Since its genesis, this litigation has sought to eradicate the effects of an official policy of racial segregation in the public schools of Oklahoma City, Oklahoma, and assure that each child enrolled in an Oklahoma City school enjoys the same right to a public education. We are now at a crossroad in the substantive and procedural life of this case and must decide whether, after our last remand, the district court followed the correct path, terminating its prior decree and finding a new student assignment plan implemented under that decree constitutional. Dowell v. Board of Educ. of Okla. City Pub. Schools, 677 F.Supp. 1503 (W.D.Okla.1987). We approach this case not so much as one dealing with desegregation, but as one dealing with the proper application of the federal law on injunctive remedies. We believe that the law in this area is unambiguous, and simply because the roots of the matter lie in school desegregation, there is no reason to depart from the longstanding principles which form the structure of that law. Upon our review, we conclude the trial court did not follow the proper path and reverse the judgment dissolving the 1972 injunctive decree. We remand the case for modification of the decree consistent with this order.

We have previously summarized the history of this case, Dowell v. Board of Educ. of Okla. City, 795 F.2d 1516, 1517, n. 1 (10th Cir.), cert. denied, 479 U.S. 938, 107 S.Ct. 420, 93 L.Ed.2d 370 (1986), tracing its metamorphosis from filing in 1961 to the generation of an equitable remedy in 1972. Dowell v. Board of Educ. of Okla. City Pub. Schools, 338 F.Supp. 1256 (W.D.Okla.1972). In 1986, when last before us, plaintiffs urged review of the district court's refusal to reopen the case to consider their petition for enforcement of the court's prior injunctive decree. The motion to reopen was triggered by the implementation of a new student assignment plan in 1984.

Until that time, defendants, the Board of Education of the Oklahoma City Public Schools, school officials, and individual board members, (the Board or defendants) operated the Oklahoma City School District (the District) under the Finger Plan, a court ordered desegregation plan prepared by Dr. John A. Finger, Jr., a Professor of Education at Rhode Island College and authority on issues of school desegregation.1 Under the Finger Plan, attendance zones were redesigned so that high schools and middle schools enrolled black and white students. Black elementary students in grades 1 through 4 were bused to previously all white elementary schools while majority black elementary schools were converted into 5th-year centers with enhanced curricula. Black fifth graders then attended the 5th-year center in their neighborhood, while white fifth graders were bused for the first time into black neighborhoods to attend class. Excepted from the Finger Plan were certain schools enrolling grades K-5, which were designated "stand alone." These schools were located in neighborhoods that were racially balanced. Kindergarten children attended their neighborhood elementary school unless their parents chose to send them to another school to join a sibling or be closer to the parent's workplace. Aside from minor alterations necessitated, for example, by a school's closing, the Board maintained the District under the Finger Plan's basic techniques of pairing, clustering, and compulsory busing, even after the district court declared the District unitary and terminated the case. Dowell v. School Bd. of Okla. City Pub. Schools, No. CIV-9452, slip op. (W.D.Okla. Jan. 18, 1977).

Seven years later, the Board adopted a new student assignment plan, the Student Reassignment Plan, (the Plan), which was implemented for the 1984-85 school year. The Plan eliminated compulsory busing in grades 1 through 4 and reassigned elementary students to their neighborhood schools. A "majority to minority" transfer option (M & M) was retained to permit elementary students assigned to a school in which they were in the majority race to transfer to one in which the student would be in the minority. Fifth-year centers would remain throughout the District and, like the middle schools and high schools, would continue to maintain racial balance through busing. The Plan created the position of an "equity officer" assisted by an equity committee to monitor all schools to insure the equality of facilities, equipment, supplies, books, and instructors. Dowell v. Board of Educ. of Okla. City Pub. Schools, 606 F.Supp. 1548, 1552 (W.D.Okla.1985). The Plan professed to maintain integrated teaching staffs in line with the District's affirmative action goal. As a consequence of the Plan, eleven of the District's sixty-four elementary schools enrolled 90%+ black children. Twenty-one elementary schools2 became 90%+ white and non-black minorities.3 Thirty-two elementary schools remained racially mixed.

In February 1985, plaintiffs filed a motion to intervene and reopen the case claiming the Board unilaterally abandoned the Finger Plan. Although the record indicated the subsequent hearing was limited to "the question of whether this case shall be reopened and the applicants allowed to intervene shall be tried and disposed of," Dowell, 795 F.2d at 1523 (emphasis omitted), the district court received evidence on the constitutionality of the Plan and disposed of all of the substantive issues defendants raised. The district court concluded the Plan was constitutional and found no special circumstances justifying relief under Fed.R.Civ.P. 60(b) to support reopening. Dowell, 606 F.Supp. at 1557.

We reversed, holding the court abused its discretion in failing to reopen the case and prematurely reached the merits of the Plan's constitutionality without permitting plaintiffs the opportunity to support their petition for enforcement of the mandatory injunction which the court had never dissolved or modified. Dowell, 795 F.2d at 1523. Key to our disposition was the reassertion of the parties' burden of proof under Fed.R.Civ.P. 60(b).4 We stated that on remand, the plaintiffs, beneficiaries of the original injunction, only have the burden of showing the court's mandatory order has been violated. "The defendants, who essentially claim that the injunction should be amended to accommodate neighborhood elementary schools, must present evidence that changed conditions require modification or that the facts or law no longer require the enforcement of the [1972] order." Id. (citation omitted) (emphasis added). Nothing in this disposition touched on the underlying constitutional issues. "[O]ur holding should not be construed as addressing, even implicitly, the ultimate issue of the constitutionality of the defendants' new school attendance plan." Id. at 1523. Remand was confined to a determination of "whether the original mandatory order will be enforced or whether and to what extent it should be modified." Id.5

During the eight-day hearing on these remand instructions, defendants6 introduced a golconda of testimony and exhibits to establish their position that substantial demographic changes in the District rendered the Finger Plan inequitable and oppressive. The inequity, the Board maintained, surfaced primarily in the burgeoning number of schools that qualified for stand-alone status, thus necessitating that black children be transported greater distances to attend racially balanced elementary schools. Defendants' expert, Dr. William A. Clark, a specialist in population geography, testified on the migration and mobility of the black population in the District. Dr. Clark was satisfied that the residential pattern that developed in the District since the implementation of the Finger Plan was not a vestige of what had occurred thirty-five or forty years before and that "black preference" accounted for the dispersal of the black population throughout the District. While recognizing that socioeconomic factors must be considered in any housing decision, Dr. Clark maintained that the most significant motivation was preference.

Dr. Finis Welch, an economist at the University of California, offered testimony on studies he conducted of the dissimilarity and exposure indices of residential areas on which the Plan was based. Dr. Welch opined that the increasing number of stand-alones would "draw down" the 5th-year centers which, he projected, would result in closing more schools in the northeast quadrant, the area of central Oklahoma City which remains majority black.

Three Board members testified about their involvement in the preparation of the Plan. The District's superintendent, several black school administrators, and various members of the community offered their views on an array of issues, from linking neighborhood schools to black achievement, to the value of parental involvement in a child's education. Ms. Susan Hermes, a member of the committee which prepared the Plan, stated that she believed "educationally it is better for a child to have family nearby." (R. IV, 390). Over plaintiffs' objection, counsel for the Board asked each witness if he or she believed the District remained unitary after implementation of the Plan. The court, also over plaintiffs' objection, asked key defense witnesses if the Plan was adopted with discriminatory intent.

Through cross-examination and in its presentation of evidence, plaintiffs offered a contrasting analysis of the issues of demographic change, the impact of the Plan, and the Board's alternative approaches of the Effective Schools Program, increased parental participation in PTA, and equity supervision. Dr. John Finger, who had prepared the original plan, rejected each of these features of the new Plan noting that Effective Schools and increased parental participation deal with different problems and cannot be substituted for a desegregated education. Dr. Gordon Foster, a professor of education at the University of Miami, testified about a student assignment plan he had prepared for plaintiffs to solve the perceived inequities of busing under the Finger Plan.

In its subsequent order, the district court initially observed it was "now aware that it should have dissolved the injunction in 1977, as pointed out in the Circuit opinion, because the Oklahoma City schools were at that time, as they are today, operating as a unitary system, wholly without discrimination to blacks or other minority students, faculty or staff." Dowell, 677 F.Supp. at 1506. Nevertheless, the court apprehended the command we framed in our prior review. "The fundamental issue the court must address is whether the School Board has shown a substantial change in conditions warranting dissolution or modification of the 1972 Order." Id. Relying on the testimony of Drs. Clark and Welch, the court concluded:

[T]he Oklahoma City Board of Education has taken absolutely no action which has caused or contributed to the patterns of residential segregation which presently exist in areas of Oklahoma City. If anything, the actions of the Board of Education, through implementation of the Finger Plan at all grade levels for more than a decade, have fostered the neighborhood integration which has occurred in Oklahoma City.

Id. at 1512.

Thus, unlinking the Board from existing residential segregation and satisfied that demographic changes rendered the Finger Plan inequitable,7 the court proceeded to examine the constitutionality of the Plan. Acknowledging that "[a] once unitary school district may lose its unitary status by partaking in intentionally discriminatory acts creating de jure segregation," id. at 1515, the court set forth the guidelines established in Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971), and Keyes v. School Dist. No. 1, Denver, Colo., 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973).

The duty of both the District Court and the Court of Appeals in a case such as this, where mandatory segregation by law of the races in the schools has long since ceased, is to first determine whether there was any action in the conduct of the business of the School Board which [was] intended to, and did in fact discriminate against minority pupils, teachers, or staff.

Dayton Bd. of Educ. v. Brinkman, 433 U.S. 406, 420, 97 S.Ct. 2766, 2775, 53 L.Ed.2d 851 (1977).

The court reviewed the evidence and concluded that not only did legitimate nondiscriminatory factors motivate the adoption of the Plan, but, also, that the Plan currently maintained a unitary district which enjoyed increased parental and community involvement and included safeguards such as the equity officer and Effective Schools Program to insure continued unitariness. While the court entertained plaintiffs' contention that the Plan did have a disproportionate impact upon some blacks in the District, it concluded that racial imbalance in the schools, without more, does not violate the Constitution, citing Milliken v. Bradley, 433 U.S. 267, 97 S.Ct. 2749, 53 L.Ed.2d 745 (1977). "It follows that a school board serving a unitary school system is free to adopt a neighborhood school plan so long as it does not act with discriminatory intent." Dowell, 677 F.Supp. at 1518. The court rejected plaintiffs' claim that the Plan is a step toward a dual school system as "ludicrous and absurd." Id. at 1524. In light of these findings of fact and conclusions of law, the district court determined the Foster Plan, plaintiffs' proposed modification of the 1972 decree, was neither feasible nor necessary.

Plaintiffs appeal this order, contending essentially that the district court misapplied the instructions on remand and misperceived the function of the unitary status achieved in 1977 to be a post-decree change in circumstances warranting dissolution of the injunction. In dissolving the injunction, plaintiffs urge the court abused its discretion by relying on clearly erroneous findings of fact.

At the outset, we must underscore this case involves an injunction upon which relief was sought pursuant to Fed.R.Civ.P. 60(b). Dowell, 795 F.2d at 1522.8 Thus, our review focuses on whether the district court abused its discretion in granting the Board's motion to dissolve the injunction and denying plaintiffs' motion to modify the relief. On appeal we will not disturb the district court's determination except for an abuse of discretion. Securities and Exch. Comm'n v. Blinder, Robinson & Co., Inc., 855 F.2d 677 (10th Cir.1988). The district court's exercise of discretion, however, must be tethered to legal principles and substantial facts in the record. Evans v. Buchanan, 582 F.2d 750, 760 (3d Cir.1978), cert. denied, 446 U.S. 923, 100 S.Ct. 1862, 64 L.Ed.2d 278 (1980). "[D]iscretion imports not the court's inclination, but ... its judgment; and its judgment is to be guided by sound legal principles." Franks v. Bowman Trans. Co., 424 U.S. 747, 770-71, 96 S.Ct. 1251, 1266-67, 47 L.Ed.2d 444 (1976) (citation omitted).

While a court's equitable power to fashion a remedy is broad and its continuing duty to modify or vacate relief inheres to the prospective nature of the relief,9 modification is subject to an exacting standard from which this circuit has not wavered. See Blinder, Robinson, 855 F.2d at 679; Equal Employment Opportunity Comm'n v. Safeway Stores, Inc., 611 F.2d 795 (10th Cir.1979), cert. denied, 446 U.S. 952, 100 S.Ct. 2918, 64 L.Ed.2d 809 (1980); Securities and Exch. Comm'n v. Thermodynamics, Inc., 464 F.2d 457 (10th Cir.1972), cert. denied, 410 U.S. 927, 93 S.Ct. 1358, 35 L.Ed.2d 588 (1973).

This standard, first articulated in United States v. Swift & Co., 286 U.S. 106, 119, 52 S.Ct. 460, 464, 76 L.Ed. 999 (1932), requires "[n]othing less than a clear showing of grievous wrong evoked by new and unforeseen conditions ... to change what was decreed after years of litigation with the consent of all concerned."10 The Court cautioned:

There is need to keep in mind steadily the limits of inquiry proper to the case before us. We are not framing a decree. We are asking ourselves whether anything has happened that will justify us now in changing a decree. The injunction, whether right or wrong, is not subject to impeachment in its application to the conditions that existed at its making. We are not at liberty to reverse under the guise of readjusting. Life is never static and the passing of a decade has brought changes to the grocery business as it has to every other. The inquiry for us is whether the changes are so important that dangers, once substantial, have become attenuated to a shadow.

Swift, 286 U.S. at 119, 52 S.Ct. at 464 (emphasis added).

Hence, to pass muster under this test, the party seeking relief from an injunctive decree "must demonstrate dramatic changes in conditions unforeseen at the time of the decree that both render the protections of the decree unnecessary to effectuate the rights of the beneficiary and impose extreme and unexpectedly oppressive hardships on the obligor." T. Jost, From Swift to Stotts and Beyond: Modification of Injunctions in the Federal Courts, 64 Tex.L.Rev. 1101, 1110 (1986). While the Swift language may also support a modification when the original purposes of the injunction are not fulfilled,11 the standard still constricts the district court's inquiry.

Placed in other words, this means for us that modification is only cautiously to be granted; that some change is not enough; that the dangers which the decree was meant to foreclose must almost have disappeared; that hardship and oppression, extreme and unexpected, are significant; and that the movants' task is to provide close to an unanswerable case. To repeat: caution, substantial change, unforeseenness, oppressive hardship, and a clear showing are the requirements.

Humble Oil & Ref. Co. v. American Oil Co., 405 F.2d 803, 813 (8th Cir.), cert. denied, 395 U.S. 905, 89 S.Ct. 1745, 23 L.Ed.2d 218 (1969). Fed.R.Civ.P. 60(b) codifies this standard.

When the relief has been fashioned and the decree entered, "an injunction takes on a life of its own and becomes an edict quite independent of the law it is meant to effectuate." 64 Tex.L.Rev. 1101, 1105. For this reason, the court's jurisdiction extends beyond the termination of the wrongdoing, Battle v. Anderson, 708 F.2d 1523, 1538 (10th Cir.1983), because an injunction seeks to stabilize a factual setting with a judicial ordering and maintain that condition which the order sought to create. The condition that eventuates as a function of the injunction cannot alone become the basis for altering the decree absent the Swift showing. Securities and Exch. Comm'n v. Jan-Dal Oil & Gas, Inc., 433 F.2d 304 (10th Cir.1970). To do otherwise is to return the beneficiary of injunctive relief to the proverbial first square. It is for this reason that Swift remains viable.12

Thus, compliance alone cannot become the basis for modifying or dissolving an injunction. United States v. W.T. Grant Co., 345 U.S. 629, 633, 73 S.Ct. 894, 897, 97 L.Ed. 1303 (1953); Jan-Dal Oil & Gas, Inc., 433 F.2d at 304.13 Nor can a mere change of conditions alter the prospective ordering of relationships embodied by a permanent injunction. The party subject to the decree must establish by clear and convincing evidence that conditions which led to the original decree no longer exist, or the condition the order sought to alleviate, a constitutional violation, has been eradicated.14 Until this showing is made, the decree stands.

Nevertheless, a permanent injunction empowered by a court's continuing jurisdiction does not presume that its underlying circumstances or the rights achieved remain static. "By its forward cast, an injunction contemplates change and must be sufficiently malleable to adapt the ordered relief to contemporary circumstances." United States v. Lawrence County School Dist., 799 F.2d 1031, 1056 (5th Cir.1986). Thus, while principles of res judicata may be applied to the factual finding of unitariness at the time the finding is made with the injunction in place, we have recognized that this past finding alone does not bar reconsideration of the decree. Dowell, 795 F.2d at 1519.

In contending there should be a different standard employed in school desegregation cases, the dissent miscasts our basic premise. We do not imply perpetual supervision of public schools by federal courts, nor do we suggest the Board is incapable of complying with constitutional mandates. We take the simple position that an injunctive order entered in a school desegregation case has the same attributes as any other injunctive order issued by a federal court, and that it is binding upon all parties until modified by the court which entered the order in the first instance. We add, as the trial court initially recognized, the injunctive order can be modified or dissolved only upon a finding of changed conditions. In this context, the intent of the defendants has little, if any, relevance.

In 1972, having found "the Defendant School Board has totally defaulted in its acknowledged duty to come forward with an acceptable plan of its own," Dowell, 338 F.Supp. at 1271, the district court held that "[p]laintiffs are entitled to a decree requiring the reasonably immediate conversion of the Oklahoma City Public Schools into a unitary school system." Id. at 1272 (citations omitted). The Board was ordered not to alter or deviate from the plan without "the prior approval and permission of the court," and the order was made binding on the Board, "its members, agents, servants, employees, present and future, and upon those persons in active concert or participation with them." Id. at 1273.

The decree embodied the constitutional mandate to eliminate "root and branch" racial discrimination enforced through a dual school system. Green v. County School Bd. of New Kent County, Va., 391 U.S. 430, 437, 88 S.Ct. 1689, 1693, 20 L.Ed.2d 716 (1968). The resulting terrain circumscribed by the injunction was later declared unitary upon the district court's finding certain components of unitariness to have been satisfied.15 "[U]nitariness is less a quantifiable 'moment' in the history of a remedial plan than it is the general state of successful desegregation." Morgan v. Nucci, 831 F.2d 313, 321 (1st Cir.1987).16 While a declaration of unitariness addresses the goals of injunctive relief, it alone does not sweep the slate clean.

Nor, in our view, does a finding of unitariness mandate the later dissolution of the decree without proof of a substantial change in the circumstances which led to issuance of that decree. Dowell, 795 F.2d at 1521; contra United States v. Overton, 834 F.2d 1171 (5th Cir.1987); Riddick v. School Bd. of Norfolk, 784 F.2d 521 (4th Cir.), cert. denied, 479 U.S. 938, 107 S.Ct. 420, 93 L.Ed.2d 370 (1986).17 Until that showing, "those who are subject to the commands of an injunctive order must obey those commands, notwithstanding eminently reasonable and proper objections to the order, until it is modified or reversed." Pasadena City Bd. of Educ. v. Spangler, 427 U.S. 424, 439, 96 S.Ct. 2697, 2706, 49 L.Ed.2d 599 (1976). It is imperative that the rights of the party for whose benefit an injunction has been entered are affected by no one unless a court determines the injunction in current form is no longer necessary to achieve the court's original objective. It is also imperative that when considering whether to vacate or modify an injunctive decree, the district court not retry "the original premises of the judgment; instead, any modification must be confined and tailored to the change in circumstance that justifies the modification." Lawrence County School Dist., 799 F.2d at 1056. Necessarily, however, in conducting a factual inquiry into the changed conditions pled, the court must reexamine whether the underlying substantive obligations are preserved. See B. Landsberg, The Desegregated School System and the Retrogression Plan, 48 La.L.Rev. 789 (1988).18

Nevertheless, in this case, unilaterally and without prior approval from the district court, as required by the injunctive decree, the Board implemented the Plan. It is uncontested that the contents of the Plan are contrary to the explicit dictates of the injunction. As we previously noted, the Board's action creates a "special circumstance which permitted plaintiffs to return to court and test the presumptions premised in the declaration of unitariness." Dowell, 795 F.2d at 1522. We so instructed the district court.

The first presumption we address, then, is whether the Board's Plan maintains the unitary status of the District since the injunction remained in effect when the Board restored neighborhood schools for elementary student assignments.19 This presumption flows from the Board's continuing affirmative duty to "accomplish desegregation," Swann, 402 U.S. at 32, 91 S.Ct. at 1284, to attain "maximum practicable desegregation," Morgan v. McDonough, 689 F.2d 265, 280 (1st Cir.1982), and to protect the constitutional rights of the class protected by the equitable remedy. Keyes v. School Dist. No. 1, Denver, Colo., 609 F.Supp. 1491, 1515 (D.Colo.1985). The remedy "must survive beyond the procedural life of the litigation." Dowell, 795 F.2d at 1521.

That thirty-two of the sixty-four elementary schools in Oklahoma City emerge from the Plan as one-race majority schools not only establishes a prima facie case that the decree has been violated and the presumption of unitariness challenged, but also satisfies plaintiffs' burden in reopening and shifts the burden to defendants to produce evidence of changed circumstances or oppressive hardship.

[I]n a system with a history of segregation the need for remedial criteria of sufficient specificity to assure a school authority's compliance with its constitutional duty warrants a presumption against schools that are substantially disproportionate in their racial composition ... The court should scrutinize such schools, and the burden upon the school authorities will be to satisfy the court that their racial composition is not the result of present or past discriminatory action on their part.

Swann, 402 U.S. at 26, 91 S.Ct. at 1281. The Board bears a "heavy burden" to show that its implementation of the Plan does not "serve to perpetuate or re-establish the dual school system." Dayton Bd. of Educ. v. Brinkman, 443 U.S. 526, 538, 99 S.Ct. 2971, 2979, 61 L.Ed.2d 720 (1979) (quoted in Clark v. Board of Educ. of Little Rock School Dist., 705 F.2d 265, 271 (8th Cir.1983)). This burden is not alleviated after a finding of unitariness when the decree remains in place but is focused on the Swift inquiry whether "anything has happened that will justify us now in changing a decree." 286 U.S. at 119, 52 S.Ct. at 464.

The Board sought to prove that substantial demographic change in the District established new conditions that were unforeseen at the time the decree was instituted and which now produced "hardship so 'extreme and unexpected' as to make the decree oppressive." Equal Employment Opportunity Comm'n v. Safeway, 611 F.2d at 800 (quoting Swift, 286 U.S. at 119, 52 S.Ct. at 464). While the record sets forth changed circumstances not unlike those contemplated by Swann, it fails to establish "the dangers the decree was meant to foreclose must almost have disappeared." Humble Oil & Ref. Co., 405 F.2d at 813.

In its factual findings, the district court relied on relocation statistics offered by defendants' experts, Drs. Clark and Welch. Although Dr. Clark's evidence indicated black families had relocated within and outside of the District,20 he conceded that his examination focused only on seven inner-city tracts and not on additional predominantly black residential tracts to the north of the studied area. (R. II, 93-94). While Dr. Clark's study establishes there is a substantial decrease in black population in these particular tracts, it reveals the same decrease for total population.21 Both on direct and cross-examination, Dr. Clark stated that the area encompassing the seven tracts underwent "land use transition," (R. II, 68) and that construction of interstate highways, I-40 and I-35, and developments in institutions, notably the hospital complex, had "dramatic impact" on population movement in some of the studied tracts. (R. II, 68).

Based on this testimony, the court concluded there was "a substantial amount of turnover in the black population residing in the east inner-city tracts." Dowell, 677 F.Supp. at 1507. This conclusion is also premised on metropolitan census data compiled from completion of a long census form. While the long census form asks the respondent whether he or she lived in the same house five years before, it does not determine whether the respondent moved out of the District or merely down the street. The basis for the turnover22 rate is thus incomplete, rendering the census form suspect.23

Using census figures, Dr. Clark calculated that the percentage of black population residing in these tracts in Oklahoma City between 1960 and 1980 decreased by 67.2%.24 Despite his statement about external forces affecting population movement, Dr. Clark concluded that "private preference" was the chief motivating factor in determining where people chose to live.25 (R. II, 113). Dr. Clark observed the "strong disinclination" of whites to move into predominantly black neighborhoods and their coincidental inclination to move out of neighborhoods that become 25 to 30% minority. (R. II, 105). He conceded that majority black areas would then be unlikely to change unless the black population moved elsewhere. (R. II, 106). The district court thus observed, "Some blacks were choosing to live within the area and others were choosing to move away. (Tr. 71)." 677 F.Supp. at 1507.

The district court also relied on the testimony of Dr. Welch who presented statistical analyses of the racial composition of residential attendance zones in the District from 1972 to 1986 and then used these figures to project racial composition in 1995. Based on Dr. Welch's calculations, the court noted that "the exposure of blacks to non-blacks almost doubled." 677 F.Supp. at 1508.26 Embracing Welch's analysis which included a ranking of the 125 school districts he had studied, the court declared, "the Oklahoma City school district experienced the eighth largest reduction in the index of dissimilarity or, in other words, the eighth greatest improvement in integration, during the period from 1968 to 1982 (Def.Ex. 27; Tr. 130-31)." 677 F.Supp. at 1508.27 The court noted that "[e]ven after implementing the K-4 neighborhood school plan, the degree of overall dissimilarity among the races attending school in Oklahoma City was less than that of Tulsa, Oklahoma, whose index was .557. (Def.Ex. 38)." 677 F.Supp. at 1508.28 Similarly, the district court relied on a comparison of the District figures following implementation of the Plan with figures from other "unitary districts in the country."29 Noting the change in the dissimilarity index from .78 prior to implementation of the Finger Plan to .24 in 1984, the district court stated "the index rose slightly to .38" in 1985 with the reintroduction of neighborhood schools. 677 F.Supp. at 1509. Contrary to the district court's characterization, the rise in the index represents a 58% increase in the ratio of blacks to non-blacks. Despite this expansion of dissimilarity, the court concluded the "increased residential integration in Oklahoma City has resulted in a much lower level of dissimilarity today in the neighborhood elementary schools (.56) than existed in 1971 before the Finger Plan was implemented (.83). (Def.Ex. 44; Tr. 187)." 677 F.Supp. at 1509.

On the basis of Def.Ex. 20, two graphs plotting enrollment figures for the District for white, black, and non-black minority students, grades K-12, the court concluded "the student body is truly multi-cultural."30 677 F.Supp. at 1509. Nevertheless, the court acknowledged the Plan created "some schools," eleven, which are 90%+ black but observed "the plan created no schools which are 90% or more white." Id. at 1510. The Plan, however, created twenty-one schools that had less than 10% black student enrollment.

The district court does not address contrasting evidence in the record. Unmentioned by the court is plaintiffs' cross-examination of Dr. Welch which produced testimony directly controverting that of Dr. Clark31 and undermined the method employed to create the figures the Board relied on to represent substantial demographic change and the oppressiveness of the decree. Noting that he used two different methods for calculating the 1974 to 1986 figures and the 1986 to 1995 figures, Dr. Welch conceded: "And I really didn't want an inconsistent forecast. I thought someone would be cross-examining me. And so I designed the procedure to be completely internally consistent." (R. III, 244). His numbers, he stated, were "guesstimates." (R. III, 246).

In addition, the court does not reference the testimony of Dr. Yale Rabin, plaintiffs' expert in population distribution. Using U.S. Census data, Dr. Rabin compared and analysed the black population in the District between 1950 and 1980. According to these census tract figures, the black population expanded from one tract in which approximately 25% of the District black population resided, to sixteen tracts 75%+ black, including 60.8% of the District's black population. He explained that as the area expanded spatially from one tract in 1950 to six tracts in 1960, thirteen tracts in 1970, and sixteen tracts in 1980, each expansion included the original all-black tracts. (R. VII, 1125-31). Dr. Rabin controverted Dr. Clark's conclusion that the black population had dropped to 16.9% in 1980 in the six tracts. "[T]he area of concentration itself has changed, and it's misleading to refer, in each successive decade, to the same six tracts as the area of concentration." (R. VII, 1133). Dr. Rabin not only recognized the substantial population displacement caused by institutional and highway development but focused the effect of Def.Exs. 7 and 8, maps showing the numbers of black families and general direction of movement in and out of the District. For example, Dr. Rabin noted that while 46 families moved into white areas from the northeast quadrant from 1974 to 1978 (Def.Ex. 7), many thousands of blacks live in the subject tracts, thus putting the significance of the turnover numbers into perspective. (R. VII, 1157). In fact, the more predominant population shift, 148 families, was within the northeast quadrant.

Most importantly, Dr. John Finger, plaintiffs' expert, underscored that the Board's statisticians had "changed the rules." (R. VIII, 1207). He explained,

There will be no schools that have less than ten percent minority, but there will be schools that have less than ten percent black. How you label these as segregated or not is what the words mean, and segregated has always been a difficult word.

(R. VIII, 1208).

Permeating the testimony on demographic change were sharply contrasting views on the impact of busing on children of "tender age." 677 F.Supp. at 1526. Numerous lay witnesses and District personnel testifying on behalf of the Board generally stated that busing young children had an adverse, emotional impact on the child.32 Defendants' expert witness, Dr. Herbert Walberg, a research professor at the University of Illinois, offered a study he completed showing that black children who were transported to school tested lower than black children who did not ride a school bus. Plaintiffs' witness, Dr. Robert Crain, who was qualified an expert on school desegregation, stated that Walberg's study was "absolutely indefensible" because it omitted critical covariant factors like socioeconomic status in the analysis. (R. VII, 1008). Dr. Crain stated that in light of the fact that half of all public school students ride a school bus and that only 5% of those children are bused for desegregation purposes, the evidence of the harmful effects of transportation on student achievement and emotional development is suspect. The district court did not reference plaintiffs' evidence on this issue.

Based on the divergent testimony on demographic change, the court concluded the Board had not taken action to cause or contribute to presently existing residential segregation but "[i]f anything, the action of the Board of Education, through implementation of the Finger Plan at all grade levels for more than a decade, have [sic] fostered the neighborhood integration which has occurred in Oklahoma City." 677 F.Supp. at 1512. Previously, in summarizing the relocation statistics, the court observed, "These relocation studies reveal the compulsory busing of black children to a certain area does not have any appreciable affect [sic] on where their parents choose to relocate. (Tr. 76-77)." 677 F.Supp. at 1508.

That demographic change of some degree occurred within the District after the Finger Plan was instituted is apparent. As Swann observed, "It does not follow that the communities served by such systems will remain demographically stable, for in a growing, mobile society, few will do so." 402 U.S. at 31, 91 S.Ct. at 1283. Nevertheless, we are reluctant to hitch the preservation of hard-won constitutional rights to numbers alone. "Unitary status is not simply a mathematical construction." Morgan, 831 F.2d at 321. As the district court observed in Keyes v. School Dist. No. 1, Denver, Colo., 609 F.Supp. at 1516, "The expert testimony in this case concerning the use of racial balance and racial contact indices, and the differing conclusions reached by the experts called by the respective parties, demonstrate once again the facility with which numerical data may be manipulated and discriminatory policies may be masked." In Oklahoma City, the sum total of all of the numbers immutably underscores the emergence of eleven all-black elementary schools and twenty-one 90%+ white and non-black minority schools, roughly half of the District's elementary schools, with the reinstitution of neighborhood black schools for the elementary grades. In fact, when the actual numbers of children attending District elementary school are run, the result is even more dramatic. Of the approximately 6,464 black students33 attending the District's elementary schools K-4, 2,990, or 46.2% of all black elementary children in the District attend the eleven 90%+ black elementary schools.34

Similarly, we are unable to conclude that these same numerical calculations support a finding that the Finger Plan became a hardship "extreme and unexpected," Humble Oil & Ref. Co., 405 F.2d at 813, because of the unintended impact of the stand-alone schools. This hardship was projected to arise if a school became stand-alone, necessitating busing black students, who had been bused into that school, even greater distances to attend an integrated school. With more students attending naturally integrated K-5 schools, the 5th-year centers in the black community would then have to close.

As viewed by the district court, the creation of Bodine Elementary School in southeast Oklahoma City as a K-5 stand-alone caused the Board to focus on the "perceived inequities" of the stand-alone feature. 677 F.Supp. at 1513. According to the minutes of the Board meeting which addressed the question of making Bodine a K-5 rather than a K-4 stand-alone,35 Board members voiced several concerns over the process of deciding which schools qualified and became stand-alone. (Def.Ex. 76). Dr. Clyde Muse, a black Board member, objected that the creation of Bodine stifled growth in the northeast quadrant and was yet another example of "a concerted effort to see to it that not only will the black community or the northeast quadrant not integrate, there also seems to be a concerted effort on somebody's part to see that it always remains impoverished." (Def.Ex. 76 at 349). Dr. Muse lamented the inevitable closing of schools in the northeast quadrant and urged the District undertake a study to determine what changes had occurred that could result in a more equitable plan for the District rather than the apparent piecemeal approach. Id. Another Board member, Ms. Jean Brody, urged the District to undertake a comprehensive study to avoid what she perceived as random planning that resulted in Bodine's becoming a K-5 stand-alone, but postponed Rockwood Elementary School's becoming stand-alone although it fully qualified and had the capacity to become a K-5 school.36

In voting to make Bodine a K-5 stand-alone, the Board rejected the advice of Dr. Paul Heath, a board member, that the K-5 concept was educationally unsound and would ultimately adversely impact the entire District. Of concern to participants at the meeting was the fact that in going to K-5 status, Bodine fifth graders would give up the opportunity to participate in special programs like strings and visual arts offered at the 5th-grade centers. (Def.Ex. 76). On the positive side, however, student reassignments necessitated by making Bodine stand-alone were not expected to impact the existing 5th-grade centers. (Def.Ex. 76).

Similarly, the trial testimony on the hardship of stand-alone schools echoed some of those concerns and underlined that the Board's planning was based on theoretical conjecture, speculative forecasting, and discretionary decision making. At the outset, Dr. Welch noted that of the eleven stand-alone schools open in 1972, only three retained this status in the District in 1984.37 (R. III, 289). The projected number of stand-alones was tied to Dr. Welch's 1995 District calculations. A senior researcher for the District, who monitored student assignments and helped prepare projections on stand-alones, stated that although ten schools were eligible for stand-alone status, only three were then stand-alone. He stated that in order to create a stand-alone, the eligible school had to have the capacity to absorb the increased number of students. (R. IV, 495-96); see also Def.Ex. 69. Internal Board memoranda also addressed the possibility of creating additional stand-alones by altering attendance boundaries, exploring reassignment options,38 and opting for either K-4 or K-5 stand-alones. (R. IV, 498). Dr. Finger stated that the original plan anticipated making as many schools stand-alone as qualified even if some busing distances increased. "But, ... these things get to be political." (R. VIII, 1201). The 5th-grade centers, he stated, were considered temporary and were designed to be incorporated into the middle schools.

The stand-alone feature, thus, emerged from the evidence as a matter of speculation tied to capacity problems, budget constraints, and local politics. Nevertheless, it was the cornerstone upon which advocates of the need to modify or dissolve the Finger Plan built their claim of hardship.

We are satisfied the evidence reveals that because of population shifts in the District, it was necessary to modify the Finger Plan. It is within the court's equitable power to modify the Finger Plan to mirror these changed circumstances, to retain the unitariness of the District, and reflect the Board's continuing duty under the decree. Just as the court can tailor the relief to modification, so too can it dissolve the injunction upon finding "that what it has been doing has been turned through changing circumstances into an instrument of wrong." Swift, 286 U.S. at 114-15, 52 S.Ct. at 462. Unfortunately, the district court perceived this duty entirely in terms of the Board's alleged discriminatory intent in adopting the Plan. This perception overlooks the essential point. Given the changes that emerge from all of the evidence presented, the court must determine whether the Plan ameliorates those conditions. Dissolution is appropriate only if the evidence unmistakably reveals the Plan encompasses the changed circumstances and maintains the continuing prospective effect of the decree.

Again, to undertake this analysis, the court must direct its attention to "the question of the withdrawal or modification of injunctive relief granted in the past ... where the Cardozo [Swift ] precepts are the operating guidelines." Humble Oil & Ref. Co., 405 F.2d at 814. Thus, while the Board's motive may be one circumstance in evaluating the effect of the Plan, it is only an element affecting the ultimate decision. An unimpeachable motive cannot obscure the essential question, does the Plan relieve the effects of changed circumstances and potential hardship? Only a positive response will merit dissolution.

The issue then becomes whether the Board's action in response to the changed conditions has the effect of making the District "un -unitary" by reviving the effects of past discrimination. The new Plan must be judged in light of the old plan to assure it mirrors actual changes in the District without so radically departing from the original decree that the rights secured by that decree are vitiated.

Swann guides our review in this inquiry by focusing our attention on the Board's continuing duty to remedy the effects of past discrimination until "it is clear that state-imposed segregation has been completely removed." 402 U.S. at 13, 91 S.Ct. at 1274; see also Columbus Bd. of Ed. v. Penick, 443 U.S. 449, 99 S.Ct. 2941, 61 L.Ed.2d 666 (1979).39 The inquiry into whether the Plan maintains unitariness in student assignments may concretely be directed to evaluating (1) the number of racially identifiable schools; (2) the good faith of school officials in the desegregation effort and running the schools; and (3) "whether maximum practicable desegregation of student bodies at the various schools has been attained." Morgan, 831 F.2d at 319. See also Ross v. Houston Indep. School Dist., 699 F.2d 218, 227 (5th Cir.1983) ("[T]he decision that public officials have satisfied their responsibility to eradicate segregation and its vestiges must be based on conditions in the district, the accomplishments to date, and the feasibility of further measures.").40 No one factor is dispositive of the determination that unitariness is preserved. However, once dismantled, the dual school system should remain dismantled.

Thus, we are troubled because the evidence indicates the Board's implementation of a "racially neutral" neighborhood student assignment plan has the effect of reviving those conditions that necessitated a remedy in the first instance. Under these circumstances the expedient of finding unitariness does not erase the record or represent that substantial change in the law or facts to warrant overlooking the effect of the Board's actions.41C.

The district court was satisfied the Plan was adopted to remedy the increased busing burdens on young black pupils, avoid closing 5th-year centers in the northeast quadrant, and eliminate the inequities of stand-alone schools. Despite the emergence of one-race elementary schools, the court found the Plan did not disturb the District's unitariness. The district court concluded that unless the Plan was adopted with discriminatory intent, a neighborhood school plan that has the effect of creating one-race schools is not constitutionally infirm.

To reach this conclusion, the court examined the remaining components of the Plan. While school faculties were not in perfect racial balance, particularly in the 90%+ black elementary schools, the court found that negotiated agreements with the teachers' union and teacher preference and seniority accounted for the imbalance and not Board policy.

The court did not address plaintiffs' exhibits 48, 50, 52, and 54. The exhibits compare elementary school enrollment with the racial composition of faculty from 1972 to 1986-87 and reflect the growing parity of imbalance between faculty and students. By 1986-87, the 90%+ black elementary schools are staffed by predominantly black teachers.42 Although the executive director of personnel testified that especially after 1985,43 teacher assignments would comply with the District's affirmative action goal of 36.9% with a 10% variance factor, the numbers belie the aspiration.

Nevertheless, the court was satisfied that recent Board action would "bring[ ] elementary faculties into racial balance in 1987-88," 677 F.Supp. at 1519, based on the statement of the District's affirmative action program planner. However, the record fails to support this conclusion with any specific evidence of change to overcome plaintiffs' documented countertrend.

The district court believed that other factors in the equation maintained the District's unitariness and offset the racial imbalance in the elementary schools. Of prime importance was the majority-to-minority transfer option which represented to the court that "parents in Oklahoma City today have a choice. No pupil of a racial minority is excluded from any school in Oklahoma City on account of race." 677 F.Supp. at 1523. The record does not support this assertion. In fact, there is little evidence to determine the effectiveness and utilization of the transfer option. Dr. Belinda Biscoe, an administrator in the department of support programs, testified that letters were sent after the Plan was implemented informing parents of the M & M option, but no follow-up was done. Dr. Biscoe expressed the concern, apparently voiced by the District superintendent, that parents needed more information about the option. (R. III, 327). Asked if the Board had studied the program to determine who was exercising the transfer option, Dr. Biscoe answered that she did not believe the numbers had been analysed. (R. III, 327). Dr. Betty Mason, the assistant superintendent of high schools, agreed that the M & M policy could not serve to desegregate the schools in the northeast quadrant (R. V, 609) and was limited by the capacity of the receiving school. Although Dr. Finger acknowledged the M & M option might work if parents understood the alternative and were willing to exercise it, he observed that often those children who most need desegregated schools would be "the least likely to take that option." (R. VIII, 1196). Another defense witness believed the transfer option was available for "convenience." (R. VI, 837). There is simply no other evidence in the record to support the court's conclusion that parents understand the availability of the option and freely exercise it. Indeed, the court's analysis of the figures indicating 332 parents exercised the option the first year of the Plan and 181 the following year suggests otherwise. 677 F.Supp. at 1523.

Likewise, the effect of the Board's desire to maintain the District's unitariness by implementing programs like Effective Schools, Student Interaction, Adopt-A-School, and the position of Equity Officer is equally undocumented. The District's Effective Schools program incorporates educational aspirations and attempts to translate those values into enhanced student achievement. (R. VI, 918-19). Although the court hailed a 13% decrease in the gap between black and white achievement test results in the District as evidence that the Effective Schools program was working, the test comparisons are flawed. The group of students studied one year is not the same studied the next year. (R. V, 744); see Def.Ex. 185. While there appeared to be some gain in achievement at eight of the 90%+ elementary schools as measured against the national average, scores at two 90%+ schools dropped. (R. VI, 942). More significantly, the meaning of the gain was not clear.44 Additional testimony established that the Effective Schools program is geared to the upper grades (R. VII, 1004-05) and tied to budgetary constraints experienced by the District. (R. VI, 881). While the testimony was consistent that the concepts of "Effective Schools" and desegregated schools are not mutually exclusive, (R. V, 693) Board witnesses suggested that increased expenditures for busing would necessarily cut into the Effective School's budget. (R. VI, 944). Most importantly, there is no evidence of specific educational programs designed for those racially identifiable elementary schools to counteract the effect of concentrating low achievement in these schools.

The Board designed the Student Interaction Plan to pair schools 90%+ black with schools that do not have significant racial minority populations. (R. IV, 394-95). Teachers were encouraged to bring students together two to four times a year and "to allow children to write letters to each other; to send video cassettes of themselves ... to have the children read the same literature." (R. IV, 395). Although the program was discretionary with each classroom teacher, the Board hoped that perhaps nine to twelve hours a year would be devoted to the Interaction plan. (R. IV, 407). In contrast, plaintiffs' witness, Dr. Crain, rejected the value of student interaction based on exchanges of letters and infrequent visits to a paired school.45 Meaningful interaction, he suggested, took place on a school athletic team or in a boy scout troop.

When the Plan was adopted, an Equity Committee chaired by an Equity Officer46 was established to oversee the District and assure that facilities and equipment were relatively equal throughout the District. While the Board could be responsive to small dollar equipment adjustments (R. VI, 840) or building maintenance problems (R. VI, 834) recommended by the committee, its ability to effect a "major facility repair" or reduce overcrowding, for example, at a predominantly black 5th-year center, was constrained by funding reductions in the District and the consequent need to seek a bond election. (R. III, 370-72).

Similarly, the Adopt-A-School program, "a partnership between the public sector and the private sector," (R. V, 716) to provide resources, speakers, tutors, and money to community schools, does not counteract the imbalance created by the reinstitution of neighborhood schools. Like the considerable testimony about the increased participation of District parents in school PTA-PTO groups,47 this evidence does not substitute for the constitutional mandate to maintain a unitary school system. In other words, while each program is laudable in principle and addresses a particular educational or community value, the program does not ameliorate the condition created by the Plan, the emergence of thirty-two effectively one-race schools.

Although the court accepted the statements of school personnel and community members who "unequivocally testified that in their opinion the Board's K-4 neighborhood school plan was not discriminatory and did not result in the recreation of a dual school system," 677 F.Supp. at 1519, the quest for discriminatory intent is not so straightforward.48 Indeed, the court asked various witnesses, lay and expert, if they believed the Plan was adopted with discriminatory intent.49 One witness was asked if he believed the court should continue its supervision of the District. (R. VI, 953). Thus assured, the court concluded the Plan retained the District's unitariness and, in fact, enhanced it by fostering a greater degree of parental and community involvement in the schools. In light of this conclusion, the court wholly rejected plaintiffs' assertion that the Plan's incorporation of the same attendance zones used prior to the Finger Plan was evidence of the Board's discriminatory intent. Citing Arlington Heights v. Metropolitan Hous. Dev. Corp., 429 U.S. 252, 265, 97 S.Ct. 555, 563, 50 L.Ed.2d 450 (1977), the court responded that "discriminatory intent may not be inferred solely from the disproportionate impact of a particular measure upon one race." In so concluding, the court narrowed its focus and overlooked the broader inquiry mandated by Swann.

Indeed, discriminatory intent cannot be ascertained by eliciting opinion testimony from witnesses, often out of context50 and accumulating those responses as substantive evidence of the motive for the Plan. "Ordinarily, only circumstantial evidence is available to establish segregative intent." Diaz v. San Jose Unified School Dist., 733 F.2d 660, 662 (9th Cir.1984). In Keyes v. School Dist. No. 1, Denver, Colo., 670 F.Supp. 1513 (D.Colo.1987), the court stated:

[S]ome discriminatory intent must be shown to prove a violation of the constitutional requirement that educational opportunity must be equally available. That intent is not, however, measured by the good faith and well meaning of individual Board members or of the persons who carry out the policies and programs directed by the Board. The intent is an institutional intent which can be proved only by circumstantial evidence.

Id. at 1516.

We also agree that neither "the foreseeability of segregative consequences" alone nor the emergence of one-race schools alone establishes a prima facie case of purposeful racial discrimination. Dayton Bd. of Educ., 443 U.S. at 536, n. 9, 99 S.Ct. at 2978 n. 9. We must look at other circumstantial evidence, "the historical background and specific sequence of events leading up to the Board's actions maintaining or exacerbating ethnic imbalance in the district schools." Diaz, 733 F.2d at 663 (citation omitted).

Swann directs "an assignment plan is not acceptable simply because it appears to be neutral." 402 U.S. at 28, 91 S.Ct. at 1282. The Court stated:

[S]uch [neutral] plans may fail to counteract the continuing effects of past school segregation resulting from discriminatory location of school sites or distortion of school size in order to achieve or maintain an artificial racial separation. When school authorities present a district court with a "loaded game board," affirmative action in the form of remedial altering of attendance zones is proper to achieve truly non-discriminatory assignments. In short, an assignment plan is not acceptable simply because it appears to be neutral.

Id. (emphasis added). Within the context of the finding of unitariness, the Plan must be judged by its effectiveness in maintaining unitary status. Davis v. Board of School Comm'rs of Mobile County, 402 U.S. 33, 37, 91 S.Ct. 1289, 1292, 28 L.Ed.2d 577 (1971). Thus, under Swann, we must total all of the evidence to decide if the district court correctly found the Plan maintained unitariness in student assignments.51

It is on this basis that we conclude the district court clearly erred in its findings of fact and consequent legal determinations. While we recognize the dedicated care and tireless, patient effort with which the court has managed and overseen this case, we cannot abandon our obligation to review all of the evidence. Thus, although there is evidence to facially support the district court's findings, on the entire evidence we are "left with the definite and firm conviction that a mistake has been committed." United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 541-42, 92 L.Ed. 746 (1948). Because the court failed to address or distinguish plaintiffs' contrary evidence, and because the court cast the evidence on which it relied in a form to provide an answer to the single question of discriminatory intent, we are convinced that the basis on which the court fashioned dissolution of the injunction was flawed.

Thus, we must focus not on whether the Plan is nondiscriminatory but whether it solves the problems created by the changed conditions in the District. We are certain it does not. While there is evidence of demographic change, that evidence does not support a return to the neighborhood schools in the elementary grades because the same neighborhoods remain predominantly white and predominantly black. Moreover, the Plan restores the effects of past discriminatory intent remedied by the decree by recreating racially identifiable elementary schools, overlooking school capacity problems, and failing to address faculty imbalance. Addressing these shortcomings does not overburden Swann 's remedial baggage.52

We are mindful that Swann contemplates actions which are systemwide in effect,53 and the Plan affects only the elementary schools, leaving the Finger Plan in place for the middle and high schools. Nevertheless, we are convinced that the impact of the Plan on the District as a whole is greater than the sum of each of these parts.54 As Judge Wisdom noted, "Infection at one school infects all schools. To take the most simple example, in a two school system, all blacks at one school means all or almost all whites at the other." United States v. Texas Educ. Agency, 467 F.2d 848, 888 (5th Cir.1972) (quoted in Keyes, 413 U.S. at 201, 93 S.Ct. at 2694). Indeed, "[t]he effect of a racially discriminatory practice is pervasive. That effect is not eradicated by merely erasing the original cause." Lawrence County School Dist., 799 F.2d at 1044.

The focus of our concern remains on the Board's duty. "Part of the affirmative duty imposed by our cases ... is the obligation not to take any action that would impede the process of disestablishing the dual system and its effects." Dayton Bd. of Educ. v. Brinkman, 443 U.S. at 538, 99 S.Ct. at 2979 (1979). Despite its cosmetic trappings, the Plan is such an impediment. While the Board's concerns are not without foundation, they do not translate into the maintenance of a unitary system. "And, it is the responsibility of boards of education and the district courts to prevent the reestablishment of such school systems." Graves v. Walton County Bd. of Educ., 686 F.2d 1135, 1143 (5th Cir.1982).

We are also satisfied based on our review of the record that modification of the Finger Plan is achievable without extreme disruption or burdensome expense. No geographical barriers have created the racial imbalance in the elementary schools. Morgan v. Nucci, 831 F.2d at 313 (unique geographic isolation caused by Boston Harbor justified single one-race middle school).55 Nor was the evidence of white flight sufficient to justify the Plan.56 See Lee v. Anniston City School Sys., 737 F.2d 952, 957, n. 3 (11th Cir.1984) (plans designed to mitigate white flight are permissible provided desegregation effort not frustrated); Liddell v. Missouri, 731 F.2d 1294, 1313-14 (8th Cir.), cert. denied, 469 U.S. 816, 105 S.Ct. 82, 83 L.Ed.2d 30 (1984) (fear of white flight is no excuse to avoid desegregation).

We do not ask that "each school be a statistical image of the [District]." Lawrence County School Dist., 799 F.2d at 1047. Nor do we suggest annual assignment alterations to mirror the ever-changing demographic makeup of this urban setting. However, on the basis of the record, it is clear that other measures that are feasible remain available to the Board " 'to improve the fit between a decade of demographic changes ... and the terms of student access to educational opportunities.' " Morgan, 831 F.2d at 318 (quoting district court order, Morgan v. Nucci, 620 F.Supp. 214, 220 (D.Mass.1985)).

Our remand remains within the framework of the injunctive relief plaintiffs achieved. "Inasmuch as an injunctive decree is drafted in light of what the court believes will be the future course of events, a court must continually be willing to redraft the order at the request of the party who obtained equitable relief in order to insure that the decree accomplishes its intended result." 11 C. Wright & A. Miller, Federal Practice and Procedure Sec. 2961, at 600 (1973) (emphasis added). Remand is not predicated on the perpetual supervision of the district court but preserves the Board's duty "to achieve the greatest possible degree of actual desegregation." Swann, 402 U.S. at 26, 91 S.Ct. at 1281.57

It is well to remember that the course we are running is a long one and the goal sought in the end--so often overlooked--is the best possible educational opportunity for all children. Communities deserve the freedom and the incentive to turn their attention and energies to this goal of quality education, free from protracted and debilitating battles over court-ordered student transportation.

Keyes, 413 U.S. at 253, 93 S.Ct. at 2719 (Powell, J. concurring in part, dissenting in part). Consistent with this goal, and with deliberate speed, the district court, on remand shall:1. Take evidence of plaintiffs' and defendants' alternatives to maintain racially balanced elementary schools within the framework of the changed circumstances that have occurred in the District;

2. Modify the Finger Plan to accommodate the changed circumstances;

3. In modifying the Finger Plan, assure that faculties achieve racial balance under the District's Affirmative Action Plan;

4. In a manner workable to the parties and the court, retain jurisdiction for a reasonable period of time to oversee the implementation and maintenance of these assignments.

The judgment of the district court is thus VACATED58 and the case REMANDED for further proceedings to modify the 1972 decree consistent with this opinion.

BALDOCK, Circuit Judge, dissenting.

It is ironic that the capstone of this court's opinion is a quote from Justice Powell,1 concerning the importance of securing quality education free from "protracted and debilitating battles over court-ordered student transportation." Keyes v. School Dist. No. 1, 413 U.S. 189, 253, 93 S.Ct. 2686, 2719, 37 L.Ed.2d 548 (1973) (Powell, J., concurring in part and dissenting in part); see Court's Opinion at 1505 (quoting same). Here, despite the school district's continued unitary status, this court retains jurisdiction and now orders the school district to racially balance the elementary schools which most certainly will require busing. Court's Opinion at 1506. The court also orders the school district to racially balance the elementary faculties. Id. The district court is to proceed "with deliberate speed." Id. at 1505. See W. Douglas, The Court Years 115 (1980) (discussing delay attributable to Justice Frankfurter's "all deliberate speed" language in Brown v. Board of Educ. (Brown II), 349 U.S. 294, 301, 75 S.Ct. 753, 756, 99 L.Ed. 1083 (1955)).

Apparently, Justice Powell's view concerning transportation of elementary students was overlooked by this court. In fact, in the very next sentence of the passage relied upon by this court, Court's Opinion at 1505, Justice Powell expressed his disapproval: "The single most disruptive element in education today is the widespread use of compulsory transportation, especially at elementary grades." Keyes, 413 U.S. at 253, 93 S.Ct. at 2719 (Powell J., concurring in part and dissenting in part). He urged a return "to a more balanced evaluation of the recognized interests of our society in achieving desegregation with other educational and societal interests a community may legitimately assert." Id.; see also Columbus, 443 U.S. at 479, 486-87, 99 S.Ct. at 2991 (Powell, J., dissenting). Justice Powell's views concerning elementary transportation and other educational objectives, while perhaps not persuasive concerning an officially dual school system in 1973 when Keyes was decided, would seem to have greater force where, as here, a school system has achieved unitary status and seeks to enhance parental involvement, retain students and improve educational opportunity through a return to neighborhood schooling, only in grades 1-4, at the elementary level. The "more balanced evaluation" of educational and social objectives is precisely what is missing from this court's resolution of the case.

This court's opinion is faulty for three reasons: 1) improper reliance upon principles concerning the modification and termination of injunctions which are inapposite in the school desegregation context, 2) inadequate recognition of the legal effect of the school district's unitary status, and 3) insufficient deference to the factual findings of the district court concerning segregative intent.

This case arises in the unusual procedural context of the district court finding the school system unitary in 1977 and terminating jurisdiction of the case. The case was revived when the school district adopted a new student assignment plan, although a 1972 decree had not been lifted. In Dowell v. Board of Educ., 795 F.2d 1516 (10th Cir.), cert. denied, 479 U.S. 938, 107 S.Ct. 420, 93 L.Ed.2d 370 (1986), this court determined that reopening the case was appropriate so that the plaintiffs would have an opportunity to be heard on the constitutionality of the plan. Id. at 1519.

Although Dowell did not express an opinion concerning the constitutionality of the neighborhood school plan at the elementary level, it cast the dispute in terms of injunction modification and termination law, and decided that a finding of unitariness does not alter the defendants' "duty to persist in the elimination of the vestiges of segregation." 795 F.2d at 1520. Of course, a finding of unitariness is inconsistent with remaining vestiges of state-imposed segregation (a constitutional violation), but in this case, the court adheres to its erroneous decision that "a finding of unitariness [does not] mandate the later dissolution of the decree without proof of a substantial change in circumstances which led to issuance of that decree." Court's Opinion at 1492.

Although the court in Dowell did not intend, even implicitly, to address the ultimate issue of constitutionality of the neighborhood school plan, 795 F.2d at 1523, the court framed the issue so as to give no practical effect to the 1977 unitariness finding. The school board retained the burden of proof that its conduct was not segregative; the plaintiffs only were required to prove that the 1972 decree had been violated. Id. In its opinion, the court has adopted that approach and has given no effect to the principle that "the differentiating factor between de jure and ... de facto segregation ... is purpose or intent to segregate." Keyes, 413 U.S. at 208, 93 S.Ct. at 2697 (emphasis in original). This case really turns on the responsibility of these defendants to ameliorate the effects of residential separation so often found in urban locations, including Oklahoma City.

It is beyond dispute that the school board cannot take actions intended to perpetuate or reestablish a dual school system. See Dayton II, 443 U.S. at 538, 99 S.Ct. at 2979 (concerning student assignment, school construction and closings); Columbus, 443 U.S. 449, at 460, 99 S.Ct. 2941 at 2948, 61 L.Ed.2d 666 (concerning school construction and closings); Swann, 402 U.S. at 21, 91 S.Ct. at 1278 (same). Here, the school board deviated from the student assignment plan contained in the 1972 decree (grades 1-4) because the plan no longer was equitable and the board sought to improve the quality of elementary education. Although the school board should have proceeded to district court to have the 1972 decree dissolved, its apparent belief that the 1977 finding of unitariness dissolved the decree was certainly understandable given the state of the law at the time. In any event, we now consider whether the 1972 decree should be enforced or modified. Court's Opinion at 1487.

At the hearing below, both parties sought to modify the district court's 1972 decree requiring implementation of the Finger Plan. The defendants sought to retain their neighborhood school plan at the elementary level. The plaintiffs sought to eliminate, at least in part, the neighborhood school plan and substitute in its place a new plan, prepared by Dr. Gordon Foster, which involved pairing and clustering schools and changing the grade structure of several elementary schools. The plaintiffs contended that the proposed plan was consistent with the 1972 decree, which they maintained had become inequitable, and they contended that the proposed plan was better than the neighborhood school plan. See rec. vol. I, doc. 17, app. A (final pretrial order--plaintiffs' contentions).

Because the school district had attained and maintained unitary status, plaintiffs were required to demonstrate that the district had lost its unitary status through purposefully segregative school board actions. When the entirety of Supreme Court desegregation precedent is applied in the context of a unitary school district, I do not agree with this court that the findings of the district court are clearly erroneous. Rather those findings are supported by the record and merely reflect the district court's choice among expert and lay testimony, which at times conflicted, concerning direct and circumstantial evidence of segregative intent.

Moreover, I question retention of jurisdiction in this case given the unitary status of the school district. Four other circuits hold that current desegregation law does not envision continued federal jurisdiction in the affairs of local school districts in perpetuity once a school district becomes unitary. See United States v. Overton, 834 F.2d 1171, 1176, 1174-77 (5th Cir.1987) ("We see Dowell's effort to declare a school district to be unitary while retaining jurisdiction as an illusion that denies the essence of unitariness.") (Higginbotham, J.); Morgan v. Nucci, 831 F.2d 313, 318 (1st Cir.1987) ("Although the [Supreme] Court has produced no formula for recognizing a unitary school system, the one thing certain about unitariness is its consequences: the mandatory devolution of power to local authorities."); Riddick v. School Bd., 784 F.2d 521, 535, 534-39 (4th Cir.1986); ("[O]nce the goal of a unitary school system is achieved, the district court's role ends."); Vaughns v. Board of Educ., 758 F.2d 983, 988 (4th Cir.1985) ("A district court's jurisdiction to grant further relief in school desegregation cases in not perpetual.... Once a school system has achieved unitary status, a court may not order further relief to counteract resegregation that does not result from the school system's intentionally discriminatory acts."); Spangler v. Pasadena Bd. of Educ., 611 F.2d 1239, 1241 (9th Cir.1979); Id. at 1247 ("Retention of jurisdiction when there is no longer a demonstrated need to monitor compliance may defeat important governmental and personal interests.... There is no warrant for retaining further jurisdiction in the case, and the district court should enter an order relinquishing all further jurisdiction.") (Kennedy, J., concurring);2 but see United States v. Board of Educ., 794 F.2d 1541, 1543 (11th Cir.1986) (per curiam) (when parties have relied on desegregation orders to make commitment to start desegregation process, it may be inappropriate to vacate such orders even after unitary status attained). Contra Dowell, 795 F.2d at 1516 (10th Cir.). This court's "semantic recasting of unitary status," Overton, 834 F.2d at 1176, in terms of prohibitory injunction law aimed at private parties and relief from judgment, Fed.R.Civ.P. 60(b), displaces the concept of unitary status and insures federal involvement, even in the absence of a constitutional violation.

To impose onerous duties on the school board, this court improperly relies upon desegregation passages involving dual systems charged with the affirmative duty to eradicate de jure segregation and its vestiges. See, e.g., Court's Opinion at 1499 (relying upon Swann 's quotation of Green v. County School Bd, 391 U.S. 430, 439, 88 S.Ct. 1689, 1694, 20 L.Ed.2d 716 (1968), 402 U.S. at 13, 91 S.Ct. at 1274, for affirmative duty); Court's Opinion at 1503 (relying upon Swann, 402 U.S. at 28, 91 S.Ct. at 1282, for discussion of neutral assignment plans which may not be compatible with affirmative duty); Court's Opinion at 1503 (relying upon Davis v. Board of School Comm'rs, 402 U.S. 33, 37, 91 S.Ct. 1289, 1292, 28 L.Ed.2d 577 (1971), for concept that desegregation plan must be effective, given remedial responsibility of local board); Court's Opinion at 1493, 1504-05 (relying upon Dayton II, 443 U.S. at 538, 99 S.Ct. at 2979, concerning board's affirmative duty not to perpetuate or reestablish dual school system and the heavy burden the board faces in explaining actions not consistent with the affirmative duty). We have a very different situation in this case--this board has met its affirmative duty and no longer is under "an unsatisfied duty to liquidate a dual system." See Dayton II, 443 U.S. at 538, 99 S.Ct. at 2979.

Under any test in which segregative intent is relevant, I cannot agree that the neighborhood school plan for the elementary grades 1-4 had "the effect of making the District 'un -unitary' by reviving the effects of past discrimination." Court's Opinion at 1499 (emphasis in original); see also id. at 1504 ("the Plan restores the effects of past discriminatory intent remedied by the decree") (emphasis in original).3 Of course, the word "un-unitary" seems unnecessary in the lexicon of school desegregation law; an "un-unitary" system is a dual system, and the inquiry here is whether the neighborhood school plan signals the return of a dual system. This court overturns the district court's decision to dissolve the injunction implementing the 1972 desegregation plan for two reasons: 1) because the district court "failed to address or distinguish plaintiffs' contrary evidence," and 2) because the district court "cast the evidence on which it relied in a form to provide an answer to the single question of discriminatory intent...." Id. at 1504. Both reasons given by this court are wanting.

Under Fed.R.Civ.P. 52(a) and Anderson v. Bessemer City, 470 U.S. 564, 574, 105 S.Ct. 1504, 1511, 84 L.Ed.2d 518 (1985), we are not at liberty to reverse the district court's factual findings unless they are clearly erroneous. Where, as here, "there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous." Id. Admittedly, the failure to address truly contrary evidence may leave the door open for the court of appeals to find internal inconsistencies or implausibility in the evidence relied upon, see id. at 575, 105 S.Ct. at 1512, but the ultimate inquiry is whether the "district court's account of the evidence is plausible in light of the record viewed in its entirety." Id. at 574, 105 S.Ct. at 1511. Considering what issues are material in this lawsuit, I would affirm the district court. There is ample evidence to support the finding that, even with the introduction of a neighborhood school plan in grades 1-4, the school board still operates a unitary system.

One of the most serious technical problems with this court's decision is the methodology of identifying racially disproportionate schools after implementation of the neighborhood plan. This court claims that 32 out of 64, or one-half, of the elementary schools are now "one-race majority schools." Court's Opinion at 1493, 1497, 1502. We are told that only "[t]hirty-two elementary schools remained racially mixed." Id. at 1487. This analysis is flawed. The district court properly was concerned with the emergence of 90+ black schools as a result of the neighborhood plan. Dowell v. Bd. of Educ., 677 F.Supp. 1503, 1509-10 (W.D.Okla.1987). Here, this court correctly notes that 11 out of 64 elementary schools are 90% + black after the plan. Court's Opinion at 1486-87. But then the court adds to the number of racially disproportionate schools by deciding that if a school has less than 10.7% black enrollment, it is not racially mixed, regardless of the presence of other minorities. See id. at 1487, 1487 n. 2. A significant portion of the data in this case only distinguished between black students and other students.4While such date clearly indicates concentration of black students,5 such an approach fails to recognize the presence of other minorities which increased since 1972; in 1985-86, non-black minorities comprised 13.3% of the elementary system enrollment (Hispanic, 6.8%; Native American, 4.2%; Asian-American, 2.3%). See Dowell, 677 F.Supp. at 1510; def. ex. 63, reproduced in Appellees' Addendum to Brief.

This is a school district in which the elementary students are 50.7% white, 36.0% black and 13.3% other minority (total minority percentage of 49.3). Dowell, 677 F.Supp. at 1510 (1985-86). If we compare the white enrollment with the aggregated number of minority students as is appropriate under Keyes, 413 U.S. at 197-98, 93 S.Ct. at 2691-92; see also Riddick, 784 F.2d at 527 n. 7 (minority percentage aggregated, not expressed solely in terms of black enrollment), and as was done by the district court, 677 F.Supp. at 1509-10, the 21 extra schools identified by this court range from 13.8% (Lafayette) to 58.5% (Willard) minority enrollment--these extra 21 schools are not one-race schools, even if they do not contain the system-wide average (40%) of black students.6 Indeed, some of these schools approach the elementary system-wide minority percentage, e.g. Lee, 44.5% minority; Willard, 58.5%. Insofar as the permissible inference of segregative intent is concerned, the 21 schools identified by this court simply do not have the same implications as the eleven 90%+ black schools.7 This is true notwithstanding Dr. Finger's reluctance to agree that schools which have more than 10% minority enrollment, but less than 10% black enrollment, are not segregated white schools. See Court's Opinion at 1496.

Indeed, one of the limitations of the 1972 Finger Plan is that it sought racial balance only with respect to black students. Rec. vol. III at 321. Dr. Biscoe, school administrator, was aware of this feature and testified that notwithstanding, it is important to look at total minority percentages, not just black, when evaluating the district. She said:

Since I work a lot with federal programs and attend a lot of those meetings, as I've gone around the country one thing I've noticed that happens in many of the school districts is that people look at the percent of minority students in their district, including Orientals, Native American students, Hispanic students.

Of course[,] if you look at our district in terms of total minority population as compared to white population, you get a totally different picture than if you just look at those data black-white.

Id. The district court correctly decided to aggregate minority percentages when reviewing the effects of the neighborhood plan.

Although the district court is criticized for its finding concerning intent, Court's Opinion at 1504, it is apparent that this court is merely reweighing the evidence and crediting the circumstantial evidence of segregative intent--namely the emergence of one-race schools. See id. at 1502-04. Of course, reweighing the evidence is not consistent with our appellate function. See 5A J. Moore & J. Lucas, Moore's Federal Practice, p 52.03 (1989). This court is under the mistaken impression that circumstantial evidence of segregative intent must be preferred in these types of cases. See Court's Opinion at 1503 (quoting Keyes v. School Dist. No. 1, 670 F.Supp. 1513, 1516 (D.Colo.1987), appeal pending, Nos. 85-2814 & 87-2634 (10th Cir. submitted Jan. 17, 1989).8 Despite the court's contrary conclusion, the district court properly weighed all of the evidence and came to a permissible conclusion that there was an absence of segregative intent necessary support a constitutional violation.

Another obvious problem with the court's resolution of this case is its reliance upon general principles concerning the modification of injunctions when principles specifically concerning the process of desegregation have been enunciated by the Supreme Court. This court relies upon the standard of United States v. Swift & Co., 286 U.S. 106, 119, 52 S.Ct. 460, 464, 76 L.Ed. 999 (1932), an antitrust case, for the proposition that, despite unitariness, the school board must show that the condition the injunction was designed to protect against has abated and the failure to modify the injunction will result in extreme hardship to the school board. Court's Opinion at 1489-91. It is not enough that the constitutional violation has been remedied. Id. at 1490-91 ("The condition that eventuates as a function of the injunction cannot alone become the basis for altering the decree absent the Swift showing."). This court severs the integral link between the desegregation decree and an extant constitutional violation contending that " 'an injunction takes on a life of its own and becomes an edict quite independent of the law it is meant of effectuate.' " Id. at 1490 (quoting T. Jost, From Swift to Stotts and Beyond: Modification of Injunctions in the Federal Courts, 64 Tex.L.Rev. 1101, 1105 (1986)). How ridiculous that an injunction, meant to effectuate the law, "becomes an edict quite independent of the law it is meant to effectuate." See id.II.

Given the court's position on insuring prospective relief even in the absence of a constitutional violation, unitariness is not a bar to reconsideration of the decree. Court's Opinion at 1491. The court tempers its reliance on the exacting standard required for modification of injunctions by citing the dissenting portion (without so indicating) of Judge Higginbotham's opinion in United States v. Lawrence County School Dist., 799 F.2d 1031, 1056 (5th Cir.1986) ("By its forward cast, an injunction contemplates change and thus must be sufficiently malleable to adapt the ordered relief to contemporary circumstances.") (Higginbotham J., concurring in part and dissenting in part). See Court's Opinion at 1491; see also Pasadena Bd. of Educ. v. Spangler, 427 U.S. 424, 437, 96 S.Ct. 2697, 2705, 49 L.Ed.2d 599 (1976) (recognizing propriety of modification when there is continuing supervision). Moreover, Judge Higginbotham, writing for a unanimous panel, quite firmly has rejected this court's approach in Dowell. Overton, 834 F.2d at 1174-77.

Swift simply does not apply to the vast majority of desegregation decrees. In Spangler v. Pasadena Bd. of Educ., Justice (then Judge) Kennedy explained the inapplicability of Swift to a district court's termination of equitable jurisdiction in school desegregation cases. Spangler, 611 F.2d at 1245 n. 5.

Swift involved the efforts of antitrust defendants who had entered into a consent decree prohibiting anticompetitive actions to modify the decree by lifting some of its prohibitions. It is doubtful the case supports the district court's retention of jurisdiction. Swift establishes general criteria for dissolution or modification of prohibitory injunctions against private wrongdoers. More recent Supreme Court desegregation decisions have established specific criteria for dissolution of regulatory injunctions imposed upon public school authorities ... [T]hese criteria recognize (1) that the proper function of a school desegregation decree is remedial, and (2) necessary concern for the important values of local control of public school systems dictates that a federal court's regulatory control of such systems not extend beyond the time required to remedy the effects of past intentional discrimination. See Milliken v. Bradley [Milliken II ], ... 433 U.S. [267,] at 280-82, 97 S.Ct. 2749, 2757-58, 53 L.Ed.2d 745 [ (1977) ].

Id. Justice Kennedy's reasoning is equally applicable to this court's retention in perpetuity of equitable jurisdiction over a unitary system, using the desegregation decree as the vehicle.

Virtually all of the cases relied upon by this court in its application of Swift involve unsuccessful attempts to obtain, modify or dissolve injunctions which forb