Full opinion text
TABLE OF CONTENTS I. 686 History of the Case II. The Parties: Who's Still Here 689 III. Demography and Geography: The city, the school district, and the students 690 A. The City of Dallas 690 B. Scholastic Enrollment 692 C. Subdistrict Organization 695 D. Summary 699 IV. Relevant Cases and Principles 701 V. The Constitutional Violation 705 VI. ' Constitutional Adequacy of the Existing Plan 707 VII. Kindergarten-Third Grade Schools 713 A. Demographics of the K-3 Schools 714 B. Time and Distance Study, K-3 Grades 724 C. Impact of Busing on K-3 Education 730 D. Opposition of Minority Parents to K-3 Busing 732 VIII. High Schools 733 IX. Fourth through Eighth Grade Schools 736 X. East Oak Cliff 739 A. Background of the East Oak Cliff Subdistrict 740 B. Implementation of the 1976 Court Order 741 C. Feasibility of Transportation to Desegregate East Oak Cliff 743 XI. Magnet High Schools 744 XII. Majority-to-Minority Transfers 748 XIII. Remedy 749 OPINION OF THE COURT SANDERS, District Judge. Today the Court decides that vestiges of state-imposed racial segregation remain in the Dallas Independent School District (DISD). The Court holds that additional systemwide transportation is not a feasible remedy for the existing constitutional violation. The Court believes, however, that effective remedies can be fashioned and directs the parties to prepare and file desegregation plans for the Court’s consideration. I. History of the Case This case is before the Court on remand from U.S. Court of Appeals for the Fifth Circuit. Tasby v. Estes, 572 F.2d 1010 (1978), reh. en banc denied, 575 F.2d 300 (1978); cert. dismissed, Estes v. Metropolitan Branches of Dallas NAACP, et al., 444 U.S. 437, 100 S.Ct. 716, 62 L.Ed.2d 626 (1980). The Circuit has directed this Court to make specific findings regarding the feasibility of using the desegregation techniques approved in Swann v. CharlotteMecklenburg Board of Educ., 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971) (Swann), to reduce the number of one-race schools in the Dallas Independent School District. This ease was originally filed in October 1970, Plaintiffs seeking the development of a comprehensive desegregation plan for DISD. The case was.tried in July 1971, Tasby v. Estes, 342 F.Supp. 945, aff’d in part, rev’d in part, and remanded with directions, Tasby v. Estes, 517 F.2d 92 (5th Cir. 1975); tried again in 1976, Tasby v. Estes, 412 F.Supp. 1192; and remanded again in 1978, Tasby v. Estes, 572 F.2d 1010 (5th Cir.), with directions to make “findings to justify the maintenance of any one-race schools that may be a part of” a new student assignment plan. 572 F.2d at 1018. The DISD was no stranger to desegregation litigation when this action was initiated, having been involved in several similar lawsuits since the 1955 U.S. Supreme Court decision in Brown II. Brown v. Board of Education, 349 U.S. 294, 75 S.Ct. 753, 99 L.Ed. 1083 (1955). A “stair-step” (one grade per year) plan for desegregation was ordered by the federal court in 1960. Implementation began at the first grade level in the 1961-62 school year, and thereafter, the DISD converted from dual attendance zones to single zones on a grade-a-year basis until 1965, when the Fifth Circuit ordered the process accelerated to include all six elementary grades as well as the twelfth grade. Dual zones were eliminated for junior high schools in 1966 and for the remaining grades ten and eleven in 1967. The “stair-step” plan merely called for the elimination of racial criteria from the school system’s admission policies. The courts did not direct DISD (and DISD did not volunteer) to take affirmative action to eradicate the vestiges of the former statutory segregated system. So, while it can fairly be said that DISD, like many another school district, moved with maximum deliberation and minimum speed to carry out the 1955 desegregation mandate of the U.S. Supreme Court, it should also be said that the federal court moved at the same pace; DISD did what the Court ordered — no more, no less. Since the filing of Tasby v. Estes ten years ago, a number of complex issues have been raised and resolved in this Court and the appellate courts. (1) Elements of the former dual system were found to remain, in violation of the Constitution, as evidenced chiefly by the number of segregated schools present in 1971. 342 F.Supp. at 947. (2) The Court was unable to find that de jure segregation had been practiced against hispanic students; nevertheless, it determined that any remedy or plan ultimately adopted would be triethnic in scope and treat hispanics as a distinct ethnic minority group for purposes of student assignment. 342 F.Supp. at 948; aff’d, 517 F.2d at 106-07. (3) The issues relating to metropolitan, interdistrict violation and remedy were separately litigated. Plaintiffs failed to make the showing of “significant segregative effect” required under Milliken I, with respect to the suburban Highland Park I.S.D., thereby foreclosing the possibility of multidistrict remedy. 412 F.Supp. at 1188-91; aff’d, 572 F.2d at 1016. Plaintiffs voluntarily dismissed their complaints against six other suburban school districts. In 1976 this Court (Taylor, J.) after considering desegregation plans submitted by DISD, Plaintiffs, a court-appointed expert and others, adopted the plan submitted by amicus curiae, the Dallas Alliance. The Alliance is a community service organization consisting of seventy-seven correspondent organizations in the Dallas area and having a forty member Board of Trustees, which seeks to assist in focusing attention on and devising solutions for a wide range of current issues facing the city. The Alliance plan had been prepared by its tri-ethnic twenty-one member Educational Task Force after an intensive four-month study. The essential features of the Alliance plan may be summarized in the following fashion. The school district has been carved into six subdistricts, four of which are geographically structured in such a way that their enrollments reflect the same racial composition as the overall DISD, plus or minus five percent. The remaining two subdistricts were exempt from the racial composition requirement. Within each of the subdistricts, students in grades kindergarten through three (K-3) were assigned according to then existing assignment patterns, with special emphasis placed on compensatory education involving individualized, diagnostic-prescriptive methods of instruction. In grades 4-8, satellite zoning and transportation techniques were used to establish desegregated schools located around the center of each subdistrict. In grades 9-12, magnet schools and majority-to-minority transfer options were the principal tools of desegregation, leaving students who declined to opt for one of these programs attending the high school in their regular neighborhood attendance zone. Exceptions, to the general student assignment principles were made for students who resided in areas which were considered integrated by residential housing patterns (the “naturally integrated areas”), and for students in the East Oak Cliff and Seagoville subdistricts. Other provisions of the desegregation plan concerned special instructional programs, transfer options, discipline policy, facilities construction and improvement, faculty assignment and training, and accountability. After the ease was returned to this Court by the Supreme Court in 1980, the parties negotiated unsuccessfully in an effort to reach an agreement on a desegregation plan. Time and distance studies were ordered by Judge Taylor and trial on the issues raised in the Circuit’s mandate was set for March 16, 1981. On March 13 Intervenor NAACP by motion requested Judge Taylor to recuse himself; Judge Taylor did so and on March 20, 1981, this case was assigned to the undersigned Judge for trial. Trial commenced April 27 and concluded May 22, 1981. In remanding this case, the Fifth Circuit expressed concern about the number of one-race schools under the 1976 Plan — specifically, K-3 schools, high schools, and schools in the East Oak Cliff Subdistrict. The Circuit stated: We cannot properly review any student assignment plan that leaves many schools in a system one race without specific findings by the district court as to the feasibility of [Swann desegregation] techniques. 572 F.2d at 1014. This Court is also directed by the Circuit “to reevaluate the effectiveness of the magnet school concept”, 572 F.2d at 1015, and “to consider assigning anglo students to the [Nolan Estes Educational Plaza ] complex” located in East Oak Cliff. 572 F.2d at 1017. II. The Parties: Who’s Still Here As is quite often the ease with protracted litigation, the cast of characters appearing before the Court has changed noticeably since 1976,' and in some respects has changed radically since these proceedings were initiated in 1971. Intervenors new and old have come and gone. Out of the fourteen or so parties who remain of record in this litigation, eight participated in the month-long hearing just concluded. Evidence was introduced at the hearing on behalf of the Plaintiffs Tasby, et al., representing a class of black and hispanic students; the Defendants, Superintendent Linus Wright and the individual members of the Board of Trustees of DISD ; the Curry intervenors, representing a group of far North Dallas parents and students; the Brinegar intervenors, representing a group of people who reside in portions of East Dallas that are multi-racial and multi-ethnic as a result of residential housing patterns; the NAACP intervenors, representing a class of black students; the Cunningham intervenors, representing two black students; and the Black Coalition to Maximize Education (“Black Coalition”), a newcomer among the intervenors, representing a class of black parents with children in DISD whose interests and positions on the issues in this case were found to differ from those of the other parties. The Educational Task Force of the Dallas Alliance, continuing in its capacity as amicus curiae, did not call witnesses or put on evidence of its own, but did participate by questioning the witnesses called by the other parties, in an effort to lend further assistance to the Court. The precise positions of each of the parties on the issues before the Court can be discerned from a reading of the proposed pretrial orders filed by each. Broadly speaking, it can be said that the Defendant DISD, the Brinegar intervenors, the Curry intervenors, and the intervenor Black Coalition have all introduced evidence in an effort to justify, constitutionally, the maintenance of all or most of the provisions of the Court’s 1976 Plan. The Plaintiffs Tasby, the intervenor NAACP, and the intervenors Cunningham, on the other hand, have sought to prove that the techniques used in the 1976 Plan are constitutionally inadequate to remedy the vestiges of past segregation and that the DISD has failed to achieve the greatest possible degree of actual school desegregation. A further comment should be made about the stance taken by the Black Coalition, a party that has only become involved in this litigation in recent months and has, consequently, not been previously identified in the reported opinions of this case. The Black Coalition is a broad-based minority community group composed of parents, patrons and taxpayers with children in the DISD, as well as representatives from a number of civic, political and ecumenical associations in the black community. Organizers of the group sought input from Dallas blacks on the issues pending before this Court by conducting a series of open meetings in the community during 1980. Leave of court was granted for the Black Coalition to intervene in the matter on January 19, 1981, after the Court was satisfied that the viewpoints of its members were not adequately represented by (and were often adverse to) the Plaintiffs and intervenor NAACP. The testimony offered by the large number of witnesses who testified in behalf of the Black Coalition has convinced the Court that there is considerable difference of opinion among sizeable segments of the minority citizenry of Dallas over the type of relief that should be ordered in this case. The Court is in no position, based on the testimony offered thus far, to determine which of the parties speaks for the greater number of Dallas blacks and hispanics. It is doubtful that such a thing can be accurately measured or surveyed, but even if it were capable of proof, this is a finding that need not be made. What is clear from the testimony is that no one party to this suit can lay claim any longer to speaking on behalf of the entire minority population as a sacrosanct “class”. As one noted legal scholar has put it: Basic principles of equity require courts to develop greater sensitivity to the growing disagreement in black communities over the nature of school relief. Existing class action rules provide ample authority for broadening representation in school cases to reflect the fact that views in the black community are no longer monolithic .... [I]t is incumbent on the courts to ensure the fairness of proceedings that will bind absent class members. By its intervention in this lawsuit, the Black Coalition merely gives formal recognition to the same undercurrents of tension and disunity among blacks that were experienced over the lengthy course of desegregation litigation in such large cities as Atlanta, Detroit, Nashville and Boston. Al- . though the multiplicity of minority viewpoints is not a situation unique to Dallas, the Court is not aware of any other case where intervenor status has been taken for the purpose of advancing the divergent positions on the record. The Black Coalition represents a substantial body of blacks who are opposed to any escalation in the use of racial balance remedies to cure the effects of school segregation. The Coalition prefers remedies designed to improve educational quality and to eliminate the disparity in academic achievement that can be attributed to past segregation, as alternatives to remedies that require pupil reassignments to non-contiguous attendance zones and mandatory transportation. III. Demography and Geography: The city, the school district, and the students A. The City of Dallas The starting point for evaluating any desegregation plan is determining the numbers and whereabouts of the school-age children affected. In the forty years following the start of the Second World War, Dallas has undergone profound changes in demography. The Dallas experience is prototypical of the rapid urbanization that has occurred during the last several decades in numerous boom towns throughout the “Sunbelt”. To a great extent Dallas has become a new city, largely suburban in its development pattern. Since 1940 the physical area embraced by the corporate boundaries of Dallas has quadrupled to include well over 400 square miles and the population has nearly tripled. To simply recite the current statistics without some indication of the historical trends that have contributed to their existence would present a very incomplete picture of the enormous task facing this equity court as it strives to devise a remedy that will retain vitality in subsequent school years. Substantial changes in the size, location, socio-economic status, and ethnic composition of the population of Dallas and DISD have occurred between each stage of this litigation. The changes have been so substantial that the Court is convinced an element of clairvoyance must necessarily be involved in the formulation of a proper decree. By way of background then, the Court offers the following sketch of the physical and social geography of Dallas. The Trinity River flowing from northwest to southeast bisects the city. On the north bank, at the city’s core, lies the central business district and beyond it the residential areas of near and far North Dallas, South Dallas and East Dallas; south of the river lie the residential areas of West Dallas and Oak Cliff. The city spans over twenty-five miles from its northernmost point to its southernmost and is twenty miles across at its widest. Further expansion in the territory of the city has been curtailed by the corporate boundaries of the suburban cities that now ring the Dallas city limits. In 1940 there were slightly less than 300,-000 citizens of Dallas, of whom slightly over 50,000 (or 17%) were black. By 1970 the total population of the city of Dallas had risen to more than 840,000, of whom roughly 25% were black. While census tallies for 1980 are only preliminary at this writing, it appears that the population of Dallas has surpassed 904,000, including more than 250,-000 blacks, or 29%. While the population of Dallas increased only 7% in the last decade, the number of people living in the surrounding suburbs rose 58%, from 576,000 in 1970 to 912,000 in 1980. The people who live in the suburbs now outnumber those who live in the city of Dallas. This suburban population boom is expected to continue unhampered throughout the next decade. In the city of Dallas proper, the major spurts of growth occurred in the early 1950’s and again in the early 1960’s. The newcomers tended to settle in neighborhoods, initially, along racial lines. 1. Blacks. In 1940 as many as three-quarters of the 50,000 blacks in Dallas lived in two areas of the city: just north of the central business district and several miles to the southeast of the central business district, though still north of the Trinity River. While there was considerable territorial annexation occurring to the north and south of the center of town, this pattern of racial location remained constant until the early 1950’s. During the fifties Dallas doubled in size and population. Large numbers of blacks moved into the area known as South Dallas, between the central business district and the Trinity. South Dallas was an older area of town that was being abandoned as whites moved into the newly built subdivisions of North Dallas. The West Dallas area between W. Commerce Street and the Trinity River was annexed to the city in 1957, adding several hispanic neighborhoods, a few whites, and a substantial number of blacks to the population. At the time of its annexation, less than 11% of the housing in this area passed code and there were only two miles of paved streets, only one 60-year-old school building, no parks, sewers or storm drains. In the area now known as East Oak Cliff (bounded by Interstate 35 on the west, the Trinity River on the north, and Interstate 45 on the east), the 1950’s saw the construction of 65% of the present housing in a series of subdivision developments. Initially, blacks resided in only two areas of Oak Cliff — in far northeast Oak Cliff along the south shore of the Trinity (directly opposite South Dallas), and in the far southeast corner of Oak Cliff near the site to which Bishop College had moved from Marshall, Texas. While most of the rest of East Oak Cliff had been originally marketed to whites, during a three-year period in the 1960’s the area changed from an almost entirely white residential area to a primarily black area. The average occupancy of these new homes by their original white owners was 2.3 years. This sudden racial crossover, taking place in a newly developed part of the city, with new streets, schools and commercial centers, was virtually unprecedented in this or any other city of its size. 2. Whites. Residential location of whites during this same period followed a different pattern, spreading to the outer fringes of the city limits — to the far north and northeast, the far east (Pleasant Grove), and in Oak Cliff, to the far southwest. The more affluent whites concentrated in the new (post-1950’s) subdivisions of far North Dallas, north of the island cities of Highland Park and University Park. Additionally, large numbers of whites participated in the tremendous expansion of outlying suburban cities that was beginning to take place. 3. Hispanics. Unavailability of separate census information makes it difficult to trace the movement of hispanics in Dallas through the same time periods. Aside from the West Dallas neighborhoods that were annexed in 1957, most of the hispanic population residing in Dallas in the 1940’s and 1950’s was concentrated in a community just north of the central business district. Most of the hispanic population presently residing in Dallas has come to the city in the late 1960’s and 1970’s. They live primarily in a wide belt that stretches from West Dallas eastward across the Trinity, through the area between the central business district and the Park Cities, and well into near East Dallas. There is also a sizeable hispanic population residing in Oak Cliff, in a wide corridor along the west side of Interstate 35E. Presently, the demographic composition of Dallas is impacted by certain related factors, which may come to bear on the task ultimately before this Court. The overall expansion in territory and population has been halted by the presence of the suburban ring. Further increases in population will be accommodated only by increased density of housing units or the construction of multi-family, apartment complexes. Recent years have seen a revitalization and expansion of the central business district; many new skyscrapers have been built or are under construction and specific plans for still others have been announced. As the population of Dallas and its suburbs increases, the number of automobiles on its already busy thoroughfares rises proportionally. Traffic congestion along the arteries and freeways leading to and from downtown in all directions is at its worst in the morning and afternoon rush hours, and will get far worse. There are no currently planned projects to relieve the freeway congestion in the foreseeable future, and any such projects, if ever approved, would take years to implement. B. Scholastic Enrollment Although the city of Dallas and the DISD include largely the same territory, their boundaries are not entirely coterminous. The school district is smaller than the city, covering approximately 351 square miles. The outer boundary of the DISD is fixed by the presence of a ring of suburban independent school districts which serve the surrounding suburban cities described above. The smaller territory of the school district can be partly attributed to the fact that a portion of the northeast corner of the city of Dallas (north of Northwest Highway) lies within the boundaries of the Richardson I.S.D. rather than DISD; this is a predominantly white residential area. The DISD is further geographically distinguished from the city of Dallas by reason of the annexation into DISD of the former Seagoville School District in 1965. Seagoville is a sparsely populated, separately incorporated rural area which is located nearly 18 miles southeast of downtown Dallas. On a diagonal line from the northernmost point of the district near the Dallas County line to the southeast corner of the district in Seagoville there is a distance of over 35 miles. The total number of persons residing within the DISD boundaries (as distinguished from the figures for scholastic enrollment) can be approximated by cumulating the statistics from each of the U.S. Government census tracts that are also located within DISD. In 1970, there were 801,211 people residing in the area served by DISD. By 1980, that figure had increased only slightly to 808,-714, which is about 100,000 fewer persons than live within the city of Dallas. While the total population statistics have levelled off over the last ten years, the same cannot be said for the number of scholastics enrolled in DISD schools. Throughout the course of this litigation DISD has maintained the position of the eighth largest city school system in the nation. At present it remains among the ten largest, notwithstanding the remarkable decline in enrollment and dramatic changes in racial composition of the district since 1970. At the time this suit was brought, total enrollment for the 1970-71 school year was 163,353, exclusive of some 1400 kindergarten students. By the end of the 1980-81 school year, enrollment had dropped to 128,-658, including kindergarten students, representing an overall decline in excess of 21%. This steady decline over the decade was produced by a combination of extremely volatile decreases and increases occurring simultaneously in the respective numbers of anglos- and minorities attending DISD schools. i. In 1970-71, the total enrollment was 58.2% anglo (95,012 students), 33.4% black (54,612 students) and 8.4% his-panic (13,729 students). By 1981, anglo enrollment had dropped by 60%, from 95,012 students to 37,989. ii. Black enrollment rose to a peak of 65,541 in the 1977 school year and has since declined slightly to 63,827, for an overall increase since 1970 of 16.9%. iii. Hispanic enrollment has ballooned in the last decade, jumping 82.2% from 13,729 students to 25,010. Thus, at present the ethnic composition of the district is 29.53% anglo, 49.61% black, and 19.44% hispanic. In a period of ten short years the DISD has undergone a transformation from a student population of 163,000 that was nearly 60% anglo to a district of 128,000 students that has a combined black and hispanic enrollment of nearly 70%. The dramatic changes in the racial composition of the DISD enrollment since 1970 are not characteristic of similar changes in the makeup of the population residing in the area served by DISD during the same period. While the school district lost 21% of its enrollment, the general population in the district remained fairly constant, showing a .93% increase. Black residents increased from 198,409 to 246,132 between the 1970 and 1980 censuses, a percentage rise of 24% since 1970 vis-a-vis a 16.9% increase in school enrollment. But, where black students comprise 49.6% of the DISD enrollment, blacks only represent 30.43% of the 1980 population. Similarly, hispanic residents increased from 51,061 to 109,697 between the 1970 and 1980 censuses, a percentage rise of 114.83% since 1970, compared to an increase in hispanic enrollment of 82.2%. Hispanics now constitute 13.56% of the population and 19.44% of the DISD enrollment. The parallel between population and enrollment trends breaks down, however, when the anglos are taken into consideration. Census figures reveal that the number of anglo residents within the district boundaries dropped from 551,741 in 1970 to 452,885 in 1980. This decline of 17.91% in the anglo population was far less drastic than the 60% reduction in anglo enrollment for the same period. And, although anglo students comprise only 29.53% of the DISD enrollment, anglos still make up 56% of the population residing within the district. Thus, while the rise in hispanic enrollment seems to reflect a similar surge in the numbers of hispanic residents, the sharp drop in anglo enrollment is not indicative of a concomitant decline in anglo population and the percentage of blacks in the schools is substantially greater than the percentage of blacks in the population. By contrast to the steady decrease in DISD enrollment since 1970, enrollment in the numerous suburban independent school districts surrounding DISD has increased steadily over the last ten years. Since the 1971 — 72 school year, total suburban enrollment has risen 24.2% while DISD enrollment decreased by 18.7%. In terms of numbers of students (/. e., arithmetic changes), suburban enrollment has increased across the board for blacks, anglos and hispanics. In fact, DISD is the only school district in the Dallas metropolitan area that has sustained a significant decline in anglo enrollment since 1970. The ethnic composition of the combined suburban school districts at the present time is 7% black, 5.9% hispanic, and 85.3% anglo. It is possible, using a variety of accepted statistical approaches, to predict with reasonable accuracy the size and ethnic composition of DISD enrollment for each of the next five school years. These projections assume a continuation of present student assignment patterns and they indicate that total enrollment is expected to continue to decrease from the present 128,658 to 122,-218 by the 1985-86 school year. The same trends that were observed in recent years in the enrollment patterns of each of the major ethnic groups of students are expected to continue as well: i. Hispanic enrollment is again expected to rise substantially from the present 25,010 students (19.44% of total district enrollment) to 32,773 students (26.8%) in 1985-86. ii. Black enrollment is projected to decline slightly from the present 63,827 students (49.61%) to 59,482 students (48.7%) in 1985-86. iii. The projections also foreshadow a continuation of the sharp decline in anglo enrollment from the present 37,989 students (29.53%) to 27,707 students (22.7%) in 1985-86. These figures portend the loss of another 29.1% of the present anglo enrollment and the increase of the hispanic enrollment by 32.8% in just four years’ time. Most of this loss will occur in the earlier grades. C. Subdistrict Organization Due to the geographic layout of the district and the low population density of Dallas as a whole, the DISD embraces a great number of students who are dispersed throughout a sprawling urban area. For purposes of devising an effective desegregation plan, as well as overall administrative efficiency, it was recommended by Plaintiffs and certain other parties submitting proposed desegregation plans at the 1976 hearings that DISD be subdivided into smaller, more manageable units. Other large city school districts, such as Boston and Detroit, have also found it expedient to create a subdistrict structure. Judge Taylor’s 1976 Order divided DISD into six geographic subdistricts, each containing elementary, middle and high schools. The subdistrict concept was devised to provide parents and students a sense of community participation and local control over their schools, which the Supreme Court has recognized as essential both to the maintenance of concern and support for public schools and to the quality of the educational process. See Miliiken I, 418 U.S. at 742, 94 S.Ct. at 3126. The Court finds that, after five school years of operation under this structure, the subdistrict concept has proved to be a useful tool, both for the day-to-day administration of the public schools and for the desegregation of the district, and concludes that there is insufficient evidence of any benefit to be obtained by displacing this organizational structure at this time. As originally designed in 1976 and implemented today, the six geographic subdistricts used for purposes of student assignments are the Northeast, Northwest, Southeast, East Oak Cliff, Southwest and Seagoville subdistricts. The Northeast subdistrict includes the central business district and much of South Dallas and extends to the northeast along East Grand Avenue and Garland Road through the Munger Place neighborhood, East Dallas, Lakewood, and the area east of White Rock Lake to the DISD boundary line. The Northwest sub-district begins just north of the central business district and includes near North Dallas, West Dallas (along the south bank of the Trinity River), the Arlington Park industrial area, the Love Field area, and the residential subdivisions of far North Dallas that lie between the Park Cities and theDISD boundary on the west and north. The Southwest subdistrict takes in what is essentially the western half of Oak Cliff, or that quadrant defined by 1-30 on the north, I-35E on the east, and the DISD boundary line on the south and west. The East Oak Cliff subdistrict is circumscribed by I-35E on the west, the Trinity River on the north, 1-45 on the east, and the DISD boundary line to the south. The Southeast subdistrict begins at the far southeast end of South Dallas, along Hatcher Street, and continues to the eastern boundary of DISD and to the southeast along the C. F. Hawn Freeway, through the Pleasant Grove area, to the subdistrict boundary at Jordan Valley Road. To the west of the Southeast subdistrict lies a large triangular region of undeveloped flood plain along the Trinity River, further separating Southeast from East Oak Cliff. The Seagoville subdistrict lies contiguous to the southeast boundary of the Southeast subdistrict and extends along the C. F. Hawn Freeway to the rural town of Seagoville, some 18 miles from downtown Dallas. This somewhat remote subdistrict takes in the area that was served by the old Seagoville I.S.D., prior to its merger into DISD in 1965. The subdistrict boundaries were not drawn with an eye towards dividing the DISD into six territories that would be roughly equal in physical dimensions. While the Northeast, Southwest and Southeast subdistricts are more or less the same size, one subdistrict is considerably larger and two subdistricts are considerably smaller than the rest. The East Oak Cliff sub-district is the smallest of the six, covering only about 21 square miles, but this subdistrict enjoys a greater population density than other areas of the district. The largest subdistrict, the Northwest subdistrict, spans approximately 90 square miles, with a lower population density than is found in other parts of the district. Instead of trying to create six areas of equal dimensions, the subdistricts were carved in such a way as to cause the ethnic composition of as many of the individual subdistricts as possible to mirror the racial makeup of the DISD as a whole. When the district court issued its memorandum opinion in March 1976, the DISD was 41.1% anglo, 44.5% black, 13.4% hispanic, and 1% other. Tasby v. Estes, 412 F.Supp. at 1208, Appendix A. Accordingly, the Final Order entered April 7,1976, called for the creation of four subdistricts each having approximately the same racial makeup, plus or minus five percent, as the entire DISD—these were the Northeast, Northwest, Southwest and Southeast subdistricts. Because of their geographical isolation from other parts of the city, two of the subdistricts created — Seagoville and East Oak Cliff — were allowed to deviate substantially from the ethnic ratio of the entire school district. As could be expected from the previous discussion of the recent trends in district-wide enrollment patterns, the decline in student population and ■ changes in ethnic composition can also be detected within each of the subdistricts to a greater or lesser degree, depending on the unique demographic circumstances inherent to each. The following table demonstrates the numerieal breakdown of enrollment by race for all grades in each of the six subdistricts for the years 1971, 1976, and 1981. Additionally, the table includes projections for the ethnic composition of each subdistrict for the school year 1985-86. K-12 Enrollment by Subdistrict: 1971-1985 I. NORTHEAST Total Anglo Black Hispanic 1971 32,088 21,687 7,967 2,259 1976 28,893 15,120 10,188 3,194 1981 25,445 10,843 9,707 4,277 (1985-86) 23,303 8,196 8,257 6,072 II. NORTHWEST 1971 42,634 23,641 12,094 6,647 1976 31,598 13,502 10,548 6,991 1981 25,525 8,715 8,612 7,519 (1985-86) 23,413 5,852 7,878 8,754 III. SOUTHWEST 1971 28,455 21,508 2,584 3,992 1976 28,184 11,081 9,429 7,255 1981 28,071 6,943 11,142 9,611 (1985-86) 33,863 4,715 14,560 14,166 IV. SOUTHEAST 1971 29,009 18,455 9,089 1,398 1976 21,328 10,998 8,467 1,769 1981 18,525 8,314 8,054 2,059 (1985-86) 20,218 8,713 8,460 2,939 V. EAST OAK CLIFF 1971 26,407 343 25,315 732 1976 27,028 349 26,223 439 1981 25,153 227 24,017 898 (1985-86) 21,421 233 20,327 842 K-12 Enrollment by Subdistrict: 1971-1985 Total Anglo Black Hispanic VI. SEAGOVILLE 1971 Data Not Provided 1976 2,307 1,862 373 127 1981 2,629 2,132 344 143 (1985-86) No Projections Made The numbers speak for themselves. The Court would simply add this gloss to highlight certain recurring themes. Between 1971 and 1981 total enrollment has declined in every major subdistrict. In the three large subdistricts north of the Trinity River there has been a dramatic drop in scholastics: Northwest has lost 40.12% of its students, Northeast has lost 20.70% of its students, and Southeast has lost 36.14% of its students. The two subdistricts south of the Trinity River have remained fairly constant in total enrollment: East Oak Cliff fell only slightly, from 26,-407 to 25,153; and Southwest was virtually unchanged, from 28,455 in 1971 to 28,184 in 1976 to 28,071 in 1981. Contrasted against this backdrop of steady decline in overall enrollment, the several ethnic components within each sub-district have exhibited even wider fluctuations over the same ten year period: i. Hispanic enrollment has risen in every subdistrict. In particular, those three subdistricts through which the wide belt of hispanic residents passes have experienced a phenomenal surge in hispanic students: in Northeast, there was an 89.33% increase between 1971 and 1981; in Northwest, the already sizeable his-panic student body in 1971 swelled another 13.11% by 1981; and in Southwest, hispanic enrollment soared by 140.75%. ii. Black enrollment during the same decade remained fairly constant, in contrast to the other ethnic groups, rising slightly in Northeast and dropping slightly in Southeast, East Oak Cliff and Seagoville. In Northwest, however, there was a more substantial decline of 28.79%, while in Southwest, black enrollment leaped 334%, from 2,584 to 11,142. iii. There has been a precipitous loss of anglos in every major subdistrict, particularly in those located north of the Trinity River. In Northeast, the anglo enrollment was cut in half by 1981 (10,844 fewer students); in Northwest, it declined 63.14% (14,926 fewer students); in Southeast, it dropped 54.94% (10,141 fewer students). On the other shore of the river, Southwest experienced a 67.71% loss of anglos (14,565 fewer students). In terms of percentages, anglos now comprise 42.60% of total Northeast enrollment (down from 67.6% in 1971), 42.73% of all Northwest scholastics (down from 55.5% in 1971), 44.88% of all Southeast scholastics (down from 63.6%), and 24.73% of all Southwest scholastics (down from 75.6% in 1971). Oddly enough, this precipitous decline in anglo scholastic enrollment over the last decade is not mirrored by a similar trend among the general population residing within corresponding subdistrict boundaries. Using census information, the population statistics for each of the subdistricts in DISD can be accurately approximated for the years 1970 and 1980. Anglo enrollment in Southwest dropped 67.71% in the last decade, while the anglo population residing in Southwest declined only 40%. Even starker was the difference between enrollment and population residing in Northeast, where anglo scholastics decreased 50% while anglo residents decreased only 19%. Without a doubt, the greatest disparity between enrollment and population statistics occurs in Northwest, where most of the affluent white suburban developments in the city of Dallas are to be found. There, the anglo exodus from the public schools measured 63.14% while the anglo population residing in the subdistrict actually increased by 4,449 people. Census figures reveal that the hispanic and black enrollment trends in each of the subdistricts more closely resemble the fluctuations in the ethnic residential patterns of the subdistricts during this period. Finally, the projections for scholastic enrollment patterns in each of the subdistricts over the next five years indicate that many of the observed trends will continue unabated. i. Between now and the 1985-86 school year, hispanic students are expected to increase by 42% in Northeast, by 47% in Southwest, and by 42.7% in Southeast. ii. Black enrollment will remain fairly constant in most districts, but will drop 15% in East Oak Cliff and increase 30.67% in Southwest by 1985-86. iii. Anglo enrollment is expected to decline another 24.41% in Northeast, 32% in Southwest, and 32.85% in Northwest. D. Summary In the interest of presenting the most complete and accurate picture of the demographic and geographic landscape of DISD, past, present, and future, the Court has laboriously examined each “tree”, branch by branch. Now to step back and look at the forest. The single most compelling characteristic of the DISD for purposes of designing an effective desegregation plan is the hard fact that DISD is and probably will remain a district composed of a majority of minority students. Together, the blacks and his-panics comprised 42% of the student body in 1970; today they account for almost 70% and, by 1985, their numbers will exceed 11% of the total enrollment. The dwindling anglo enrollment will continue to dwindle, as every year a greater number and percentage of anglo students are graduated from the system than enter the kindergarten and elementary grades. This anglo decline does not reflect a comparable outmigration of anglo residents from DISD, particularly in the northeast and northwest corners of the district. Suburban enrollment is overwhelmingly anglo and increasing in number every year. Regardless of whether or not these occurrences are related to one another or were in some way caused by these proceedings, they are in themselves “interesting statistics.” While the Court places emphasis on the demographic feature of declining anglo enrollment in shaping a remedy for this case, it should not be assumed that the Court lends its imprimatur to the theories of “white flight” or “white avoidance” advanced by some of the parties to this suit, most notably the Curry Intervenors. This oft-sung refrain — that any new measures designed to increase the amount of desegregation will cause white flight and “resegregation” in the schools — has fallen on deaf ears in virtually every court where it has been uttered. See, e. g., United States v. Scotland Neck Cty. Bd. of Educ., 407 U.S. 484, 491, 92 S.Ct. 2214, 2218, 33 L.Ed.2d 75 (1972); United States v. DeSoto Parish School Bd., 574 F.2d 804 (5th Cir.), cert. denied, 439 U.S. 982, 99 S.Ct. 571, 58 L.Ed.2d 653 (1978). Moreover, as was ably demonstrated by counsel for Plaintiffs, it is impossible to say that school desegregation is the only impetus for the outmigration of school-aged whites from the system. In some cases, such outmigration may in fact be an unfortunate response to student assignment patterns imposed by the school board or the court on unwilling individuals. However, such factors as job relocations, housing costs and availability, or the inmigration of singles or families without school-aged children might also be causes of declining white enrollment. It is likely that all of these factors contribute to the phenomenon. The Court, therefore, cannot stay its hand out of fear of “white flight”. That does not mean that the Court must ignore the unpleasant realities of population changes as it devises a plan that will redress the vestiges of a past wrong and promote greater, lasting desegregation. Stout v. Jefferson County Bd. of Educ., 537 F.2d 800, 802 (5th Cir. 1976). Thus, the trends and projections of ever declining anglo enrollment in a heavily minority district, juxtaposed against the continued existence of a majority anglo resident population within the district and the ever-increasing anglo enrollment in suburban school districts, are conditions which must be factored into the final calculus. A second characteristic, every bit as critical as the enrollment trends and projections, is the unusual physical size of the district and the manner in which the several ethnic populations are dispersed throughout it — /. e., DISD’s “social.” geography. There are generally three types of housing patterns: neighborhoods that are primarily anglo, neighborhoods that are primarily minority, and neighborhoods that are residentially integrated, with shifting ethnic populations. Throughout most of the district, there is great physical separation of the races. In many large cities, the ethnic map of the city would resemble a doughnut, with white residents forming a ring around a black inner-city core. In such instances, dividing the district into pie-shaped wedges will accomplish considerable racial mixing on a smaller scale. Dallas, however, resembles a pie in which one whole “wedge” is made up of black residents, from the center of the district all the way to its outermost boundary. The predominantly black areas in East Oak Cliff are separated from the white neighborhoods, primarily in far north Dallas and northeast Dallas, first, by the Trinity River and its bottomlands, then, by the central business district with its congested traffic, then, by a buffer zone of integrated housing overlayed by a wide arc of hispanic residents, and finally, by the separate “Park Cities” in North Dallas and by White Rock Lake in northeast Dallas. The problem is further exacerbated by population density, which is extremely high in South Dallas and Oak Cliff (where most blacks live) and much lower in far north and northeast Dallas (where most anglos live). These, then, are the unique features in the light of which any desegregation plan must be evaluated. IV. Relevant Cases and Principles During the years since this litigation began, the higher federal courts have decided many significant cases. The principles announced in these cases guide this Court. Accordingly, the Court now turns to an examination of some of those cases and the facts out of which they originated. Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971), cited by the Fifth Circuit in remanding this case (572 F.2d at 1014, 1015), involved a metropolitan school system with 84,000 students (71% white and 29% black), and is chiefly noted for its approval of mandatory cross-district transportation to achieve desegregation. Chief Justice Burger, for a unanimous court, found that the school board “had totally defaulted in its acknowledged duty to come forward with an acceptable plan . . .,” 402 U.S. at 24, 91 S.Ct. at 1280, and held that busing a substantial number of students was an acceptable method of achieving desegregation. The Court also noted that the existence of some one-race schools “is not in and of itself the mark of a system that still practices segregation by law.” Id. at 26, 91 S.Ct. at 1281. Swann was preceded by unanimous Supreme Court decisions in Green v. New Kent County School Board, 391 U.S. 430, 88 S.Ct. 1689, 20 L.Ed.2d 716 (1968) and Davis v. Board of School Commissioners of Mobile County, 402 U.S. 33, 91 S.Ct. 1289, 28 L.Ed.2d 577 (1971). In Green the school system had only two schools and 1300 pupils —740 black and 550 white. There were no attendance zones; each school served the entire county. At issue was “a freedom of choice” plan, which the Court struck down as ineffective because it was not achieving desegregation. In Davis, the school district had 91 schools and 73,500 pupils, divided 58% white and 42% black, with 22,000 pupils already being transported for purposes not related to desegregation. The Supreme Court held, 402 U.S. at 37, 91 S.Ct. at 1292, that “.. . the district judge or school authorities should make every effort to achieve the greatest possible degree of actual desegregation, taking into account the practicalities of the situation.” In 1973 the Court addressed school desegregation in a case from Denver, Colorado, Keyes v. School District No. 1, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973). Denver had 119 schools and approximately 97,000 pupils, divided 66% anglo, 14% black, and 20% hispanic. In that case the Supreme Court originated the well known Keyes presumption: intentionally segregative school board actions in a meaningful portion of a school system, . . . creates a presumption that other segregated schooling within the system is not adventitious. It establishes, in other words, a prima facie case of unlawful segregative design on the part of school authorities, and shifts to those authorities the burden of proving that other segregated schools within the system are not also the result of intentionally segregative actions. 413 U.S. at 208, 93 S.Ct. at 2697. Keyes also stated another important, if lesser known, principle: In Swann, we suggested that at some point in time the relationship between past segregative acts and present segregation may become so attenuated as to be incapable of supporting a finding of de jure segregation warranting judicial intervention. (Citations omitted). We made it clear, however, that a connection between past segregative acts and present segregation may be present even when not apparent and that close examination is required before concluding that the connection does not exist. Id. at 211, 93 S.Ct. at 2698-99. In 1974 and again in 1977 the Supreme Court considered school desegregation in Detroit. Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974) (Milliken I); Milliken v. Bradley, 433 U.S. 267, 97 S.Ct. 2749, 53 L.Ed.2d 745 (1977) (Milliken II). The Detroit school system had approximately 257,000 students, of whom 71.5% were black. Milliken I chiefly concerned an interdistrict remedy ordered by the District Court, which the Supreme Court invalidated. Milliken II approved the use of remedial education programs as a method of eliminating the effects of prior segregation in a large urban district with a large minority student population, and also held, inter alia, that the purpose of a desegregation decree is “to restore the victims of discriminatory conduct to the position they would have occupied in the absence of such conduct.” 433 U.S. at 280, 97 S.Ct. at 2757. The Court further noted that an order requiring a “particular degree of racial balance” is “infirm as a matter of law”, Id. n.14, and approved the District Court’s plan “to prevent the disruption, by massive pupil reassignment, of racially mixed schools in stable neighborhoods”, 433 U.S. at 288 n.19, 97 S.Ct. at 2761 n.19. The Supreme Court has twice dealt with school desegregation in Dayton, Ohio. Dayton Board of Education v. Brinkman, 433 U.S. 406, 97 S.Ct. 2766, 53 L.Ed.2d 851 (1977) (Dayton I); Dayton Board of Education v. Brinkman, 443 U.S. 526, 99 S.Ct. 2971, 61 L.Ed.2d 720 (1979) (Dayton II). The Dayton schools had about 54,000 students, 43% of whom were black. Dayton had 69 schools, including 21 one-race black schools and 28 one-race white. Columbus Board of Education v. Penick, 443 U.S. 449, 99 S.Ct. 2941, 61 L.Ed.2d 666 (1979) was handed down the same day as Dayton II. The Columbus, Ohio, school system had 96,000 students, 32% of whom were black, with 172 schools, half of which were one-race schools. Columbus and Dayton II strongly reaffirmed the Keyes presumption and the continuing affirmative duty of the school board to establish a racially nondiscriminatory system. Chief Justice Burger, concurring in Columbus criticized the “unknown and unforeseeable affirmative duty to desegregate for the past 25 years”, Id. at 469, 99 S.Ct. at 2952; Justice Rehnquist, dissenting, said that the Court was making the Keyes presumption “essentially . . . irrebuttable”, Id. at 508, 99 S.Ct. at 2962. For over a generation, in scores of opinions, the Fifth Circuit has consistently required of school systems the “maximum desegregation practically achievable.” See, e. g., United States v. Seminole County School District, 553 F.2d 992, 995 (5th Cir. 1977). The Circuit has been unwavering in its commitment to school desegregation; in recent years its resolve has, if anything, become more firm. In Carr v. Montgomery County Board, 377 F.Supp. 1123 (M.D.Ala.1974), aff’d 511 F.2d 1374 (5th Cir.), cert. denied 423 U.S. 986, 96 S.Ct. 394, 46 L.Ed.2d 303 (1975), Chief Judge (now Circuit Judge) Frank Johnson, holding that the Montgomery County, Alabama, school system had a satisfactory desegregation program, made several cogent observations particularly relevant to the case at bar: Thus, it appears that a balance must be reached, one unquestionably subtle in its implications: while school system segregation must be actively disestablished, racial quotas for student population are not to be instituted. Id. at 1133. This Court desires to emphasize that the remaining predominantly black schools in this school system under the board’s plan cannot be effectively desegregated in a practical and workable manner. ... As this Court has observed time and time again in school desegregation cases, racial quotas and busing to achieve racial quotas are not required by the law. Id. at 1135. . . . every formerly all-white school in the Montgomery school system will ... be substantially desegregated. Id. at 1139. Calhoun v. Cook, 522 F.2d 717 (5th Cir. 1975), affirming 332 F.Supp. 804 (N.D.Ga. 1971) concerned the Atlanta, Georgia, school system. Atlanta had a student population 85% black and 15% white, a dramatic change since 1958 (70% white, 30% black). In 1975, after trying for years to effect desegregation, Atlanta had 92 one-race black schools out of 148. The school district had “substantially met” its goal of 30% black enrollment in every white school. The Fifth Circuit held that “based on live, present reality” the Atlanta school system was “free of racial discrimination.” 522 F.2d at 720. The Circuit stressed that the school board was controlled by blacks and that blacks were in the majority in the school administration. The Circuit also noted that the busing of blacks to white schools was opposed by a large group of black parents, and that time and distance studies showed that busing was impracticable as a desegregation remedy. In Lemon v. Bossier Parish School Board, 566 F.2d 985 (5th Cir. 1978), however, the Circuit refused to rule out busing as a remedy where there were three virtually all-white schools within two miles of an all-black school. U.S. v. DeSoto Parish School Board, 574 F.2d 804 (5th Cir.), cert. denied, 439 U.S. 982, 99 S.Ct. 571, 58 L.Ed.2d 653 (1978) concerned a rural school district with a student population 60% black and 40% white and no concentrated residential segregation. The district had 11 schools; several of the five black schools were in close proximity to several of the six predominately white schools. The Fifth Circuit ruled that white flight was not a valid argument against busing but said: This is not to say that a school board or Court must ignore a likely danger of an exodus of white students from a school system. ‘[I]n choosing between various permissible plans a chancellor may . . . elect one calculated to minimize white boycotts. ... He may not refuse to adopt a permissible plan and elect or confect one which preserves a dual system because of such fears.’ Stout v. Jefferson County Bd. of Educ., 1976, 5th Cir., 537 F.2d 800, 802. Id. at 816. This Court has also reviewed U.S. v. Valdosta Board of Educ., 576 F.2d 37, 39 (5th Cir.) cert. denied, 439 U.S. 1007, 99 S.Ct. 622, 58 L.Ed.2d 684 (1978) and Lee v. Tuscaloosa Schools, 576 F.2d 39, 41 (5th Cir. 1978) which emphasize that “a systemwide racial balance” is not the goal of desegregation; the goal is “to cure the continuing effects of the dual school system.” One-race schools, however, are subject to close scrutiny. Anderson v. Dougherty County Board of Education, 609 F.2d 225 (5th Cir. 1980). In U.S. v. Texas Education Agency (Lubbock I.S.D.), 600 F.2d 518 (5th Cir. 1979), the Fifth Circuit remanded the Lubbock school case for additional findings of fact. In Lubbock there were 32,000 students, divided 59% white, 27% Mexican-American, and 13% black. The Circuit held, in effect, that the District Court’s findings did not overcome the Keyes presumption: The district court’s task on remand is to determine how much ‘incremental segregative effect’ the School Board’s [past] intentional discriminatory acts had on the residential distribution of the Lubbock school population as presently constituted, when that distribution is compared to what it would have been in the absence of such intentional segregative acts. 600 F.2d at 527. . . . the School Board continues to bear the burden to show that its intentional [past] segregative acts did not contribute to the current segregation of those schools. Id. at 528. Lee v. Macon County, 616 F.2d 805 (5th Cir. 1980) involved a district which, like DISD, has long been entangled in desegregation litigation. Ruling that the district court’s reasons for retaining one-race elementary schools were “legally insufficient” the Circuit held: Unlike the Carr court’s conclusion that logistical difficulties prevented dismantling every one-race school, the present district court’s conclusion implies that the racial imbalance in Tuscaloosa’s elementary schools results not from de jure segregation but rather from demographic changes occurring since [the first attempt to desegregate].... Not until all vestiges of the dual system are eradicated can demographic changes constitute legal cause for racial imbalance in the schools. Id. at 809. The Circuit also ruled in Macon County that opposition to busing by experienced educators of both races was not decisive on the issue. However, the Circuit did note: Focusing on the target of a unitary system rather than a systemwide racial balance, the court may devise a constitutional plan that temporarily .or permanently leaves one or more racially identifiable elementary schools, or that omits some of the earlier grades from the busing program. 600 F.2d at 812. See, also, Flax v. Potts, 464 F.2d 865, 868 (5th Cir.), cert. denied, 409 U.S. 1007, 93 S.Ct. 433, 34 L.Ed.2d 299 (1972); Lee v. Lee County Board, 639 F.2d 1243, 1247 n.4 (5th Cir. 1981). More recently in Valley v. Rapides Parish School Board, 646 F.2d 925 (5th Cir. 1981) and U.S. v. Texas Education Agency (South Park I.S.D.), 647 F.2d 504 (5th Cir. 1981), the Fifth Circuit has emphasized the continuing affirmative duty to eliminate all vestiges of previous state-imposed segregation, at the same time stressing that the school system must be viewed as a whole and not school-by-school. From these and other representative cases, several of which are cited later in this opinion, this Court gleans the following principles relevant to the case sub judice: 1. Since Brown v. Board of Education, 349 U.S. 294, 75 S.Ct. 753, 99 L.Ed. 1083 (Brown II) in 1955, every school district which operated a de jure racially segregated system has been under an “affirmative duty to take whatever steps might be necessary” to eliminate all the vestiges of past school segregation. Green, supra, 391 U.S. at 437-38, 88 S.Ct. at 1693-94. 2. The Keyes presumption (that past segregative actions in a “meaningful portion” of a school system are presumed to be the cause of current segregated schools, absent “countervailing evidence”), as strengthened in 1979 in Dayton II and Columbus, may be irrebuttable. Demographic changes are insufficient to rebut the Keyes presumption. See, e. g., Lee v. Macon County, supra. But the Keyes presumption pertains to liability, not to remedy, or feasibility of remedy, which is the principal task before this Court. 3. There is a presumption against the continued existence of one-race minority schools; they will be tolerated only after “close scrutiny” by the courts. But racial quotas and racial balance in individual schools are neither a goal nor a requirement; in fact, they are disapproved. The school district must be viewed as a whole and not school-by-school; a constitutional plan may leave some racially identifiable schools if further desegregation is not feasible. 4. A school desegregation plan should be flexible and sensitive as well as effective and must take into account the “practicalities of the situation” in the particular district. The goal is to make whole the victims of past unlawful