NAACP, JACKSONVILLE BRANCH, Plaintiff-Appellant, v. DUVAL COUNTY SCHOOL, Defendant-Appellee, 273 F.3d 960

HILL, Circuit Judge: Forty-one years ago, this litigation began. The original complaint sought the desegregation of the Duval County, Florida school system. Five district court judges have presided over the case since its inception, and, four times, two different circuit courts of appeals have been asked to review one of their decisions. In this fifth appeal, we must decide whether the present district court correctly determined that this litigation should come to an end because the school system has achieved unitary status. We agree with the district court that the answer is “yes.” I.

The original complaint in this case was filed on December 6, 1960. See Braxton v. Bd. of Pub. Inst. of Duval County. It sought an injunctive order desegregating the Duval County public schools. In 1963, after a finding that the county was operating a de jure dual school system, in which black and white children were required to attend separate schools, the district court began supervising the desegregation of the county’s schools. In 1971, after the Supreme Court held that mandatory busing of students to eliminate racial disparities was a permissible, and sometimes necessary, desegregation remedy, Swann v.

Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 22, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971), the district court ordered the Duval County School Board (the “Board”) to implement the mandatory reassignment of students. Mims v. Duval County Sch. Bd., 329 F.Supp. 123 (M.D.Fla.1971). Elementary and junior high schools were clustered into groups which were converted to grade centers to produce student bodies with 21% to 34% black students. Id. at 134-35. Students were bussed within each cluster to achieve the mandated racial balance. Id. at 130-31.

The judgment was affirmed on appeal, 447 F.2d 1330 (5th Cir.1971) and, as amended from time to time, primarily to accommodate the opening of new schools, the decree governed the parties for the remainder of the 1970’s and well into the 1980’s. During the pendency of the Mims injunction over the next 19 years, every school in the entire school district, which previously had been all-black or all-white, had substantial numbers of students of the other race in attendance for one or more years. By 1989, only 18 of the 142 schools operating in Duval County were identifiably black.

By 1990, however, both plaintiffs and the Board had come to the conclusion that mandatory busing was particularly onerous to the students of Duval County and not in their best interests. After almost thirty years of litigation, the parties entered into settlement negotiations. They agreed that mass busing should come to an end. They also agreed that future de-segregative efforts by the Board should focus almost entirely upon the elementary schools.

Finally, they agreed that they wished to measure the Board’s success in achieving unitary status by the goals outlined in the “Stipulation and Agreement,” filed on June 18,1990. After some technical corrections and refinements, the “Corrected Stipulation and Agreement” (the “CSA”) was established as the defining document outlining the goals that the Du-val County School Board (the “Board”) must achieve in order to attain unitary status. The CSA, thus, represents a roadmap to the end of judicial supervision of the Duval County school system.

Its 33 paragraphs contain a series of steps which the Board agreed to undertake to achieve unitary status for the school district. Most of these steps are addressed to attaining greater racial balance in student enrollment in the county’s schools, especially its elementary schools. “Attachment C” to the CSA lists 28 elementary schools that the parties expected to become identifiably black with the end of the Mims injunction’s mandatory student assignment and busing. Prior to the injunction, thirteen of these schools had been all-white; fifteen had been all-black.

During the years of mandatory student assignment, all the Attachment C schools were attended by students of both races. After mandatory student assignment was replaced with the CSA’s “local attendance zones,” these schools became identifiably black because their attendance zones had become 96% black. The parties, therefore, directed much of the CSA to improving the racial balance at these schools. The CSA established “a desegregative goal of at least 20% black students and 45% white students” at these Attachment C elementary schools.

The Board was required, with community input, to implement and aggressively promote magnet programs as incentives to attract white students to these schools. In addition, the CSA required the Board to permit majority to minority transfers in order to improve the racial balance at these schools.

The Board was also required to commit $60,000,000 for the “renovation, substantial rehabilitation, or replacement of core city schools.” With respect to middle and high schools, the C.SA specifically designated three middle and four high schools, which, based on their attendance areas, were expected to remain or become racially identifiable and directed that they operate magnet programs to attract other-race students. The CSA also set goals with respect to achieving racial equality in faculty and staff hiring and placement; transportation; extracurricular activities; and facilities and capital expenditures.

On July 14, 1990, the district court approved and adopted the CSA. It retained jurisdiction to monitor its implementation and to enforce it. The CSA was implemented during the 1991-92 school year. In 1996, the Board moved the district court to declare that the Duval County school district had met the constitutional requirements for unitary status as set forth by the Supreme Court in Bd. of Educ. of Oklahoma City Pub. Sch. v. Dowell, 498 U.S. 237, 111 S.Ct. 630, 112 L.Ed.2d 715 (1991), Freeman v. Pitts, 503 U.S. 467, 112 S.Ct. 1430, 118 L.Ed.2d 108(1992), and Missouri v.

Jenkins, 515 U.S. 70, 115 S.Ct. 2038, 132 L.Ed.2d 63 (1995), and also that it had fulfilled its contractual obligations under the CSA. After a three-week evidentiary hearing in 1997, and over a year for the preparation of briefs, the case was orally argued on August 24, 1998.

In a 140 page opinion that exhaustively catalogued every paragraph of the CSA and carefully evaluated the record evidence of the Board’s efforts and success in achieving each paragraph’s stated goal, the district court found that the Board had acted in exceptional good faith in its efforts to comply with the CSA and that it had substantially achieved its goals. The court also held that the Board had fulfilled its constitutional obligation to eliminate the vestiges of de jure segregation and to desegregate Duval County’s schools in good faith and to the extent practicable.

Accordingly, the district court declared that the school district was unitary in all respects, vacated all prior injunctions, and dismissed the case. This appeal followed. We review the district court’s conclusion that Duval County has achieved unitary status for clear error. Manning v. Sch. Bd. of Hillsborough County, 244 F.3d 927, 940 (11th Cir.2001) cert. denied, - U.S. -, 122 S.Ct. 61, 151 L.Ed.2d 28 (2001) (unitary status is a finding of fact); Lockett v. Bd. of Educ. of Muscogee County, 111 F.3d 839, 841-42 (11th Cir.1997) (Lockett II).

Under this standard of review, a district court’s findings are entitled to substantial deference. “Where there are two permissible views of the evidence, the [district courtj’s choice between them cannot be clearly erroneous.” Manning, 244 F.3d at 940 (quoting Lockett II, 111 F.3d at 842) (internal quotation and citation omitted). Moreover, only in circumstances when “a district court applies an incorrect legal standard which taints or infects its findings of facts, [do] such findings lose the insulation of Rule 52(

a) and judgment based thereon cannot stand.” Id. at 940-41 (internal quotations and citations omitted). We review the district court’s

interpretation and application of the law de novo. Id. II. School boards that formerly operated a dual school system, in which black students attended one set of schools and white students another, have been “clearly charged with the affirmative duty to take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch.” Green v. County Sch. Bd. of New Kent County, 391 U.S. 430, 437-38, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968) (emphasis added).

A unitary system is one in which there is presently no de jure racial segregation, and the vestiges of former de jure segregation have been eliminated to the extent practicable. Freeman, 503 U.S. at 494, 112 S.Ct. 1430 (“If the unlawful de jure policy of a school system has been the cause of the racial imbalance in student attendance, that condition must be remedied”). Until these goals are achieved, the Supreme Court has ordered district courts to supervise the desegregative efforts of school boards that formerly practiced de jure segregation. Lockett II, 111 F.3d at 842 (citing Brown v.

Bd. of Educ., 349 U.S. 294, 301, 75 S.Ct. 753, 99 L.Ed. 1083 (1955)). To be entitled to the end of federal court supervision, a formerly dual school system must be able to prove that it has (1) complied in good faith with the desegregation decree, and (2) eliminated the vestiges of prior de jure segregation to the extent practicable. Jenkins, 515 U.S. at 88, 115 S.Ct. 2038 (quoting Freeman, 503 U.S. at 492, 112 S.Ct. 1430). See also Manning, 244 F.3d at 942; Lockett II, 111 F.3d at 843.

In this case, the parties and the district court have agreed that the CSA defines the specific goals the Board must meet and the steps it must take in order to attain unitary status. “For enforcement purposes, consent agreements are interpreted under the principles of contract law.... [T]he obligations required of each party to a consent decree must be found ‘within its four corners, and not by reference to what might satisfy the purposes of one of the parties to it.’ ” Jacksonville Branch, NAACP v. Duval County Sch. Bd., 978 F.2d 1574, 1578 (11th Cir.1992) (citations omitted).

The Board, of course, must also meet its constitutional obligation to eliminate the vestiges of de jure segregation to the extent practicable. In evaluating the Board’s fulfillment of this obligation, the district court must examine six areas: (1) student assignments; (2) facilities; (3) faculty; (4) staff; (5) transportation; and (6) extracurricular activities. Green, 391 U.S. at 435, 88 S.Ct. 1689. Since the Board operated de jure segregated schools in the past, there is a presumption that any current racial disparities in these areas are the result of its past unlawful conduct. Keyes v. School Dist.

No. 1, 413 U.S. 189, 208-09, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973). The Board “bears the burden of showing that any current imbalance [in these areas] is not traceable, in a proximate way, to the prior violation.” Freeman, 503 U.S. at 493, 112 S.Ct. 1430. See also Lockett II, 111 F.3d at 843 (district court must presume that substantially disproportionate racial compositions within the schools is constitutionally violative) (citing Swann, 402 U.S. at 25, 91 S.Ct. 1267). This presumption may, however, be overcome.

The Supreme Court has recognized that some racial imbalances in our schools today are caused by external forces, such as demographic shifts, which are not the result of segregation and which are beyond a school board’s control. See Jenkins, 515 U.S. at 102, 115 S.Ct. 2038; Pasadena City Bd. of Ed. v. Spangler, 427 U.S. 424, 434, 96 S.Ct. 2697, 49 L.Ed.2d 599 (1976); Swann, 402 U.S. at 22, 91 S.Ct. 1267.

If a school board can prove that such factors have substantially caused current racial imbalances in its schools, it overcomes the presumption that segregative intent is the cause, and there is no constitutional violation. Id. See also Manning, 244 F.3d at 944; Lockett II, 111 F.3d at 843. Where there is no constitutional violation, a school board is under no duty to remedy racial imbalances. Id. The Supreme Court has made clear that: Racial balance is not to be achieved for its own sake. It is to be pursued when racial imbalance has been caused by a constitutional violation.

Once the racial imbalance due to the de jure violation has been remedied, the school district is under no duty to remedy imbalance that is caused by demographic factors. Freeman, 503 U.S. at 494, 112 S.Ct. 1430. See also Manning, 244 F.3d at 941.

Furthermore, even when remedying the effects of de jure segregation, the Constitution does not require rigid racial ratios. “The purpose of federal supervision is not to maintain a desired racial mix at a school.” Manning, 244 F.3d at 941 (citing Pasadena, 427 U.S. at 434-37, 96 S.Ct. 2697 and Freeman, 503 U.S. at 494, 112 S.Ct. 1430). “The constitutional command to desegregate schools does not mean that every school in every community must always reflect the racial composition of the school system as a whole.” Swann, 402 U.S. at 23-24, 91 S.Ct. 1267. Such a goal would require federal supervision in perpetuity.

On the contrary, a complete return to local control of school systems is the ultimate goal of all judicial supervision because “[fjrom the very first, federal supervision of local school systems was intended as a temporary measure to remedy past discrimination,” and desegregation decrees “are not intended to operate in perpetuity.” Dowell, 498 U.S. at 247-48, 111 S.Ct. 630. See also Manning, 244 F.3d at 942 n. 25. We turn now to our review of the district court’s determination that the time has come to return control over its school system back to Duval County. III. A.

The Green factors: Faculty and staff hiring and 'placement; transportation; extracurricular activities; facilities and capital expenditures. It is undisputed that the Board aggressively recruited black faculty and staff and has achieved the CSA’s goal of a 70% to 30% faculty racial composition system-wide. Nor is it disputed that the Board provided transportation in a non-discriminatory fashion or that it transported students to magnet programs far from their homes at no cost to them.

The evidence also supports the district court’s finding that the Board made all extra-curricular activities available to all students throughout the county without regard to race. Finally, the district court found that it is “beyond dispute” that the Board has substantially complied with all of the requirements contained in the CSA with respect to facilities improvements and repairs, spending over $60,000,000 to modify and improve existing schools, most of which were located in inner-city areas.

Thus, we conclude that there is no clear error in the district court’s finding that the Board has achieved unitary status with respect to five of the six Green factors. B. Student Assignment With respect to student assignment, it is undisputed that the Board implemented a voluntary school choice plan, terminated mass busing, and designated 71 schools as magnet schools, all as the CSA required. The district court also found that the Board allocated appropriate funding for the operation and improvement of its magnet programs.

It operated a separate office with staff dedicated to achieving the CSA goals, including magnet program development and recruitment. The Board has also engaged in substantial and continuing efforts to recruit white students to its racially identifiable black elementary schools, each of which has a magnet program. The district conducts a “Magnet Mania” recruitment fair, distributes a cata-logue to parents, maintains a telephone hotline, and directly recruits parents to encourage magnet enrollment of their children.

Furthermore, while not contractually obligated to, the Board did use several other methods to increase racial balance at certain schools, including capping enrollment, majority to minority transfer programs, attendance boundaries and limits on curriculum offerings at certain schools. The evidence at trial was that the Board’s efforts have substantially achieved the CSA’s student enrollment goals. In 1991-92, the first year under the CSA, 65 of the 139 elementary, middle and high schools met its goal of a student population that was more than 20% but less than 55% black.

By the 1996-97 school year (the year application was made for unitary status), 88 of 138 schools, or 64% of the district’s schools, were in compliance with the CSA’s student enrollment goals. One year later, in 1997-98, the number of schools in compliance rose again, to 91 out of 140, or 65%. At the elementary school level, the compliance rate went from 42% (42 of 100) at the inception of the CSA (1991-92 school year), to 60% (60 of 100) in 1997-98. Middle schools went from 59% compliance in 1991-92 (13 of 22), to 77% compliance in 1997-98 (17 of 22).

High schools were 59% compliant in 1991-92 (10 of 17) and 76% (13 of 17) compliant in 1997-98. Despite this overall substantial compliance with the CSA’s enrollment goals, there are 26 identifiably black schools in Duval County. The following table illustrates this fact. Two things are clear from this table. First, the Mims injunction was successful in eliminating the vestiges of de jure discrimination. Under its mandatory student assignment orders, all students were assigned to schools based upon racial balance goals and then large numbers were bussed to achieve those goals.

During this time, the number of racially identifiable schools in the system was significantly reduced. As the table shows, from 1971 to 1991, the number of schools with 75% or more black students went from 21% to 15%. As we observed earlier, the evidence at trial was that all the racially identifiable Attachment C elementary schools in the core city were attended by students of both races during those years. All but two of the fifteen schools formerly operated solely for black students reached what later became the CSA racial balance goals for at least a three-year period before the implementation of the CSA.

The Board had broken the pattern of all-black enrollment at the schools it formerly operated solely for black students during de jure segregation, and had successfully eliminated the vestiges of de jure segregation in all the Attachment C schools. Then, immediately upon the establishment of the CSA’s local attendance zones, the number of racially identifiable schools, as the parties had expected, rebounded to 21%.

Although the percentage of racially identifiable schools has now declined to 18%, the Board acknowledges that under the CSA, the school district has experienced an increase in the total number of racially identifiable schools. It denies, however, that this is the result of any resegregative policies, either intended or unintended. The Board points out that 19 of the 26 racially identifiable schools in operation today are the Attachment C elementary schools located in the core city.

Testimony presented at trial persuasively demonstrated that their racial identity has been substantially caused by the decline in the number of white students in these schools. The table graphically illustrates this decline. With the advent in 1971 of mass busing, substantial numbers of white students began to leave the public school system. In 1969, the Duval County public school student population was 72% white. By 1998, only 59% of that population was white. As a result, the percentage of black students went from 28% to 40%.

The Board’s racial enrollment statistics demonstrate that much of the decline in the white student population occurred in the schools of the core city, ultimately leaving it with a school-age population that is over 96% black. A former school district superintendent testified that even the 13 formerly de jure all-white elementary schools in the core city became identifiably black schools with the end of mass busing. Additionally, the de jure white high school in the core city also became identifiably black.

Furthermore, the Board presented evidence that during this same period of time black student mobility has sometimes caused formerly racially balanced schools to become racially identifiable black schools. For example, the residential areas surrounding two naturally integrated elementary schools experienced an increase in black students from approximately 20-30% in the 1989-90 school year to approximately 40-60% in 1996-97. The residential areas surrounding two additional elementary schools went from 20-30% in the 1989-90 school year to 30-40% in the 1996-97 school year.

These dramatic increases in the number of black school-age children residing in the areas surrounding these naturally integrated schools persuaded the district court that “the loss of these naturally integrated schools was the result of housing demographics, and not any intentional or unintentional policy operated by the Board.” Finally, the Board’s expert demographer’s evidence was that even though the black student population is declining in the core city (6% decline), as black families move to the outlying areas of the city (26% gain in the areas surrounding the inner city, and 41% gain in the outermost areas), more black than white students move into the core city, from both outside and outlying areas of the county.

The district court concluded that: In response to the School Board’s direct evidence and expert testimony attributing present racial imbalances in some schools to demographic factors beyond the Board’s control, the NAACP has failed to present any persuasive evidence that such racial Imbalance existing at any elementary, middle, and senior high school is the result of anything other than racial isolation in housing patterns and the continued existence of “white flight” in Duval County. The Board has simply not contributed to or perpetuated the current imbalance in any way.

To the contrary, the Board has made enormous efforts to counter the effects of past discrimination and present-day choices by parents. Since the NAACP was unable to demonstrate that the continued existence of identifiably black schools in Duval County is a vestige of prior de jure segregation rather than the result of these demographic or other non-discriminatory factors, the district court held that the county was entitled to unitary status on this Green factor as well. We agree.

As we have noted above, in such cases a school district is under no duty to remedy racial imbalances that are caused by demographic factors. Freeman, 503 U.S. at 494, 112 S.Ct. 1430. As the Supreme Court wrote in Swann: Neither school authorities nor district courts are constitutionally required to make year-by-year adjustments of the racial composition of student bodies once the affirmative duty to desegregate has been accomplished and racial discrimination through official action is eliminated from the system.

This does not mean that federal courts are without power to deal with future problems; but in the absence of a showing that either the school authorities or some other agency of the State has deliberately attempted to fix or alter demographic patterns to affect the racial composition of the schools, further intervention by a district court should not be necessary. 402 U.S. at 31-32, 91 S.Ct. 1267. In Freeman, the Supreme Court added: The effect of changing residential patterns on the racial composition of schools, though not always fortunate, is somewhat predictable.

Studies show a high correlation between residential segregation and school segregation.... Where desegregation is a product not of state action but of private choices, it does not have constitutional implications. It is beyond the authority and beyond the practical ability of the federal courts to try to counteract these kinds of continuous and massive demographic shifts. To attempt such results would require ongoing and never-ending supervision by the court of school districts simply because they were once de jure segregated.

Residential housing choices, and their attendant effects on the racial composition of schools, present an ever-changing pattern, one difficult to address through judicial remedies. 503 U.S. at 495, 112 S.Ct. 1430. We believe the district court correctly concluded that this is such a case. Although the Duval County school district as a whole is undeniably in substantial compliance with the goals of the CSA, white flight and voluntary residential patterns have re-segregated a number of the core city’s schools.

There is no evidence that any Board policy contributed to the reemergence of these schools or perpetuated their existence in any way. “While those charged with desegregation must not shrink from the threat of white flight, school officials who have taken effective action have no affirmative fourteenth-amendment duty to respond to those who vote with their feet.” Ross v. Houston Indep. Sch. Dist., 699 F.2d 218, 225 (5th Cir.1983). Nor does the existence of racially identifiable schools in Duval County necessarily trump the Board’s claim to be operating a unitary system.

Pasadena, 427 U.S. at 424, 96 S.Ct. 2697; Freeman, 503 U.S. at 493-94, 112 S.Ct. 1430; Jenkins, 515 U.S. at 116, 115 S.Ct. 2038 (Thomas, J., concurring). Id. In Freeman, for example, the Supreme Court affirmed a finding of unitary status in a school district in which 50% of the black students attended schools that were over 90% black in a system where blacks comprised only 47% of the student body. Furthermore, 5 of the 22 high schools were over 90% black, and another 5 were over 80% white. Finally, 18 of the 74 elementary schools were over 90% black and 10 were over 90% white. 503 U.S. at 476-77, 112 S.Ct. 1430.

In rejecting the argument that such racial imbalances precluded a finding of unitary status, the Court held that “the school district is under no duty to remedy imbalance that is caused by demographic factors.” 503 U.S. at 494, 112 S.Ct. 1430. We, too, have had occasion to consider the impact of racially identifiable schools on a school district’s request for unitary status. In Manning, the district court denied unitary status to the Hillsborough County school system in which approximately 90% of the system’s schools were racially balanced, but which also had 17 racially identifiable schools.

In reversing the district court, we held that the existence of these racially identifiable schools did not preclude a finding of unitary status since the county proved that external forces were a substantial cause of the racial imbalances. 244 F.3d at 944. Other courts have also ended federal supervision despite the existence of racially identifiable schools in a district. The First Circuit Court of Appeals upheld a finding of unitary status in a school district where 8 schools enrolled a student body that exceeded 80% black, and 5 schools enrolled a student body that was in excess of 80% white. Morgan v.

Nucci, 831 F.2d 313, 320-22 (1st Cir.1987). The court held that “maximum practicable desegregation” is a “practical, not theoretical standard,” and “[ujnitary status is not simply a mathematical construction.” Id. at 321. Instead of focusing on racial quotas, the court looked to the school district’s history of good faith in both the operation of the school system and the implementation of the court’s student assignment orders. Id.

Similarly, the Fifth Circuit affirmed unitary status for a district in which 55 out of 226 schools had student populations that were 90% or more black, in a district that was 20% white, 38% black, and 42% his-panic. Ross, 699 F.2d at 228. Several district courts have also declared school districts unitary despite the continued operation of numerous racially identifiable schools. See e.g., United States v. Unified Sch. Dist. No. 500, 974 F.Supp. 1367 (D.Kan.1997); Stell v. Bd. of Pub. Educ., 860 F.Supp. 1563, 1580 (S.D.Ga.1994); Flax v.

Potts, 725 F.Supp. 322, 328-29 (N.D.Tex.1989), aff'd and remanded on other grounds, 915 F.2d 155 (5th Cir.1990). Even an increase in the number of racially identifiable schools during the period of federal court supervision does not preclude a finding of unitary status. In Manning, we affirmed unitary status even though every one of the 17 racially unbalanced schools in Hillsborough County had become so during the pendency of the district court’s supervision. Id.

In Ross, the Fifth Circuit affirmed unitary status despite the fact that the number of racially identifiable schools in Houston had increased by 33 during the district court’s supervision. 699 F.2d at 227-28. “Constructing a unitary system,” the court said, “does not require a racial balance in all of the schools. What is required is that every reasonable effort be made to eradicate segregation and its insidious residue.” Id. The NAACP, however, insists that, unlike these other school districts, Duval County is not entitled to unitary status until it eradicates its racially identifiable schools.

It argues that under the CSA, the Board is obligated to achieve the “maximum practical desegregation.” So long as the district contains racially identifiable schools, it concludes, the Board is not in substantial compliance with its contractual duty. The Supreme Court has made quite clear, however, that the Constitution does not require a school board to eliminate the vestiges of past discrimination “to the maximum extent practicable.” Jenkins, 515 U.S. at 90, 115 S.Ct. 2038; Freeman, 503 U.S. at 492, 112 S.Ct. 1430; Dowell, 498 U.S. at 250, 111 S.Ct. 630; Manning, 244 F.3d at 942.

The Court expressly rejected this test in Jenkins, and held that the proper test was whether the deficiencies “attributable to prior de jure segregation had been remedied to the extent practicable.” 515 U.S. at 101, 115 S.Ct. 2038 (emphasis added). See also Manning, 244 F.3d at 943 n. 28. Under the correct constitutional standard, then, Duval County is entitled to unitary status if it has eliminated the vestiges of de jure segregation to the extent practicable. It is not responsible for the segregative effects of external forces over which ■ it has no control. Freeman, 503 U.S. at 494, 112 S.Ct. 1430.

Since the Board has demonstrated that white flight and voluntary residential isolation, and not its policies, substantially caused the racial identifiability of some of its schools, they are not the vestiges of de jure segregation and the Board is under no constitutional obligation to combat the demographic factors which produced them. Id. What then do we make of the CSA’s requirement that the Board “take steps to the maximum extent practicable” to achieve its racial enrollment goals?

Interpreting the contract in its constitutional context, the CSA requires the Board to implement policies that will desegregate Duval County’s schools to the maximum extent practicable within the parameters of the county’s particular characteristics. The district court correctly found that, within this context, the Board has performed its obligations to the maximum extent practicable. C. The Board’s Good Faith Performance Under the CSA The NAACP’s final claim on appeal is that the Board has not implemented the CSA in good faith. The Board has both a contractual and a constitutional obligation to do so.

Dowell, 498 U.S. at 249-50, 111 S.Ct. 630. To be entitled to unitary status, not only must a school system eliminate the vestiges of de jure segregation to the extent practicable, but “local authorities [must] have in good faith fully and satisfactorily complied with, and shown a commitment to, the desegregation plan.” Id. See also Manning, 244 F.3d at 942. Failure to do so can justify continued judicial supervision of a desegregation decree. Id.

The CSA was designed to achieve the maximum practicable desegregation within a system of voluntary choice, which would be guided and molded by the existence of attractive magnet programs at formerly racially identifiable schools. The district court found that the Board not only timely implemented its centerpiece — the magnet program — but that it generously funded and aggressively marketed it. The court also found that the Board has not only taken the other steps required of it by the CSA’s 33 paragraphs, but that it has done so in a timely and inclusive manner.

All the evidence indicates that the Board has consulted extensively with the NAACP in implementing the magnet programs, deciding on new student assignments and building new schools. During this time, the Board has never been found to be in violation of any provision of the CSA. The district court concluded that the Board has exhibited enormous good faith in performing its agreement to craft a school system which protects student and parent choice while vigorously encouraging a race neutral distribution of students throughout the system.

As with the Board’s compliance with the CSA, however, the NAACP views the good-faith glass as half-empty. It contends that the Board has faded to utilize “standard school desegregation techniques mentioned in the CSA” to eliminate its racially identifiable schools. Like the claim that the district is not in substantial compliance with the CSA, however, this argument is tainted by the NAACP’s adoption of the wrong legal standard. See Manning, 244 F.3d at 948-44.

The Constitution does not require a school board to remedy racial imbalances caused by external factors, such as demographic shifts, which are not the result of segregation and are beyond the board’s control. Freeman, 503 U.S. at 495, 112 S.Ct. 1430. Therefore, the Board is not obligated to employ a particular desegregative method because it might remedy these racial imbalances. Id. at 490, 112 S.Ct. 1430. The Board’s constitutional obligation is only to eliminate the effects of a constitutional violation.

The continued existence of a some racially identifiable schools within the district is not, by itself, proof of a constitutional violation, nor has the NAACP offered any. Without proof that the Board’s present or former segregative policies or practices are the cause of current racial imbalances, it is under no constitutional duty to employ desegregative techniques not required by the CSA to combat the demographic factors that are. Furthermore, we are without power to order it to do so. Milliken v.

Bradley, 433 U.S. 267, 282, 97 S.Ct. 2749, 53 L.Ed.2d 745 (1977) (“[F]ederal-court decrees must directly address and relate to the constitutional violation itself ... [and] exceed appropriate limits if they are aimed at eliminating a condition that does not violate the Constitution.”). Nor do we find any merit in the NAACP’s contention that the Board has “acted and omitted in classically segre-gative manners.” The crux of this complaint is that the Board should have built new schools in the so-called “grey areas” between the black core city and the white outlying suburbs.

This desegregative technique, it is argued, would have helped to achieve the maximum desegregation possible in the district. The evidence at trial was that the Board and the NAACP met on several occasions to discuss proposed sites for new schools. The former superintendent of the school system even proposed the use of the “grey area” sites for some new construction. After study, however, it was determined that each of the proposed grey sites had serious geographical deficiencies. Many were in highly commercialized areas, and several had serious pollution issues.

Ultimately, none of these sites was deemed suitable for new school construction. The Board then created a Desegregation Committee in 1995 to consider and recommend new school sites, but the NAACP refused to participate. The committee recommended the construction of 10 new schools. These recommendations were incorporated into a five-year facilities plan that was distributed to the NAACP later that year. The evidence was uncontradict-ed that all but one of these schools were built in areas for which the Board projects strong or moderate growth in student enrollment, including black students.

The Board’s expert testified that black student mobility to these areas of naturally integrated housing patterns will eventually increase the desegregation of the district’s schools. In view of the foregoing, we conclude that the NAACP’s mistaken view that the CSA and the Constitution require the Board to employ the maximum desegrega-tive techniques available has led it to conclude that the Board has acted in bad faith in its site selection for new schools.

Neither the CSA nor the Constitution require the Board to build new schools in the most integrated parts of the county regardless of whether these sites correspond to the areas of maximum population growth. The district court correctly found that the Board has fulfilled its constitutional and contractual obligation to act in good faith in the selection of new school sites in a race-neutral way. We agree. rv. In sum, we conclude that the judgment from which the appeal is taken is due to be affirmed. An affirmance implies that appellants have lost.

In a meaningful way, however, that implication is not justified here. This judgment means that appellants have accomplished what they, decades ago, set out to do. They challenged a rigidly maintained, de jure system of school segregation and sued to bring it into compliance with the constitutional requirement of equal protection under the law. We say today that they have succeeded. If this judgment is counted as a loss for appellants, it is so because they have won. Furthermore, none should read more into this judgment than it contains.

With its implementation, the Duval County school system may be out of the courthouse, but it is not out of the reach of the Constitution, the Bill of Rights, and the laws of this land. Nothing in this judgment authorizes conduct contrary to these laws. The Board, and the people of Duval County who, in the end, govern their school system, must be aware that the door through which they leave the courthouse is not locked behind them. They will undoubtedly find that this is so if they fail to maintain the unitary system we conclude exists today. The judgment of the district court is AFFIRMED. .

The appeals filed prior to 1981 were heard by the Court of Appeals for the Fifth Circuit. After the split in 1981, the appeals have been heard by this court. . For a variety of reasons, the case name has changed over time. . In 1986, the district court dissolved the Mims injunction because it found that the Duval County school system had achieved unitary status. On appeal, we found no fault with the district court’s finding of unitary status with respect to the areas of extra-curricular activities and transportation, but that further efforts were necessary in administrative staff, teacher and student assignments.

Jacksonville Branch, NAACP v. Duval County Sch. Bd., 883 F.2d 945 (11th Cir.1989). . By agreement of the parties, an identifiably black school has a student body over 75% black. . Duval County operates one of the largest school districts in the nation, consisting of 851 square miles, a geographic region repre- senting two-thirds the size of Rhode Island. Additionally, the county is separated by a large river and several smaller rivers causing some sections of the city to be isolated from other parts.

The transporting of students over the rivers and the sheer size of the county significantly increase the difficulty of busing in Duval County. . A majority of the middle and high schools were already in compliance with the student enrollment goals the CSA would establish and there were more racially identifiable elementary schools. . .

While none of these mutual goals was expressly stated in the Agreement, we agree with the district court’s finding that all are clearly implicit in it, and were also expressed explicitly in the Pretrial Statement filed by the parties prior to the hearing on unitary status, the arguments of counsel and the evidence presented at the hearing. . This appeal seeks a determination that the goals of the CSA have not been met. . The CSA is 25 pages in total length, including attachments, containing 33 paragraphs roughly divided as follows: (

a) paragraphs 1-10 incorporate the broad goals, purposes, and time frame informing the intent of the parties in the CSA implementation process; (

b) paragraphs 11-17 address requirements pertaining to Attachment C schools (identifiably black elementary schools); (

c) paragraphs 18-19 address requirements pertaining to Attachment D schools (identifiably white elementary schools); (

d) paragraph 20 addresses requirements pertaining to Attachment E schools (naturally integrated schools based on integrated housing patterns); (

e) paragraphs 21-22 contain additional requirements pertaining specific middle schools; (

f) paragraph 23 contains additional requirements pertaining to specific high schools; (

g) paragraph 24 addresses issues surrounding new school site selection and construction and requires the formation of a facilities committee consisting of representatives of both parties and members of the community; (

h) paragraph 25 requires the development of detailed plans to achieve non-discriminatoiy operational procedures; (

i) paragraphs 26-27 require the attainment of faculty and staff hiring goals as set forth by the Mims injunction; (

j) paragraph 28 requires the Board to take steps to increased black student and faculty representation in the gifted program; (

k) paragraph 29 requires a re-evaluation of black students classified as mentally retarded; (

l) paragraphs 30-33 establish the Board’s affirmative obligation to desegregate “to the maximum extent practicable” and that unitary status shall not be achieved until the Board maintains three years of racial equality in all areas of school operation. . A declining white student population during the Mims years caused the elementary schools that had formerly been operated solely for white students under state-mandated segregation to become identifiably black schools. See discussion in

section III. B. infra. . One school originally listed in Attachment C, is not included in this discussion because it is located at the beach in a zone that Attachment C was not supposed to include. . The NAACP’s expert witness testified that, in practice, this formulation means a black student enrollment of 20% to 55%. Some schools at the outer geographic areas of the county — the beaches and the far west — were given a goal of within +/-10% of the zone-wide racial composition at its grade organization level. .

The district court found that the CSA requires only the elementary schools listed in Attachment C to achieve these numerical goals because they applied only to schools in Zones I-V, which include only elementary schools. The NAACP argues that these goals apply to all schools in the district. The issue is really of no moment since both the district court and the Board have used these benchmarks to measure the racial enrollment effects of the middle and high school magnet programs. .

"Magnet schools, as generally understood, are public schools designed to promote integration by voluntarily drawing students away from their neighborhoods and private schools through distinctive curricula and high quality.” Missouri v. Jenkins, 495 U.S. 33, 40 n. 6, 110 S.Ct. 1651, 109 L.Ed.2d 31 (1990). Under the CSA, the Board was required to establish and maintain magnet programs at "racially identifiable" elementary schools and at several middle and high schools. .

The CSA required the Board to "intensively recruit” white and black students to attend opposite race schools and to provide transportation "at no cost” for these students to attend such schools. . Within a year, the NAACP filed a motion to modify the CSA. Among other requests, it urged the court to require the Board to implement a system of "controlled choice” as an additional desegregative technique. The district court denied the motion, Jacksonville Branch, NAACP v. Duval County Sch.

Bd., 978 F.2d 1574 (11th Cir.1992), and, on appeal, we found no error in the denial as to most of the issues raised, including the refusal to modify the CSA by ordering a "controlled choice” student enrollment program. On two issues, we remanded for a development of the record. . The parties have established a very detailed set of specific requirements in the CSA’s thirty-three paragraphs. The district court carefully analyzed the Board's achievement of each of these requirements. See

section III infra. . The NAACP contends only that the Board has not been in substantial compliance for at least three years as required by the CSA. The district court specifically found that the Board had been in compliance since the 1995-96 school year and the NAACP presented no evidence to the contrary. Therefore, this finding is not clearly erroneous. . Since participation in extra-curricular activities is voluntary, the law does not require, nor does the CSA establish, racial balance goals for this Green factor.

Eliminating racial distinctions is required, however, and the district court correctly found that the evidence is that this has been done. . The Board did cap white enrollment at three schools, one of which was identifiably white. Those students who were “capped out” were encouraged to attend magnet programs at one of the preferred option schools. The Board provided free transportation to those students who did so. . The Board has used a majority to minority transfer program to assist in desegregating naturally integrated, identifiably black, and identifiably white schools.

Free transportation has been provided to transferring students. The evidence was that requests for transfers within the elementary system are denied if they are counterproductive to the Board's desegregative efforts. In 1996, the Board mailed transfer letters to more than 36,000 parents, resulting in only 332 transfer requests. The Board contends that these low numbers demonstrate that students attending racially identifiable schools are doing do because it is their parents' preference. .

The Board changed the attendance boundaries for one high school in order to increase black enrollment and that school has been in compliance with the CSA goals since that time. No changes were made in the attendance boundaries of the racially identifiable elementary schools, however, because the evidence was that such changes would have made little difference in the racial composition of these schools since most of them are in the overwhelmingly black core city. .

The Board did not, however, limit the curricular offerings at any of the racially identifiable elementary schools because it believed it necessary to provide each student at every elementary school with a complete curriculum. In the Board’s view, limiting the curriculum at an elementary school, even in order to foster movement of students out of racially identifiable schools, was not consistent with the goals of the CSA or sound educational policy. Instead, the Board sought to encourage student movement by enhancing the curricular offerings at various racially identifiable schools through the magnet programs. .

Highly successful academic magnet programs have succeeded in transforming three of the racially identifiable middle schools and two of the racially identifiable high schools into schools with diverse student bodies. Two middle and two high schools in the core city, however, remain identifiably black. The magnet programs at these schools have been modified in a continuing effort to recruit minority students. .

These were the elementary schools identified in the Mims injunction as the "vestiges of de jure segregation.” The dissent incorrectly implies that these schools "remained” racially identifiable throughout the entire period of federal court supervision. The dissent also incorrectly maintains that the neighborhoods in which these schools are located are now, and have always been, over 90% black, when the record evidence is that these neighborhoods once supported 13 all-white elementary schools and only became over 90% black with the white flight that was precipitated by mass busing. .

In the final analysis, the dissent gives no weight at all to the "root and branch” destruction of de jure segregation which occurred with the massive busing under the Mims injunction, not because it didn’t occur soon enough, but because it occurred too soon to count! In rejecting all pre-CSA events as irrelevant to the issue of unitary status, the dissent ignores the first thirty years of federal supervision of the Duval County school system. During these years, the Board ended de jure segregation and instituted mass busing to eliminate the vestiges of that de jure segregation.

Then, having broken the link to the past, the Board began the final phase of its decades-long desegregative effort by entering into the CSA, a contract which expressly recognized the demographic shifts in the student population and which established an aggressive magnet program, as well as other steps, designed to ameliorate the resulting racial imbalances. The Board's efforts under the CSA have resulted in a 65% compliance rate with its racial enrollment goals.

Ignoring this context, and pretending that the Board’s success in eliminating the vestiges of de jure segregation does not count if it occurred before 1991, is the only premise upon which the dissent can build a case that "the Board did not eliminate the vestiges of segregation at its formerly de jure black schools during the CSA enforcement term.” (This may be the first time that a school district has been judicially chastised for achieving too much, too soon!) But this is not the law.

The Supreme Court has made clear that, once the vestiges of de jure segregation have been eliminated, as they were in Duval County under the Mims injunction, any re-emergence of racially identifiable schools resulting from demographic changes (over which the Board, of course, has no control) cannot be attributed to the Board’s former de jure segregative policies. Freeman, 503 U.S. at 494, 112 S.Ct. 1430. Nor does the Board have any constitutional obligation to combat these factors. Id.

On the issue of whether the Board has met its contractual obligation to ameliorate the impact of these demographic shifts, the district court's cata-logue of findings of fact attesting to the Board's good faith and often extraordinary efforts in achieving a 65% compliance rate with the CSA's racial enrollment goals overwhelmingly support our conclusion that it has. . It should be noted, however, that the overwhelming majority of the schools in the system remained in a desegregated status. . The most recent data is for the 1997-98 school year. .

With respect to all elementary schools, however, there has been a decline in racially identifiable schools. In the 1991-92 school year, 25 elementary schools had a black population in excess of 75%. By the 1997-98 school year, only these 19 were left. . The NAACP has not disputed the district court's findings of fact concerning the "white flight” that preceded the CSA, nor the statistical data underlying them. . Actually, the table exaggerates the white .student population because it puts all non-black students into the "white” category, including Asians, Hispanics, and Native Americans.

The student population characterized as "white” by the school board was down to 54% in 1997. . Under the Mims injunction, the de jure black high school in the core city was paired with its all-white counterpart, which was located virtually across the street. The district court revised the attendance zones for each, expecting to produce two totally integrated high schools with a white population of approximately 40% each. Within two years, however, both high schools were overwhelmingly black (90% + ). They remain so today. . Dr. Milan Mueller conducted a seven-year study of black student mobility in Duval coun-1y.

His study was designed to reveal the impact of such mobility on racial balance in school enrollments. . The parties agreed in the CSA that certain of the schools in the district were "naturally integrated” prior to implementation of the CSA due to residential patterns. . Current demographic trends indicate that the school system may become even more racially balanced as the black student population continues to shift out of the core city. In the interim, the success of the Board’s magnet school program can be expected to further reduce the number of racially identifiable schools. .

Indeed, as Justice Powell once noted, to expect a total absence of one-race schools would deny the demographic and economic realities of most major cities. See Estes v. Metropolitan Branches of the Dallas NAACP, 444 U.S. 437, 445-46, 100 S.Ct. 716, 62 L.Ed.2d 626 (1980) (Powell, J., dissenting from dismissal of writ of certiorari as improvidently granted). . We find no merit in any contention raised by the NAACP which we do not discuss. .

The CSA stated that the Board "may adopt other techniques to achieve desegregation in its schools, including, but not limited to: capping the enrollment; re-designing the student attendance zones; and limiting the curricular offering at certain magnet programs.” Although not mandatory, the Board has employed some of these techniques at some schools as discussed above. . The NAACP contends, and the dissent implies, that the Board should have implemented a “controlled choice” plan to achieve "maximum” racial balance in the district’s schools.

Under such a plan, groups of schools are clustered and each student is provided with a form with which to rank preferences for attendance within the cluster, in-eluding at least one school with a student body predominantly of the other race. When racial enrollment limits are reached at one school, the remaining students of that race are assigned and bussed to a school based on the racial composition of the school. The purpose of such a plan is the maximum possible desegregation of a school district without regard to the causes of the offending racial imbalances. This the Constitution does not require.

Furthermore, the NAACP concedes that implementation of such a plan in Duval County would require a return to substantial busing of students to achieve the desired racial mix. This the CSA not only does not require, but, in fact, was specifically implemented to avoid. . The exception was an elementary school built in the core city to replace an unsafe school. .

The NAACP also contends that the Board is guilty of bad faith because it has not filled empty seats in core city schools with white students in an attempt to maximize desegregation but rather has used portable classrooms to keep white students at white schools. The district court correctly found that there is no evidentiary support for either of these contentions. The evidence was that the empty seats in the core city would in no way accommodate the pppulation growth in the outlying areas without increasing those seats by busing black students out.

The Board’s facilities superintendent also testified that portable classrooms are not used at schools where student enrollment is not within the CSA goals.