consent decree is a unique hybrid of non-judicial settlement and court order. See United States v. City of Chicago, 717 F.2d 378, 382 (7th Cir.1983). As such, both the court and the parties share the responsibility for their creation. This particular decree requires for its ultimate success cooperation between federal officials and the Chicago School Board. Unfortunately, somewhere during the course of the last six years, the consensual nature so necessary to the 1980 plan disappeared, leaving a void into which the district court, in protection of its order, has stepped. Abandonment of the consensual aspect of a decree may be appropriate in certain situations, but given the stature of the signatories and the scope of the agreement in terms of students, dollars, and time, it was, and still is, imperative that the district court make every effort to settle this dispute before imposing its own determination of appropriate remedial relief. As we made clear in the previous opinions, the district court is not without weapons to advance the parties into a dialogue. Paragraph 15.1 certainly requires good faith efforts by both sides. Good faith is not unilateral; it requires each party’s best efforts in conjunction with, and in consultation with, its opposite. In deciding whether bad faith exists, one crucial factor would be the willingness of a party to enter into discussions designed to resolve the dispute; another would be the conduct of the party in such negotiations. A party’s refusal to bargain in good faith can serve as the basis for a court-imposed resolution of the case. By accepting the parties’ position that judicial resolution represents the only hope for an end to the dispute, the district court did not consider this crucial intermediate step, and thus did not pursue the more restrained possibility of first providing the parties with guidance as to what is acceptable under U 15.1. B. The absence of judicial restraint is further evidenced by the specifics of the district court’s interpretation of what compliance with H 15.1 requires. While the district court’s tripartite analysis is both so *291 phisticated and thorough, it does not find sufficient support in the record. It stands as a plausible compromise between the parties’ positions, but a “fair” middle ground cannot be imposed on the parties unless there is a legal basis supporting it. There is little in the ambiguous language of 1Í 15.1, see Board of Education I, 717 F.2d at 382, or the extrinsic evidence surrounding its negotiation, that provides any basis for specific answers to the questions raised by the scope, pipeline, or share issues. The only thing that was clearly contemplated by both the parties and the court in November 1980 was that a cooperative Department of Education would make a good faith effort to provide the Board with a substantial amount of funding. This type of “best efforts” clause does not mandate specific sources of funds. As was conceded by counsel for the Board at oral argument, there exists a wide “band” of amounts and methods of funding that could satisfy U 15.1. Clearly, the government retains discretion under the decree to choose among the possible combinations of sources of funds. The consent decree does not limit this discretion; the only limits are those imposed by Congress and fiscal realities. The decree itself allows for its “scope” to be as broad as declared by the district court or as narrow as argued for by the government with no meaningful basis for choosing between the extremes. Similarly, 1115.1 says nothing about any specific type of priority consideration — the “pipeline” issue. The entire question of priorities arose not out of the decree but out of the government’s November 10 plan and the second appellate opinion. The decree contemplates some type of priority for the Board, but is notably silent as to the nature of that priority vis-a-vis other departmental priorities. Once again this creates the possibility of a number of “priority treatments” that could satisfy the terms of the decree. The district court arrived at its conclusion regarding priorities, as well as scope, by noting that the Department of Education’s “top of the list” priority had failed to ensure adequate funds, and then reasoning because of this that the decree could not have contemplated the government’s approach. This logic, however, is not persuasive. It is possible that given a different attitude and fiscal setting within the Department, the government’s “top of the list” priority could result in substantial funding within the meaning of 1115.1. Thus, it cannot be said that 1115.1 “means” what the district court claims it does. The consent decree, like a contract, cannot be interpreted by taking a hindsight view of conduct that is facially in compliance but produces an unsatisfying result and deciding therefore that the agreement must mean something more. When an agreement provides for the type of flexibility in performance inherent in a “best efforts” clause, the court and the parties must accept the inherent uncertainty with respect to the specifics of performance, or void the agreement on the grounds of vagueness. See infra, Part IV. The arbitrariness inherent in the district court’s interpretation of 1115.1 is further, and perhaps most vividly, illustrated by its resolution of the share issue — the application of the scope and pipeline holdings. Having created a detailed analytical framework for interpreting the decree, the district court’s attempt to apply it resulted in a rough approximation of what the government can afford without significantly injuring other grantees. The figures arrived at reflect no findings concerning the relative needs of competing grantees, particularly the effect of a specified percentage reduction on those grantees. Without a detailed knowledge of other grantee’s needs, a Herculean task in and of itself, the district court’s conclusion that others would not be significantly injured is without basis. Moreover, these numbers do not necessarily reflect consideration of the Board’s needs: while the amount awarded was substantially more than originally provided by the government for fiscal year 1984, it is but a fraction of the $103 million that Judge Shadur found the Board needed. While the remedy from an equitable standpoint may be Solomonic, when viewed as an *292 exercise in contract interpretation, it is without foundation. The potentially arbitrary result reached by the district court is in any event premature. In fashioning “make whole” remedies for breaches of ambiguous “best efforts” clauses, courts may often have to be flexible in determining, the amount of performance that would constitute best efforts during the period in question. The primary flaw in the district court’s decision is that the result is grounded not only in the power of the court to remedy a breach of the decree but also on the premise that these amounts represent the only possible compliance with H 15.1. In other words, the remedy is not cast in terms of the district court’s concept of what level of performance constitutes good faith; the court expressly declined to reach this issue. Instead through its remedy, the district court in essence declared that there existed an identifiable level and type of performance under the decree. This labelling is inconsistent with our holding that if 15.1 requires best efforts, a standard that by definition contemplates a broad range of possible funding arrangements. Thus, we cannot affirm the district court’s resolution of the share issue not only because it is premature, if not arbitrary, as a remedy (see supra, Part IIA), but also because it is derived from an unwarranted construction of what compliance with If 15.1 requires. C. We then must proceed to describe what compliance with If 15.1 entails. Paragraph 15.1 requires the government to make every good faith effort to find and provide funds for the Board’s use. As with any best efforts clause, If 15.1 can be satisfied by any of a wide range of possible levels and types of performance that comport with the exercise of “good faith” by the obligor. 8 In this sense, the Board did not receive any commitment for a specified type of consideration by the government, let alone a specified amount of funding. What the Board “contracted” for was a process —a process under which the Board receives the maximum amount of funds the government, acting in good faith, determines it can provide. Thus, the only standard involved in evaluating the government’s actions is the Department of Education’s good faith. The only funding program constituting a breach or violation of 1f 15.1 is one proposed in bad faith, which can be defined as attempting to avoid or minimize the government’s obligation under the decree. Good faith, however, is not a term that exists in a vacuum. The nature and circumstances of the underlying obligation help to determine what constitutes good faith. In the present case, the obligation to make a good faith effort to provide funds is placed on an administrative agency, the Department of Education. 9 As such, best *293 efforts should be defined with due regard for administrative decisionmaking. The issue here is not whether the Department has a duty to obtain funds for the Board; this was resolved in the first appellate opinion. Rather, the obligation of the Department relates to its treatment of the Board in the process of allocating funds in its budget. This allocation process is far from entirely discretionary. Congressional appropriations carry with them certain express limitations, as well as statements of congressional desires to have certain programs funded. Moreover, the Secretary is subject to constraints imposed by the Department’s own administrative regulations. Nevertheless, the Secretary does have a substantial degree of discretion with respect to the funding of certain programs and the funding of certain grantees within specific programs. The consent decree serves as an additional constraint on this discretion. See Board of Education II, 744 F.2d at 1306 n. 7. What this means is that the government has obligated itself to consider the Board’s desegregation plan in apportioning its budget. The Department of Education must seek to maximize the amount of funds going to the Board in the same way it would seek to maximize the funds going to any one of a large number of priorities competing for their shares of a limited budget. 10 The Department naturally makes decisions concerning the relative importance of each program and those programs deemed less important receive a lesser share. This is the nature of the administrative process and the decree does nothing to change it. The relative priority the Secretary attaches to the desegregation plan in the budgetary process is an administrative decision entitled to deference, in the absence of an intent to avoid the 1115.1 obligation as evidenced by manifestations to that effect or a method of funding that by its very nature *294 reflects an absence of serious consideration of the Board’s needs. In essence, good faith requires the government to consider the II 15.1 obligation in exactly the same way as it would treat any other funding priority it had with regard to the monies available under a program for which the particular priority is eligible. See supra note 10. Notwithstanding the government’s reading of the second appellate opinion, there is nothing in K 15.1 that necessarily limits the programs from which funds can be provided to those earmarked for desegregation purposes. 11 In order to give the obligations under the decree full good faith consideration, the government must evaluate the possibility of funding any element of the Board’s plan under any source of funds for which that element meets the statutory and administrative criteria. 12 This does not mean that the Board must receive funding from non-desegregation programs, but the government can no longer cling to the position that such programs stand outside the reach of 1115.1. The decree contains no requirements or prohibitions on sources of funding. The government has the burden of creating a funding package that represents the maximum it can provide the Board, using all sources that the Department, through its administrative process, determines are available for this purpose under existing criteria and with due regard for the needs of other funding priorities. The district court’s corresponding burden is to determine whether the government accorded the desegregation plan full and fair administrative consideration in an effort to maximize the amount of funds provided to the Board. Significantly, there is no place under this standard for residual funding — funding in which the government makes its budgetary decisions as if the decree did not exist, and after meeting all other needs, gives the Board whatever is left over. 13 “Best efforts,” by definition, requires an affirmative attempt to provide funds. Paragraph 15.1 thus cannot be considered a secondary obligation outside the normal funding process. By signing the decree the United States obligated itself to integrate the Chicago desegregation plan into the fabric of its adminis *295 trative system of distributing the monies contained in the budget. All of this leaves a wide range of acceptable funding arrangements. The goal of the remand in the second appellate opinion was for the district court to evaluate whether the “priority treatment” discussed in the course of the second appeal, see supra note 10, fell within the range of acceptable funding possibilities. Clearly, the scheme the government proposed could be within If 15.1 if it provided the maximum amount of funding under existing program criteria. 14 Similarly, the funding arrangement contained in the district court order represents an acceptable alternative vision of the process in which the Secretary gives greater weight to the decree vis-a-vis other funding priorities. These are potential examples of good faith in this setting. There is an additional aspect of good faith under the consent decree which should be emphasized. The decree contemplates a broad plan designed to alleviate the problems of racial segregation in one of the nation’s largest public school systems. This type of effort was, and continues to be, beyond the means of the Board. The problem was solved, at least in part, by federal cooperation. Under the decree the government is more than a conscripted supplier of dollars that passively receives and evaluates requests for funds. Rather, it is an integral part of the plan contemplated by the decree. Although at this point successful cooperation may be a pipedream, 1115.1 requires at a minimum that both parties attempt to establish a federal-municipal dialogue concerning the needs of the Board and the resources available in the federal system. The government cannot view the Board as just another grantee whose needs happen to be more acute: the government has a substantial obligation to the Board that goes beyond, for example, what it has to other grantees under Title IV. Similarly, the Board cannot devise its plan based on the notion that the government has an obligation to fund all educational programs that the Board on its own initiative decides to include under the label of desegregation. Neither this court nor the district court will further tolerate the parties’ position that one of them must emerge the victor. Paragraph 15.1 imposes a continuing relationship between the parties, the responsibility for which cannot be avoided through litigation over the precise obligations imposed by the funding provision. This type of rigidity, in the absence of any attempt at amicable settlement, is itself a violation of the good faith/best efforts required by H 15.1. We are not unmindful of the record the United States has built before the district court. 15 For purposes of remand on the issue of good faith, however, the past history of this litigation, while relevant for establishing a factual background and for evidencing the government’s attitude toward the decree, is not dispositive. The focus should be on the parties’ response to this opinion and the nature of any funding arrangement that is presented to the district court. Thus, following a reasonable period during which the parties, with the *296 district court’s guidance, can attempt to comply with this opinion, the district court should review whatever solution the government presents, whether old or new, under the good faith standard. This entails close examination of the nature and amount of the funding; the extent to which the 1115.1 obligation received full consideration in the bureaucratic process; the conduct and attitudes of the parties in the wake of this opinion; and any manifestations of intent to avoid the responsibilities accepted in 1980. While the standard of review set forth here is necessarily cast in general terms, the district court is sufficiently familiar with the review of administrative discretion and the application of good faith standards to allow it meaningfully to exercise its oversight powers without supplanting the continuing responsibilities of the parties. See generally Camp v. Pitts, 411 U.S. 138, 93 S.Ct. 1241, 36 L.Ed.2d 106 (1973); Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971). III. Part of the role of the district court as set forth in Part II is to step into the breach should this procedure go asunder. A finding of bad faith by the district court cannot be based solely on the nature of the funding and the record developed by the parties. 16 Should the court find reason to call into question the good faith of either party, appropriate evidentiary hearings must be held; Since good faith is in an important sense subjective, it may become necessary for officials of the Department of Education and the Department of Justice or representatives of the Board to give testimony in order to evaluate the administrative or Board treatment of the obligation under the decree. If the district court finds that any party has acted in bad faith, it then must take necessary remedial steps to protect the integrity of its order. See supra, note 15. These measures are important not only because they insure the ultimate success of the decree but because they provide the incentives for respecting the obligation under 1115.1. Should the district court find bad faith, it would have two alternative remedial courses depending on the degree of bad faith. The first, and most drastic, remedy is contempt. Board of Education II, 744 F.2d at 1308. This dramatic remedy is appropriate where an unequivocal judicial command is callously ignored, see Ferrell v. Pierce, 785 F.2d 1372, 1378 (7th Cir.1986), or where steps are taken to subvert the decree. Not every instance of bad faith, however, requires the use of contempt. Given the somewhat nebulous nature of good faith under II 15.1, an attempt to comply could amount to bad faith, while not constituting direct defiance of the orders of the court. The court must distinguish between a failure to comply through inaction or inadequate performance on the one hand, and defiant or subversive actions that constitute contempt on the other. The alternative approach, where the conduct is not so egregious, is for the court to exercise its inherent power to protect the decree by ordering, based on appropriate factual findings concerning the Board’s needs and the government’s means and treatment of other funding priorities, what the district court determines constitutes “best efforts.” In the case of governmental bad faith, this amount would be the maximum amount the government could pay out of available funds under the parameters set forth in Part II of this opinion. The nature of this inquiry would be similar to that undertaken by the district court in its determination of the share issue in its most recent order. 17 The crucial difference *297 is that such an order should be based on the power of the court to fashion a remedy for a breach of the decree rather than an inappropriately specific interpretation of H 15.1. In the former situation the approximations required in establishing amounts are less troubling because they carry less impact as to the precise obligation of the government in subsequent years. It is important to emphasize that the district court’s role in crafting such a remedial order is neither to punish the government nor present a windfall to the Board. Rather, the goal must be to determine how much the Board could have received with due regard for statutory, regulatory, and prudential guidelines imposed on or enacted by the Department of Education, budgetary limitations, the needs of the Board, and the needs of competing funding priorities. IV. As a final matter, the district court on remand may find it advisable to reevaluate the consent decree. If the decree is violated by the bad faith actions of one party, the court must take actions to protect the integrity of the court order embodied in the decree. In the absence of bad faith, however, the court may find that a result acceptable under 1115.1 is insufficient to allow the Board to fulfill the desegregation and remedial education goals of the decree, or that despite the parties’ efforts, the problems inherent in the decree are too intractable for judicially supervised resolution. The affirmative duties placed on a party by a consent decree are inevitably based on predictions of achievability. These predictions can, and often are, shown to be unworkable in practice despite the good faith efforts of the parties. Philadelphia Welfare Rights Organization v. Shapp, 602 F.2d 1114, 1120-21 (3d Cir.1979). It is in such settings that the court can on its own motion vacate the decree pursuant to Rule 60(b)(5). See, e.g., United States v. Swift & Co., 286 U.S. 106, 114, 52 S.Ct. 460, 462, 76 L.Ed. 999 (1932); Duran v. Elrod, 760 F.2d 756, 758 (7th Cir.1985); Alliance to End Repression v. City of Chicago, 742 F.2d 1007, 1020 (7th Cir.1984) (en banc) (dictum); Philadelphia Welfare, 602 F.2d at 1120-21; J. Moore, 7 Moore’s Federal Practice K 60.26 (1985). In hindsight, entry into and approval of a decree that contemplates wide-reaching and expensive changes in a large, financially strapped school system without explicit commitments for outside funding by the federal government may have been ill-advised. Moreover, the last six years have demonstrated that consensual, rather than judicial, solutions are not easily available. While the decision of one of the signatories to ignore its obligations cannot serve as the basis for finding the decree to be impractical, the past can and should be undone where, despite good faith attempts to resolve the issue, no amount of funding sufficient to implement the Board’s plan at an acceptable level can be provided. In this event the court has two options. First, it can reevaluate the current racial and educational situation in the Chicago public schools and, in conjunction with the parties, attempt to modify the decree so as to align costs with available funds while maintaining the goals of the initial decree. Second, if there is no readily identifiable alternative to the Board’s current plan, and it may be unlikely that such an alternative exists, the court should consider vacating *298 the decree, releasing both parties from the obligations thereunder. Such a course of action would reactivate the government’s 1980 complaint charging the Board with constitutional violations arising out of its schools’ segregated conditions. The court then would have to take immediate steps to set the matter for trial. “When an equity decree affects other people besides the parties to it, the judge must take account of the interest of those people — the public interest — in his decision whether to grant or deny equitable re-lief____ This is true whether the judge is being asked to approve a decree ..., or interpret a decree ..., or, it seems evident, modify a decree.” Duran, 760 F.2d at 759 (citations omitted). If this decree is not capable of assuring the school children of Chicago the rights they allegedly are being deprived of, it is manifestly unjust to require the taxpayers to continue funding both sides of this protracted litigation. Should there be unconstitutional conditions in the Chicago public schools (notably this has not been admitted by the Board), the public school children of Chicago may deserve more than this decree may be able to provide. V. For the reasons stated above, the order of the district court is vacated and remanded for proceedings consistent with this opinion. Accordingly, on remand the district court should take the following steps: 1. The court should allow sufficient time for the government, acting in conjunction with the Board, to craft a funding arrangement that falls within the range of possible good faith resolutions under 1115.-1, as defined by this opinion. 2. Should the district court find grounds to question either the government’s or the Board’s good faith, the court should conduct evidentiary hearings regarding the good faith afforded the decree in administrative funding allocations. 3. Should the court find that either party acted in bad faith in accord with the parameters set out in this opinion, it should: a. issue contempt citations against appropriate officials if the Board’s or the government’s conduct constitutes an attempt to defy or subvert the court’s order; or b. in the case of bad faith by the government, craft remedial relief that, to the best of the court’s abilities, approximates what the Board would have received if the government had exercised good faith in allocating funding. This remedy is appropriate where the government’s breach consists of inadequate performance rather than more egregious conduct. 4. Should the court find that an otherwise appropriate funding arrangement under the decree is inadequate in light of the Board’s needs under the desegregation plan, or should it otherwise conclude that the problems vexing enforcement of the decree are insoluable, it should take measures to vacate the decree and set the underlying matter of segregation in the Chicago schools for immediate trial. Circuit Rule 18 shall apply on remand. Footnotes [1] In a recent order the district court described the current state of relations between the parties: We have commented earlier on the parties’ bitterness toward each other, which persists to some extent despite their ability to reach some agreements. If the parties are ever to work smoothly together, some of their attitudes and assumptions must change. The United States seems ever-ready to attack a Board proposal rather than eager to work with the Board in designing a reasonable proposal; it presumes inadequacy, taking the attitude, "What’s wrong with the Board’s proposal," rather than, “How can the Board’s proposal be made workable?" While the government should scrutinize the Board’s proposals carefully, it should he more flexible in its approach. In turn, the Board should move from its presumption that virtually every government criticism is illegitimate and prompted by ill-will. The Board’s suspicion is well founded in the history of the case, but even the Board has admitted that the United States has heen more cooperative lately. Despite its history of possible bad faith, the United States surely has legitimate concerns and has raised some legitimate criticisms. United States v. Board of Education of the City of Chicago, No. 80 C 5124 at p. 10 n. 8 (N.D.Ill. July 17, 1986). [Available on Westlaw, DCTU database.] [2] 744 F.2d at 1304. The reference in the second opinion to the failure of non-judicial resolution was a comment on the failure of the parties and the courts to reach a consensus up to that point in time. It was not a determination that the federal-municipal standoff was irreconcilable by any means other than judicial resolution. The value of the type of consent decree entered here lies in its service as an outline for a future working relationship between its former antagonists. The courts must exercise the utmost restraint to avoid assuming complete control of the relationship in the absence of a finding that agreement is impossible or is precluded by one or both parties’ bad faith. [3] There is some dispute as to whether the representations made by the government on the second appeal can be viewed as part and parcel of the November 10 plan. The November 10 plan provided the Board with a "competitive priority” in desegregation programs that provided operational support for local agencies. At the time there were no such programs. The plan provided no priority with respect to the desegregation funds available in Title IV or the Discretionary Fund. The district court concluded that the government’s offer of "top of the list priority” with respect to Title IV and the Discretionary Fund was a new position taken for purposes of the appeal and was distinct from the November 10 plan. 621 F.Supp. at 1315. The government takes the position that its counsel at oral argument expanded on the November 10 plan *285 and that the plan as now defined was accepted by this court. The government's assumption concerning the effect of the second appeal is unwarranted. Should we have wanted to adopt the oral representations of the government, the opinion would have explicitly embraced these representations either as an adaptation of the November plan or as sui generis. Clearly the second appellate opinion did not do this. See supra, Part II. The government bases its conclusion in part on the responses and questions at oral argument. These statements are used by the government to provide a context for the interpretation of the opinion. This type of interpretation of judicial precedent is unacceptable. This opinion, like any judicial opinion, stands as the sole expression of the court; statements made at oral argument do not determine the scope of an opinion unless explicitly incorporated into the opinion. Thus, while the opinion clearly refers to the government counsel's representations as part of the basis of its conclusion that the Board is entitled to a priority with respect to available desegregation funds, 744 F.2d at 1305, the precise definition of these terms cannot be found by analyzing the transcript of oral argument. [4] While the district court did not base its decision on bad faith, it did acknowledge that bad faith was an alternative grounds for its decision and one that was arguably supported by the record: (b) In addition, this pattern of failure to comply arguably shows that the United States has made every effort to minimize and to avoid its obligations under the Consent Decree and the provision of funding to the Board. This pattern of conduct could constitute a bad faith violation of the Consent Decree under the standards discussed in Part 11(B) of the Court of Appeals’ 1984 Opinion. Accordingly, all of the proposed provisions of this court’s Order described above could, as a supplemental matter, be grounded upon these bad faith violations. Besides the overall pattern of conduct, certain particular actions could constitute bad faith, such as the Yates-Weicker Activities, the activities noted in footnote 12 of the Second Opinion, the conduct surrounding the excess funds, see Conclusions in Chapter 4, and the woeful lack of search activities. For now we express no firm opinion on the bad faith issue, since it has not been fully addressed by the parties. However, the parties should be prepared to discuss the issue at the next status conference. 621 F.Supp. at 1443. [5] The district court held that this amount would be reduced to $1.5 million if the government would acquiesce to a direct grant to the Board rather than a block grant to a grantee who provides services to the Board. The government did agree to this form of payment in the district court’s final remedial order of December 9, 1985. [6] The district court found there to be no conflict with the second appellate opinion’s proscription against requiring legislative initiative by the executive in its requirement of reprogramming. The district court concluded that the separation of powers concerns underlying this court’s conclusion were mooted by the fact that: (1) there are no competing programs for these funds; (2) reprogramming is not formal legislative activity but an informal way of seeking legislative approval of an action that could be taken without consultation with Congress; and (3) these funds were earmarked by non-binding line item appropriations within larger appropriations under whose criteria the Board was eligible. [7] At that time this court invoked Circuit Rule 18 to order the transfer of this case from the district judge who approved the decree and handled this litigation up to the second appeal. [8] Because compliance under a good faith standard cannot be limited to a specific funding arrangement, the most that can be provided by this court’s opinion is general guidelines. Thus, the more restrained role for the district court envisioned by this court is necessitated by the broad language of fí 15.1. It would certainly be easier if the decree could be read to require a specific type and level of performance. We are not so fortunate, however, and thus general guidelines must be set because that is as far as If 15.1 permits us to go. [9] The government in general rather than specifically the Department of Education is the signatory of the decree. Thus the burden of meeting the obligations imposed by f 15.1 is not the sole responsibility of one administrative agency. During the course of this litigation the Department, as the agency generally responsible for federal financing of the type of matter involved here, has logically become the focal point of the funding debate. Clearly the Department must serve as the primary, if not the only, source of federal funds. As to funds outside of the Department’s budget, the second opinion makes clear that ft 15.1 does not reach funds that, consistent with the intent of Congress, cannot be used for the Board’s school desegregation plan. 744 F.2d at 1306. Thus, ff 15.1 does not provide a blanket priority that covers the entire federal budget. Should there exist funds outside the Department which can be used under existing criteria as a source of funding for a legitimate element of the Board’s plan, the government has an obligation, consistent with the good faith standard of this opinion, to search for such funds and to at least consider the Board’s program among the other competing grantees for *293 such funds. See generally notes 10 and 12, infra. [10] The word “priority” like many words in this complex case has acquired its own special meaning during the course of this litigation. In the second appeal when the government referred to priority in its "top of the list” priority, it meant that the Board would receive preferential treatment with regard to specific grant programs. In this sense "priority” was used as an adjective to describe the treatment that the Board would receive relative to other competing grantees. Such “priority treatment” should be a foregone conclusion under jf 15.1. It would be antithetical to the good faith nature of the provision for the government to not put the Board at the top of any list of grantees for funds earmarked for desegregation. Board of Education II, 744 F.2d at 1305 n. 6. When we use the word "priority” in this opinion we mean to use it as a noun. Thus, we mean to refer to the decree as a priority of the Department of Education rather than a method of treatment. In apportioning the discretionary portions of its budget the Department has a number of goals which it seeks to accomplish by both allotting funds to programs and earmarking funds within a given program in a certain manner. Consideration as a goal of the Department, or more accurately a priority of the Department, is analytically distinct from priority treatment vis-a-vis competing grantees. In its simplest form, the government must consider the decree as one of those things that the Department wants to or is obligated to fund rather than a document which makes one grantee’s needs more pressing than another. This does not mean that good faith consideration of the decree as a Departmental priority could not ultimately result in a situation where maximum funding was provided through preferential treatment under existing grant programs. The second appellate opinion clearly accepted this as a viable possibility for compliance with H15.I. These semantic distinctions, while admittedly slight, must be emphasized in order to prevent further distortions of this court’s opinions. This opinion is intentionally broader in scope than our previous opinions. The process of administrative consideration under the decree encompasses a vast array of funding sources and funding arrangements. Nothing in this opinion should be read to be a preference, on behalf of this court, for any particular funding program. Our concern is that the process employed by the government constitutes its "best efforts” to provide all available funding under 1f 15.1. Artic- • ulating the precise parameters of the obligations under the decree is complicated by the hundreds of pages already written in which certain words have taken on their own special meaning. In the end, one of the many things this case illustrates is the ability of the legal mind to creatively redefine the obvious and to draw find semantic distinctions from the general. [11] At the present time those funds that the Secretary has the discretion to use for desegregation purposes are Title IV of the Civil Rights Act of 1964, 42 U.S.C. § 2000c; Chapter 2 of the Education Consolidation and Improvement Act of 1981, 20 U.S.C. §§ 3811-3863 (this includes the so called “Secretary’s Discretionary Fund"); the Magnet Schools Assistance Program of the Education for Economic Security Act, 20 U.S.C. §§ 3911-4074. Clearly, a decree designed to desegregate a school system should be given its greatest weight with respect to funds specifically earmarked for desegregation purposes. This does not mean that aspects of the Board’s plan which qualify under the criteria of non-desegregation programs are not entitled to funds from such programs. It does provide, however, some basis for distinguishing the priority for desegregation funds as opposed to other sources of funding. [12] As was emphasized in the second opinion of this court, 744 F.2d at 1306, an important element of the government’s obligation, and thus crucial to good faith under 1[ 15.1, is to search through its various sources of funds, within the Department of Education and possibly outside it, see supra note 9, for funds that can be made available to the Board. “The best proof that the government is fulfilling this duty would be the assignment of personnel to the task of periodically reviewing federal funding programs, in the Department of Education and in other federal agencies, for unemcumbered funds that may be used to advance the Board’s desegregation plan.” 744 F.2d at 1306-07. We are aware that the district court found the government’s search efforts inadequate. The district court, however, refrained from finding bad faith and, on the basis of this record, we will not reach that conclusion. On remand the district court should reevaluate the government’s search efforts in the broader context of reviewing the entirety of the government’s response to its obligation under a good faith standard. [13] The district court found that this was precisely the type of funding scheme used by the government with respect to Title IV funds. We have no cause to review this finding since it, like all factual findings by the district court, is vacated by this opinion. New factual findings and conclusions of law based on this opinion are required. Therefore, we express no opinion on the government’s past funding methods except to note that residual funding does not constitute good faith under f 15.1. [14] It is important to note that the good faith obligation applies to the criteria that the Department sets up to evaluate competing priorities. The government can set up general "across-the-board” criteria, but requirements that are designed to defeat the Board’s priority to certain funds would constitute bad faith under f 15.1. [15] In dealing primarily with the possibility of bad faith by the government we do not mean to trivialize the potential for bad faith on the part of the Board. It should not be forgotten that the Board is not an innocent before the courts. The root of this action is the Board’s action, or lack thereof, with respect to the nagging racial problems in the school system. While this in no way detracts from any failure of the federal government to honor its obligations under the decree, it evidences the fact that the Board has primary responsibility for the funding and implementation of its plan. The district court, thus, must also exercise its oversight functions with respect to the Board. The Board’s good faith in searching for all available non-federal sources of funds, in allocating available funds to the most vital programs, in searching for the most economic way of accomplishing the decree’s goals, and in limiting the programs under the plan to those which are necessary to materially advance the plan’s purpose should be monitored. [16] It is important to reemphasize that a determination of bad faith by the government cannot be based on the administration’s public manifestations of an ideological position hostile to the goals embodied in the decree or the pursuit or enactment of general policies that, although deleterious to the Board, are designed to apply uniformly. See Board of Education II, 744 F.2d at 1307. Direct manifestations of bad faith must be specifically directed at the decree. [17] One of the most troubling aspects of the district court’s opinion on remand is the ques *297 tion of the availability of excess funds which can be reprogrammed to allow their receipt by the Board. Excess funds in programs that could have been used to fund the Board’s desegregation plan, under the approach adopted in this opinion, are evidence of the government’s lack of good faith. The reprogramming issue is more difficult. The district court clearly is correct in distinguishing the type of reprogramming disallowed in the second opinion from the reprogramming of excess funds for which no competing grantees exist. This does not mean, however, that there are not legitimate reasons, apart from animus to the decree, for refraining from seeking the reprogramming of, or from simply ignoring, line item restrictions enacted by Congress. At this time we do not have an adequate record concerning the nature and past use of reprogramming in such a setting to definitively determine that failing to go before Congress for this purpose constitutes bad faith.