Robinson Ex Rel. Robinson v. Shelby County Board of Education

ALGENON L. MARBLEY, District Judge,

dissenting.

In reversing the district court’s carefully considered judgment denying unitary status to the Shelby County Schools on the subjects of student desegregation, faculty desegregation, and extracurricular activities, the majority’s reasoning comes to this: the parties have jointly stipulated to dismissal and such compromises should be encouraged by the courts, especially in the divisive realm of school desegregation.

I have no particular quarrel with this general principle, but I do not believe that the virtues of compromise can compensate for the lack of evidence substantiating that the County has in fact eliminated, to the extent practicable, all remaining vestiges of unlawful discrimination. Nothing — not the agreement of the parties jointly to seek dissolution of the desegregation decree, not the number of years that this case has been pending and the general progress in race relations nationwide that has occurred in that time, and not the eagerness of the courts or school boards to restore local control over community schools — can substitute for evidence showing the Board’s compliance with the desegregation decree. The evidence in fact reveals that among the forty-four schools for which the Board has data, two thirds of them are not in compliance with the flexible benchmark set forth by the district court for measuring racial balance. Furthermore, the district court has been managing this case since 1963. Based on the court’s knowledge and experience with this case, it is in the best position to judge the evidence and determine whether dissolution of the desegregation decree is appropriate at this time. Because the parties have not carried their burden of showing that the racial disparities that continue to plague the County’s schools are not the vestiges of past unlawful discrimination, I would affirm the district court’s judgment.

I. BACKGROUND

A. The History of Desegregation in Shelby County

The majority opinion does not recount the history of this lengthy and complicated case. Because I believe that we must consider the entire record to evaluate properly the district court’s judgment, I set forth the most important facts of the case below.

1. 1963-1971

The years between 1963 and 1971 were by far the most active in the case’s history. This period was marked by the County’s early intransigence in adopting appropriate measures to de-segregate its schools, the intervention of the Department of Jus *658 tice (the “Government”) to pressure the County to take its desegregation obligations seriously, and ultimately, in 1971, the district court’s approval (following a reversal and remand by this Court) of a comprehensive plan to eliminate the racial identifiability of all the County’s schools.

On June 12, 1963, nine years after Brown v. Board of Education, 347 U.S. 483 , 74 S.Ct. 686 , 98 L.Ed. 873 (1954) was decided, twenty-one public-school students brought this class action against the Shelby County Board of Education (the “Board”) seeking a declaratory injunction that the public schools were unconstitutionally segregated and an injunction requiring the Board to integrate them. (Joint Appendix (“JA”) 156.) In response to the Plaintiffs’ complaint, the Board denied any wrongdoing, but nonetheless submitted a plan to the district court, which approved it on March 17, 1964. (JA 156— 160.)

This first effort amounted to no more than a “freedom-of-choice” plan, and it barely qualified as that. “Freedom-of-choice” plans purported to put an end to segregated schools by permitting African-American students voluntarily to choose to attend the all-white schools from which they had long been excluded. The Board’s plan, however, erected numerous obstacles to exercising free choice. It conditioned the transfer of African-American students into white schools on a showing of good behavior, acceptable academic performance, sufficient family income, and psychological stability. (JA 158.)

In 1966, two years after the district court approved the Board’s freedom-of-choice plan, the Government intervened in the case. . (JA 160.) In response to a Government motion criticizing the Board for lack of progress in desegregating its schools, the district court entered an order making minor modifications to the decree, but these changes still did not promise to make the mandate of Brown a reality. (JA 162-63.) By 1967, four years after the Plaintiffs filed suit, and thirteen years after Brown , 100 percent of white students attended the formerly all-white schools and 98.7 percent of African-American students attended the formerly all-black schools. (JA 165.)

On January 19, 1967, in response to another Government motion contending that the Board was shirking its desegregation obligations, the district court again modified the decree. (JA 284.) The district court tinkered with the plan to make it slightly easier for African-American students to transfer to the formerly all-white schools, but the anemic “freedom-of-choice” approach still ruled the day. Although continuing to fall short in terms of desegregating the student bodies of the Shelby County Schools, the district court’s January 19, 1967 order was notable because for the first time it spelled out precise benchmarks for desegregating the County’s faculties. (JA 284.) The court held that a faculty would be regarded as desegregated when its racial composition reflected the County-wide composition within a deviation of ten percentage points. (JA 286.) To that end, the court ordered the Board to fill all faculty vacancies with teachers whose race was under-represented in the school at issue. (JA 286.) In addition, the district court ordered the Board to implement a program to recruit white teachers to work in schools whose faculties were predominantly African-American and African-American teachers to work in schools with predominantly white faculties. (JA.287.)

To measure the Board’s compliance with these requirements, the district court ordered it to file certain reports on a regular basis. First, before filling a faculty vacancy with a teacher of the over-represented *659 race, the Board was required to notify the Court of its intention to do so, as well as explain its efforts to transfer or hire a teacher of the under-represented race and why those efforts failed. (JA 288-89.) In its July 26, 2007 order declining to dissolve the desegregation decree, the district court found that throughout this litigation, the Board has never once filed this report before transferring/hiring a teacher of the over-represented race to fill a vacancy. 1

Second, the Board was instructed to submit, on August 1 of each year, a report providing data on the racial make-up of the faculty in each school, as well as the number of vacancies that were filled by faculty of the over-represented race. (JA 289.) The August 1 report was to be supplemented on October 1 of each year. (JA289.)

On August 1, 1967, the Board submitted its first annual report on teacher desegregation. This report showed that the Board employed 1500 teachers. Pursuant to the district court’s order that the Board re-assign teachers of the opposite race “in all cases in which the transfer can be accomplished without seriously impairing the educational program,” (JA 287), the Board reported that it had re-assigned just 128 teachers, or 8.5 percent of the total number. The Board’s October supplement showed that of the 200 vacancies filled for the upcoming school year, only twelve of these were filled by teachers of the underrepresented race.

On May 27, 1968, the Supreme Court decided Green v. County School Board of New Kent County, 391 U.S. 430 , 88 S.Ct. 1689 , 20 L.Ed.2d 716 (1968). There, the Court held that “freedom-of-choice” plans were generally inadequate to satisfy the mandate of Brown . Leaving no doubt about the gravity of the responsibility facing local school boards, the Green Court stated, “The burden on a school board today is to come forward with a plan that promises realistically to work, and promises realistically to work now.” Id. at 439 , 88 S.Ct. 1689 . School boards were instructed to take whatever remedial steps were required to eliminate racial discrimination “root and branch.” Id. at 438 , 88 S.Ct. 1689 .

The district court issued another order on July 17, 1968, clarifying the Board’s desegregation responsibilities in light of Green . (JA 171.) As to faculty desegregation, the court reiterated that the ratio of African-American to white teachers in each school was to reflect the County-wide ratio within a margin of ten percentage points. The court held that this ratio had to be satisfied in the County’s elementary schools (defined as grades one through six) by the start of the 1968-69 school year and that it was to be met in the remainder of the County’s schools by the start of the 1969-70 school year. (JA 172.) The court further ordered the Board to prepare a new desegregation plan designed to accomplish these goals and otherwise eliminate the racial identifiability of the County’s schools. (JA 172.)

On August 15, 1968, the district court provided additional guidance about the preparation of-the Board’s new desegregation plan. (JA 173.) For the first time, the district court set a benchmark for desegregating the County’s student bodies. The court ruled that insofar as feasible, the Board should assign students so that the ratio of African-American to white students in each school reflected the Countywide ratio, within a margin of ten percentage points. (JA 173.)

The Board submitted its post-Green plan for the district court’s review on Jan *660 uary 15, 1969. This plan did not come anywhere close to achieving the student racial balance that the district court had ordered. (JA 175.) The Board admitted that its plan did not require the desegregation of high school students, but instead permitted them to remain in the high schools that they were already attending. (JA 175.) As to faculty desegregation, the Board sought relief from the obligation of achieving any particular racial ratio among high school teachers, let alone that specified by the district court (a black-to-white ratio in each school that reflected the ratio in the County as a whole, plus or minus ten percentage points). (JA 174.) The Board argued that the work of desegregating high school faculties was made more difficult by the fact that these teachers were licensed in particular subject areas.

The Plaintiffs and the Government opposed the Board’s post-Green plan as inadequate. (JA 175-78.) The district court nonetheless approved it.

On April 6, 1970, the district court approved certain modifications to the Board’s desegregation plan. See Robinson v. Shelby County Bd. of Educ., 311 F.Supp. 97, 104-05 (W.D.Tenn.1970). Importantly, the court backed away from its earlier insistence that the County achieve a student racial composition within each school that reflected the County racial composition within ten percentage points. Rather, the court held that the Board was not required to achieve any particular degree of racial balance among its students and that the most that could be asked was for the Board to “honestly draw[] unitary geographical zone lines, that is, zones not gerrymandered to preserve segregation ....” Id. at 102 . The court stood by its prior pronouncements with respect to faculty desegregation but added the caveat that efforts to de-segregate high school teachers should be tempered by considerations of their qualifications. The court stated, “to the extent feasible, in the light of the qualifications of the teachers and the need for teachers of particular qualifications in the [secondary] school, such teachers will be assigned and transferred so that the ratio of white to Negro teachers in each school will be, within a tolerance of 10%, the same as in the system as a whole.” Id. at 105 .

On appeal of the district court’s April 6, 1970 order, this Court remanded the case for further consideration of the County’s desegregation obligations. See Robinson v. Shelby County Bd. of Education, 442 F.2d 255 , 258 (6th Cir.1971). This Court held that the district court had misapprehended the extent of the County’s affirmative duty to undo the effects of its past discriminatory conduct. Id. at 258 (“Where there has been a history of state-imposed segregation of the schools, it is not sufficient to adopt a plan which, out of context, might be seen as nondiscriminatory but which does not do as much to disestablish segregation as an alternative proposal which is feasible and pedagogically sound.”).

In response, the district court approved a revised desegregation plan in August 1971. See Robinson v. Shelby County Bd. of Educ., 330 F.Supp. 837, 843-47 (W.D.Tenn.1971), aff'd 467 F.2d 1187 (6th Cir.1972). This plan was largely that advanced by the Board, but the district court also accepted certain suggestions from the Title IV Center at the University of Tennessee, as well as the Government. Id. at 843. This latest effort apparently dealt exclusively with student desegregation. Id.

2. 1971-2006

The Board’s 1971 desegregation plan approved by the district court marked the last substantial revision to the Board’s de *661 segregation efforts in the history of this case. It was also virtually the last time that either the Plaintiffs or the Government challenged any of the Board’s decisions as contrary to its desegregation obligations, or argued that the Board was not complying with the plan’s requirements. Perhaps due in large part to the lack of adversarial litigation, the record is rather thin on evidence showing what progress, if any, the Board made toward dismantling the vestiges of unlawful discrimination in its schools. It appears that neither the Plaintiffs, the Government, nor the district court required much in the way of statistical data tracking the racial composition of the County’s students and faculty over time.

To begin with, very little can be gleaned from the record about what was happening with student desegregation. There is almost no evidence documenting the racial composition of each school from year to year (and comparing that to the Countywide ratio), the indicators the Board used, if any, to gauge its progress, the obstacles the Board confronted, or how it made decisions in operating the County’s schools to ensure its full compliance with the desegregation decree. The definition of “success” and the Board’s path to arrive there are not clear. For instance, between August 20, 1974 and August 3, 2004, the district court entered more than fifty consent orders modifying the decree as to such things as school-attendance zones and new-school construction. (JA 192-234.) It appears that the Board rarely provided information about the impact these modifications would have, on the student racial composition of each of the schools. The record suggests that the Board often did no more than conclusorily state that the changes were not expected to have any deleterious effects on the desegregation plan. (See, e.g., JA 219-26.)

Moreover, Plaintiffs rarely interposed any objections to the Board’s plans and the same was true of the Government. On one of the few occasions that serious opposition to the Board’s plans was lodged, the district court found that the Board had allowed improper considerations to trump its desegregation duties. The dispute centered on the Board’s desire in 1985 to add ten new classrooms to an overwhelmingly white elementary school to alleviate overcrowding. (JA 203.) The Government opposed the construction project (it does not appear that the Plaintiffs joined the Government), claiming that the Board impermissibly wanted to avoid utilizing the excess capacity of nearby elementary schools with significant African-American populations. (JA 203-04.) On April 8, 1986, the district court denied the Board’s petition to build the extra classrooms. (JA 209.) The court agreed with the Government that the Board had failed to give due consideration to its desegregation obligations, and that it had elevated “community pride” (meaning the community pride of the white elementary school students and parents) in preserving present enrollment at the over-crowded elementary school over its duty to eliminate unlawful discrimination. (JA 209-10.)

The record regarding faculty desegregation between 1971 and 2006 is only slightly more robust. After the Board’s initial report on August 1, 1967, the record is silent about the Board’s efforts to re-assign teachers. As noted above, the Board reported that it had re-assigned 8.5 percent of its teachers, effective during the 1967-68 school year. The record does not disclose whether the Board undertook any more reassignment efforts after 1967. Certainly, the Board does not state in its briefing to this Court that it did so, nor does it point to any evidence of such. Thus, it appears that in response to the district court’s January 19, 1967 order, the *662 Board had done all the faculty re-assigning it intended to do by August 1,1967.

With respect to teacher transfers to fill vacancies, the district court had instructed the Board to file annual reports showing the number of openings filled by teachers of the over-represented and under-represented race. Between 1974 and 1979, the Board’s reports show that half the time, a majority of vacancies were not filled by teachers of the under-represented race (as ordered by the district court). (JA 133; 194-95; 198-201.) It appears that the Board stopped reporting this statistic altogether by 1981. Neither the Plaintiffs, the Government, nor the district court seems to have taken issue with the Board’s abandonment of this requirement. In addition, although the Board was supposed to notify the court before filling a vacancy with a teacher of the over-represented race and justify its inability to employ an opposite-race teacher, the record is bereft of any evidence that the Board ever did so, and the Board does not dispute that it did not.

Besides the Board’s lackluster track record as to faculty re-assignment and vacancies, the record discloses another deeply disturbing trend. Between 1974 and 2006, the total percentage of African-American faculty at the County’s schools declined markedly. The Board’s 1974 annual report showed that thirty-three percent of the County’s faculty were African-American, which correlates closely to the fact that around this same period, thirty percent of the County’s students were African-American. In 2006, only fifteen percent of the County’s faculty were African-American, even though around this same period, a significantly higher percentage, thirty-four percent, of the County’s students were African-American.

In August 1989, the district court was so disturbed by the decline in the Board’s employment of African-American teachers that it asked the parties to address the question of whether the Board’s reports “for the reporting period of August 1, 1985, to August 1, 1989, indicate an employment practice or policy within the Shelby County Schools which achieves a gradual but definite decline in the number of black teachers employed by the school system.” (JA 214.) The Board responded that the problem was attributable to fewer African-American people becoming teachers. (JA 214.) The Government had a different position, noting that between 1972 and 1989, the number of white teachers had more than doubled, but the number of African-American teachers had remained virtually the same, that the Board had “recruited almost 10 times as many white applicants as black applicants,” and that its offer rate for white candidates was much higher than for African-American candidates. (JA 215.) The Government also cited statistics showing that the Memphis City Schools (located within Shelby County, but not part of the Shelby County school system) did a much better job of recruiting African-American applicants, even though the starting salary for teachers in Memphis was slightly less than that offered by the Board. (JA 215-16.) The Government also argued that the Board was much less proactive than it could have been in its outreach to colleges and universities, particularly to historically African-American colleges. (JA 216.) On July 3, 1990, the district court ordered the Board to submit supplemental annual reports regarding its minority recruiting practices. (JA 218.)

The Board filed regular supplemental reports on minority recruiting between 1990 and 2006. These reports both confirm that the Board expanded its outreach to colleges and universities, including historically African-American colleges, and provide data on the offer rates for Afri *663 ean-Ameriean and white candidates. In the early years, offer rates for African-American candidates sometimes exceeded those for white candidates; but by 1997, offers to white candidates always outpaced offers to African-American candidates, sometimes significantly. See, e.g., JA 228 (October 1999 report showing that twenty-six percent of African-American applicants interviewed through the central office were given offers, compared to fifty-one percent of white applicants).

Thus, the extent of the Board’s compliance with the desegregation decree between 1971 and 2006 is difficult to assess. During that time, the Plaintiffs never challenged any of the Board’s decisions. Similarly, the Board did not apprise the court (and the court apparently did not inquire) of any factors that impeded its progress desegregating as much as practicable or any steps it took to alleviate those impediments. The Board did not even bother to submit the required reports on faculty desegregation. Whether the racial disparities that continue to mark the Shelby County Schools are due to reasons wholly unrelated to the ongoing effects of unlawful discrimination, as the majority contends, or whether the Board could and should have done more to attain its desegregation objectives, are not answered by the record as it existed at the time the parties moved for dissolution of the decree. Moreover, as described below, the parties did not fill this evidentiary gap at the two hearings conducted by the district court to consider their motion.

3. The Joint Motion to Dissolve the Desegregation Decree

On August 14, 2006, the Plaintiffs and the Board jointly moved for an order dissolving the desegregation decree and declaring the public schools “unitary.” The parties asserted that they had fully complied with the decree and that they had met the standards articulated by the Supreme Court for eliminating the effects of past de jure segregation. Despite the joint nature of the application, the district court recognized that it “ha[d] an obligation to independently evaluate these factors and the evidence to assure that the system as it exist[s] now is truly unitary and color blind.” (JA 1187.) The. court therefore held evidentiary, hearings on January 26, 2007 and July 23, 2007.

The two hearings consisted largely of anecdotal testimony by school officials and parents. For example, Superintendent Bobby G. Webb testified that his staff worked diligently to adjust school attendance zones to make sure that the County has “good community-based schools and have them as diverse as we possibly can.” (JA 1204.) He stated, “I can tell you without a doubt that — that all of the staff that I have to work with ... certainly love and respect every child regardless of their color, background or whatever.... ” (JA 1205.) When asked by the district court what evidence it should consider in determining whether the Board’s efforts to promote color blindness will continue in the absence of court monitoring, Webb responded: “I personally can assure you that as long as I’m superintendent there will be no discrimination in any shape, form, or fashion.” (JA 1230.)

Assistant Superintendent of Planning and Student Service Maura Sullivan testified that- she had worked closely with Plaintiffs’ counsel over the years to adjust attendance zones based on enrollment projections and assess the impact of the adjustments on school demographics. (JA 1239-40.) Assistant Superintendent Lois Williams described the County’s efforts to recruit minority teachers. She identified the biggest barrier to minority-teacher recruitment as the shrinking teaching pool *664 overall and the shortage among minority candidates in particular. (JA 1315.) The Board’s goal, according to Williams, is for fourteen percent of the teachers in each of the County’s schools to be minorities. (JA 1302.)

During the July 23, 2007 hearing, the Plaintiffs and the Board called two parents to testify. ' The first, Ricky Jeans, had himself been a student in the Shelby County Schools in the late 1960s and had been one of the first African-American students to attend a formerly all-white school. (JA 1378-79.) Jeans testified in favor of dissolving the desegregation decree, stating, “I think that it’s time that we — we look backwards and see where we started at and look at the improvement of the system and what has gone on down through the years and support maybe dropping the system from this point on, but I am a big advocate of the Shelby County school system.” (JA 1379.) The second, Brenda Gipson, testified that the Shelby County Schools had become more diverse both in their student bodies and faculties since she had been involved with them through her children. Gipson testified that school administrators are “for an environment of inclusiveness — inclusiveness, they go out of their way to make sure that everyone feels important in the school system.” (JA 1387.) Finally, Gipson testified that some schools in the County have a largely African-American student body but that “obviously most of the ones in my area are predominantly white.” (JA 1388.)

The evidence presented at the two hearings was thus largely anecdotal and based on the personal views of the interested parties and two parents. Neither the Board nor Plaintiffs put any school-desegregation experts on the stand to testify about the Board’s performance since the desegregation decree was imposed. Neither party put any witnesses on the stand to opine that the Board had achieved desegregation to the extent feasible or that it had availed itself of all opportunities to desegregate. Finally, there was no expert testimony about demographic changes in the County or about how such changes should be interpreted in light of the fact that the Board had formerly practiced de jure segregation.

B. The District Court’s Order

On July 26, 2007, , the district court issued a sixty-two page opinion, supplemented by a one-hundred-and-sixteen page “Procedural Appendix” of the case. The district court granted the parties’ joint motion as to facilities, transportation, and staffing, finding that there was no evidence that the Board had been discriminatory over the last few decades in managing these aspects of the schools. (JA 115.) A declaration of unitary status was therefore found to be appropriate as to these three areas. The court, however, denied the joint motion as to student assignment, faculty assignment, and extracurricular activities and held that the court would continue to supervise these areas. (JA 116.)

1. Student Assignment

The district court found that by 1971, the Board had begun to make real progress toward desegregating the student bodies of its schools. (JA 106.) Still, “at least a quarter of the schools remained outside the target racial composition range established by the [court].” (JA 107.) The court also noted that by 1984, the proportion of African-American students in each school continued to be “uneven,” with three schools out of thirty-four having fewer than five percent African-American students and three having between fifty and seventy percent. (JA 108.) These disparities have only widened. The district court found that by the 2004-05 *665 school year, only seventeen of the County’s forty-six schools had a racial make-up that reflected, within ten percentage points, the racial make-up of the County as a whole. (JA 108.) Moreover, the County’s new “state-of-the-art” high school is expected to have an eighty-eight percent or higher African-American student population. (JA 108.)

The district court concluded that “[t]he Board has made no showing that racial balance is infeasible either generally or with regard to certain schools.” (JA 111.) The court further found that “the record does not indicate that the County has at any time accomplished its objectives.... ” (JA 111.) The racial compositions of the student bodies of the individual schools relative to the County as a whole are important, reasoned the court, because such concrete and detailed statistics enable a court to gauge progress towards eliminating the “racial identifiability” of schools, which is the key to achieving unitary status. Without a statistical comparison, the court said it would be left to “rely on anecdotal evidence, gut feelings, and assurances from defendants that they ‘love and respect every child regardless of their color — hardly a proper basis for making momentous legal decisions.’ ” (JA 110.) At the same time, the court recognized that “racial balance is not a strict requirement and there may be mitigating circumstances which would allow a school system to operate one or more schools with a predominance of one race or another without running afoul of its equal protection obligations — ” (JA 109.)

With these principles in mind, the district court directed the Board to work toward achieving racial balance in all its schools, which it defined as a student racial composition in each school that mirrors the County-wide ratio within a margin of fifteen percentage points. (JA 117.) The court characterized this benchmark as a “flexible goal” and “a starting point in analyzing the Board’s success in desegregation.” (JA 117.)

2. Faculty Assignment

The district court declined to grant unitary status as to faculty assignment, finding that the Board had not exhibited good-faith compliance with the court’s prior desegregation orders. (JA 102-03.) The court found it “somewhat astonishing” that the Board reassigned only 128 teachers out of 1500 in response to the order directing that this be done “in all cases” provided that the educational program was not seriously impaired. (JA 103.) The court was also troubled by the lack of full compliance with the requirements for transferring teachers when vacancies arose. The district court concluded that

Strict compliance with the new faculty integration requirement would have almost surely effected a very rapid transition to a system where the faculty at each school was reflective of the system-wide racial composition. This seems apparent because practically any vacancy at a school whose faculty was racially disproportionate would have set off a chain of intra-system transfers which would only stop when either complete racial balance was achieved or a vacancy resulted which was impossible to fill from within the system.

(JA 102.) Finally, the district court observed that the Board had “utterly] disregard[ed]” the requirement that it notify the court before filling a faculty opening with a teacher of the over-represented race. (JA 103.) The district court concluded that the wide gulf between the proportion of African-American teachers at various schools in the County today, ranging from a low of five percent to a high of *666 thirty percent, showed that the Board’s efforts were wanting. (JA 112.)

When it came to fashioning a remedy, the court disavowed its earlier approach of aiming to have a percentage of African-American teachers in each school that mirrored the percentage of African-American teachers in the County. (JA 113.) This approach, said the court, “created the perverse incentive to allow the overall black faculty representation to slip.... ” (JA 113.) The court feared that the Board could reduce the County-wide percentage of African-American teachers by terminating African-American teachers. As the County-wide percentage fell, so too could the percentage required in each school, without raising the specter of any disparity between the two. (JA 113.)

Having identified what it deemed to be a fatal flaw with its longstanding method of measuring progress toward faculty desegregation, the district court came up with another method. It ruled that, henceforth, the proper comparison would be between the percentage of African-American teachers in a school and the percentage of African-American students in the County. (JA 114.) Thus, since the County’s African-American student population stood at thirty-four percent in 2007, each school should endeavor to achieve that proportion of African-American teachers. The district court ruled that this approach was consistent with ensuring that African-American students were not “isolated” in their educational experience as a result of too few African-American teachers in their classrooms. (JA 114.) The court stated that “if a black child is allowed to attend a previously all-white school, but is denied educational guidance which includes teachers of the student’s own race, the student is unacceptably isolated and deprived of a full educational experience.” (JA 114.) The Court concluded that a concern about the possible “isolation” of African-American students was an “unarticulated principle that ... animates the Supreme Court’s emphasis on the importance of a diverse faculty.” (JA 114.) Thus, the district court ordered the Board to work toward a faculty racial balance that would mirror the proportion of African-American students in the County as a whole, within a margin of fifteen percentage points. Just as with the student-assignment goal, the court treated this faculty goal as “flexible” and “a starting point in analyzing the Board’s success.... ” (JA 117.)

3. Extracurricular Activities

The district court also declined to grant unitary status with respect to the Board’s sponsorship of student extracurricular activities, but the district court did not separately explain this decision. 2

Following the January hearing, the district court asked the Board to submit additional statistical data relating to several different aspects of the County schools. One of the things the Court requested was “demographic data on students participating in all extracurricular activities.” (JA 479.) The Board admitted that it has not typically maintained records tracking this type of information. Owing to a similar request by the Tennessee Board of Education, however, the Board compiled what information it could, which turned out to be a chart showing the percentage of African-American and white students in thirty out of the County’s forty-eight schools that participate in extracurricular activities. (JA 486.) Thus, data for eighteen schools was completely omitted. Moreover, the *667 chart provides only the students’ participation rate by race; it does not explain what the Board or individual schools deem to be extracurricular activities, or African-American and white students’ participation rates within specific activities.

4. Other Aspects of the District Court’s Order

The district court acknowledged that its management of the case had not been optimally tailored to helping the Board achieve unitary status. The court said that its “failure to adopt clear and unequivocal guidelines for achievement of the [c]ourt’s goals is in large part responsible for the fact that the County is seeking unitary status some forty-four years after this suit was first filed____” (JA 116.) Pointing to the fact that the Plaintiffs and the County had been in “lock step” with one another for much of the last thirty years and that there had been “few, if any, contested issues, and only nominal litigation,” the district court acknowledged the need for it to take a more hands-on approach to supervising the litigation with an eye to bringing it to a conclusion in the near future. The court stated that it had “largely served to ‘rubber-stamp’ the County’s unopposed construction and zoning requests with little or no meaningful review of how such proposals contributed to or detracted from the County’s overall progress toward unitary status.... The joint motion to dismiss compels the [cjourt to resume a more substantive role in bringing the County school system’s desegregation process to a legitimate closure.” (JA 115-16.)

Besides the “flexible” benchmarks the district court set for evaluating the Board’s student and faculty desegregation efforts, the district court also set timetables for achieving them. The court set October 2012 as the target date for full compliance. (JA 119.) After three years of frill compliance, the court said it would dissolve the desegregation decree. (JA 119.) \

Next, the district court ordered the Board to submit data annually concerning the racial composition of each school’s students and teachers. The Board was further instructed to provide information about “mitigating factors, including infeasibility of further desegregation and shifting demographics, as appropriate.” (JA 118.) The Board was to continue submitting notices of school construction plans and attendance-zone modifications to the court until the desegregation decree was dissolved. Contrary to the perfunctory nature of similar reports filed in the past, the court clarified that all such plans are to include’a comprehensive discussion of what impact the construction and attendance-zone changes will have on the Board’s desegregation efforts. (JA 119.)

Finally, pursuant to Federal Rule of Civil Procedure 58, the district court said it would appoint a special master to consider the statistical data supplied by the parties and to make annual reports and recommendations to the court. (JA 117-18.) The court instructed the parties jointly to select a special master with the requisite qualifications, which the court defined as a “neutral expert in educational research, preferably with experience in desegregation issues....” (JA 118.)

II. ANALYSIS

The record shows that the district court was both thorough and careful in considering the petition to declare the Shelby County Schools unitary. As a threshold matter, the court held two public hearings, both of which involved witness testimony, argument by counsel, and the opportunity for participation by members of the community. All parties were allowed to call *668 and cross-examine witnesses, and the court asked questions as well. Following each hearing, the court asked the parties to provide specific additional information. Just a few days after the second hearing, the court issued its opinion which, at sixty-two pages, gave comprehensive consideration to the entire history of the case, identified and discussed the relevant Supreme Court authority, and set forth a precise plan designed to eventually eliminate the need for the desegregation decree and restore local control. The court also appended to its opinion a 116-page procedural history of the case.

Before specifically addressing each of the three areas in which the district court declined to grant unitary status, it is important to clarify what the district court’s order did and did not do. Contrary to the Board’s argument (repeated throughout its brief), the district court did not order the Board to comply with a “strict racial quota.” True, the district court held that the Board should endeavor to achieve a ratio of white to African-American students in each school that mirrors the racial composition of the County-wide student population within fifteen percentage points, and that the Board should seek to achieve the same racial balance with respect to faculty. But the district court plainly characterized this as a “flexible goal” and “a starting point in analyzing the Board’s success in desegregation.” Correctly citing applicable Supreme Court precedent, the district court expressly rejected a rigid approach, saying that “racial balance is not a strict requirement and there may be mitigating circumstances which would allow a school system to operate one or more schools with a predominance of one race or another without running afoul of its equal protection obligations under the Constitution.” The district court also specifically instructed the Board to submit annually information describing “mitigating factors” that make further progress toward racial balance infeasible.

The Board appeals the district court’s order in its entirety. Although the Government supported a declaration of unitary status below, it does not appeal the district court’s rulings retaining supervisory control. The Government says that it has “chosen not to appeal the denial of unitary status in light of the deferential clear error standard of review that applies to that determination.” (Govt.Br.29.) With respect to the district court’s modification of the decree, establishing a “flexible” student-assignment goal, the Government concedes that it is “hard pressed to argue on appeal that the district court’s ruling to the contrary was clear error, or that the student assignment goals that the court put in place to remedy the remaining racial imbalance in Shelby County’s schools are an abuse of discretion.” (Id.) The Government, however, is appealing the district court’s teacher-assignment goal on the ground that the court’s tying of the faculty racial ratio to the student racial ratio was an abuse of discretion.

1. The County Has Not Carried Its Burden as to Student Desegregation 3

The Board argues that it has fully complied with the court’s orders regarding *669 desegregating its students. The record evidence is simply inadequate to support this assertion.

As an initial matter, there is no indication that the Board regularly reported to the Court about the racial make-up of each of its schools relative to the County as a whole. Statistical data of this sort is largely absent from the record. Consequently, we cannot know how the Board fared with respect to the court’s original student-desegregation goal (identical racial ratios in the schools and County within a deviation of ten percentage points). As to those years for which there is more comprehensive data, the district court found that the Board’s results were unimpressive. For instance, the court found that in 1974, at least a quarter of the County’s schools had not attained “the target racial composition range established by the [c]ourt in 1968 and reiterated in ... 1971____” (JA 106-07.) The court went on to note that “[e]uriously missing from the Board’s submissions was any explanation of why one school had only 2% African-American students while a handful of others had as much as 71%.” (JA 107.) The district court found that by 1984, “[t]he racial balance in the schools continued to be uneven, with three schools out of 34 having fewer than 5% African-American students and three having between 50% and 72%.” (JA 107-08.) Finally, the district court noted that in the 2004-05 school year, only seventeen of the County’s forty-six schools “had a racial makeup that was reflective, within ten percentage points, of the 32% African-American student composition of the district as a whole.” (JA 108.)

It is also difficult to evaluate the import of what little statistical data is available. It is impossible to assess whether that data reflects all the desegregation that was feasible in Shelby County, or whether that data only reflects the desegregation that the County found comfortable. This is the case because the record is virtually silent about what obstacles the Board confronted in meeting its desegregation obligations throughout the 1970s and 1980s, how it handled those obstacles, and whether it pursued all practicable forms of relief or limited itself to only the most convenient ones.

The Board relies heavily on a chart detailing statistical data about the proportion of African-American students in each of its schools during the 2002-03- and 2007-08 school years. The Board claims that this evidence shows that County schools have become more diverse over the last five years. The Board’s data shows, however, that among the forty-four schools for which it has data (five schools were new), the African-American student population in 2007-08 still varied widely, from a low of three percent to a high of ninety percent. If the County-wide African-American student population was thirty-four percent in 2007, as the Board claims, then only fifteen of the forty-four schools for which the Board has demographic data would satisfy the district court’s benchmark for measuring racial balance, namely plus or minus fifteen percentage points from thirty-four percent (racially balanced schools using this metric would be those with an African-American student population of between nineteen and forty-nine percent). 4 Just ten of the County’s forty-four schools *670 would satisfy the district court’s earlier goal of a school racial composition that mirrored the County’s within a margin of ten percentage points.

Finally, the Board insists that any current disparities in the racial composition of its schools are caused by demographic factors unrelated to its history of segregation and over which the Board has no control, and the district court has no authority to correct. The Board also argues that annexations by the City of Memphis have affected, and will continue to affect, the African-American population within the County, making it a “moving target.” (Board Br. 9.) This argument may very well have some explanatory force, but the Board did not put on sufficient evidence to substantiate it. As noted above, the witness testimony at the two hearings consisted of little more than bare assertions that the Board had done all it could, that discrimination was not tolerated, and that parents and administrators were proud of the school system’s accomplishments. The Board did not present expert testimony about the shifting racial demographics within the County and how those changes affected the Board’s desegregation efforts. Nor did it present expert testimony about the impact of the annexations by the City of Memphis. The Board’s lack of evidentiary support on these issues stands in marked contrast to the evidence others in the Board’s shoes have presented in petitions to dissolve a desegregation decree. See, e.g., Freeman, 503 U.S. at 471, 112 S.Ct. 1430 (noting presentation of expert testimony about demographic changes in the school system since the desegregation decree was entered); Manning, 244 F.3d at 936 (noting presentation of “reports on attendance boundaries, demographic reports, and expert testimony” to address question of whether the racial identifiability of seventeen of the school system’s 150 schools was traceable to the system’s prior discriminatory behavior or to something else). Even the Government agrees that most of the evidence presented at the July hearing (in which the Government participated) was anecdotal. The Government notes that, “[t]he Board presented no minutes or other records documenting its decision-making and provided no demographic statistics or census maps beyond enrollment figures and charted attendance zones.” (Govt. Br. at 21.)

For these reasons, I would affirm the district court’s judgment denying unitary status on the issue of student desegregation.

2. The County Has Not Carried Its Burden As To Faculty Desegregation

The Board contends that the district court erred in failing to grant unitary status pertaining to faculty integration. The Government contends that the district court erred in correlating the percentage of African-American teachers in each school with the percentage of African-American students in the County overall. Importantly, neither party challenges the district court’s factual findings that the County did not fully comply with the court’s directives to (1) reassign faculty between schools, (2) transfer opposite-race teachers to fill vacancies, and (3) file the required notices and reports with the court. The Government even concedes that the racial imbalance among the County’s teachers “create[s] a presumption against unitary status for faculty” and that “[ajssignment of teachers according to flexible guidelines would not be an abuse of discretion.” The ■ Government’s position — opposing the district court’s particular remedy but not disagreeing that some type of remedy may be appropriate — implies that it does not oppose a remand for *671 the district court to re-formulate its faculty-desegregation remedy.

As described above, the district court modified the desegregation decree to tether the racial composition of each school’s faculty to the racial composition of the County’s students. The Government opposes this exercise of the district court’s remedial authority as an abuse of discretion, arguing that the proper comparison is the one the district court previously endorsed but has since abandoned: comparing the proportion of African-American teachers in each school to the proportion of African-American teachers County-wide.

The Supreme Court has never ruled on whether faculty desegregation may be accomplished as the district court has suggested — through a flexible comparison of the proportion of African-American teachers in each school to the proportion of African-American students in the district as a whole. The Government relies on Wygant v. Jackson Board of Education, 476 U.S. 267 , 106 S.Ct. 1842 , 90 L.Ed.2d 260 (1986) for the proposition that the Supreme Court would disapprove of such a comparison. In Wygant , white teachers with more seniority were laid off ahead of African-American teachers, some of whom were merely on probationary status. Id. at 272 , 106 S.Ct. 1842 . The school board reasoned that this policy was justified' in order to preserve the gains made in African-American hiring. Id. at 288 , 106 S.Ct. 1842 . The district court concluded that the layoffs did not violate the Equal Protection Clause, in part, because it determined that African-American students were entitled to role models of their own race. Id. at 274 , 106 S.Ct. 1842 . The Supreme Court rejected the district court’s embrace of “the role model theory,” concluding that “by tying the required percentage of minority teachers to the percentage of minority students, it requires just the sort of year-to-year calibration the Court stated was unnecessary in Swann.” Id. at 276 , 106 S.Ct. 1842 . Moreover, the Supreme Court reasoned that “[cjarried to its logical extreme, the idea that black students are better off with black teachers could lead to the very system the Court rejected in Brown ....” Id. The Court was apparently concerned that if the African-American student population shrinks over time, that fact could be used to justify hiring fewer African-American teachers. See id. at 276 , 106 S.Ct. 1842 . Conversely, if the African-American student population increases over time, that would require hiring more African-American teachers, which could increase the racial identifiability of the school. See id. Wygant, however, is distinguishable from this case because the policy imposed resulted in layoffs, as opposed to faculty integration through transfers and reassignments.

As the Government indicates, this Court has held that a district court overseeing a desegregation decree does not have the power to set a racial quota for the hiring of African-American teachers. In Oliver v. Kalamazoo Education Association, 706 F.2d 757 (6th Cir.1983), we stated that “students ... do not have a constitutional right to attend a school with a teaching staff of any particular racial composition .... Rather, ... all that the students are entitled to is the sustained good faith effort to recruit minority faculty members so as to remedy the effects of any past discriminatory practices.” Id. at 762 (internal quotation marks and citation omitted). Oliver , however, is distinguishable from this case because a rigid racial quota had been imposed for hiring teaching staff, as opposed to a flexible goal, which is what was imposed in this case.

Wygant and Oliver are therefore not on all fours with this case. Moreover, there *672 is no reason to fear that the district court’s formulation will tend to increase the racial identifiability of schools, or require a year-to-year re-calibration, as the Government contends. The district court tied its faculty hiring goal to the African-American student population in the County as a whole, not the African-American student population in each school. Thus, a school with a student body that is fifty-five percent African-American need not aim for a faculty that is fifty-five percent African-American. Rather, using the 2007 County-wide statistic of thirty-four percent African-American students, such a school would endeavor to attain a faculty that is thirty-four percent African-American, plus or minus fifteen percentage points.

For these reasons, I would hold that the district court did not abuse its discretion in tying the percentage of African-American teachers in each school to the percentage of African-American students in the County as a whole.

3. The Board Has Not Carried Its Burden With Respect to Extracurricular Activities

No more need be said about this. There is no basis for reversing as clearly erroneous the district court’s denial of unitary status as to extracurricular activities in light of the almost non-existent record on this subject.

III. CONCLUSION

For the reasons articulated above, I respectfully dissent.

. The Board does not dispute this finding.

. The Board presented minimal evidence on this subject, which may explain the absence of the district court’s explanation on its decision.

. In the specific setting of a school desegregation class action, ''[w]here the relief sought in the district court is the dissolution of a[] [desegregation decree], the order of the district court is subject to a mixed standard of review.” Manning ex rel. Manning v. School Bd. Of Hillsborough County, 244 F.3d 927, 940 (11th Cir.2001). The district court’s application of the law is subject to de novo review, while the court's factual findings, including its determination that a school district has.not achieved unitary status, fall under the *669 clearly erroneous standard of Federal Rule of Civil Procedure 52(a). Id. (citations omitted); Holton v. City of Thomasville School Dist., 425 F.3d 1325, 1336 (11th Cir.2005) (citations omitted).

. The Government agrees. It says that "[i]n 2007, about 60% of the schools varied by more than 15% from the system-wide racial ratio, with some as high as 90% black.” (Govt.Br.37.)