Bradley v. Milliken, 460 F. Supp. 299 (E.D. Mich. 1978)

*300 *301 Louis R. Lucas, Memphis, Tenn., Thomas Atkins, Boston, Mass., for plaintiffs.

George T. Roumell, Jr., Theodore Sachs, Detroit, Mich., George L. McCargar, Jr., Asst. Atty. Gen., Lansing, Mich., for defendants.

DeMASCIO, District Judge.

We have re-examined carefully the demographic data characterizing the Detroit City Schools to determine whether further desegregation of Regions 1, 5 and 8 is possible. We undertake this task following an appeal of our remedial guidelines to the U.S. Sixth Circuit Court of Appeals. That court held our August 15, 1975 remedial guidelines for desegregation "insufficient as to Regions 1, 5 and 8" and remanded "for further consideration in regard to the[se] three central regions." Bradley v. Milliken, 540 F.2d 229, 239, 240 (6th Cir. 1976).

When considering all of the Regions collectively, the late Judge Stephen Roth found in 1972 that an adequate remedy based on pupil reassignments limited to the corporate limits of Detroit would be impossible. The U.S. Supreme Court did not disturb this finding. See Bradley v. Milliken, 402 F. Supp. 1096, 1104 (E.D.Mich.1975). We also agreed with Judge Roth's conclusion. We stated that:

If Detroit's school population were more equally divided between black and white or if the desegregation area were sufficiently large to permit greater equalization, it would be possible to diminish the inevitable limitations on the task of eliminating racially identifiable schools in the district.

* * * * * *

Limitations may be imposed by the desegregation area. For example, the black proportion of the population can be so great that racial balance will inevitably result in majority black schools. In such an area, only two alternatives are available: The desegregation area must be enlarged or flexibility must be permitted in defining a desegregated setting. 402 F. Supp. at 1102, 1131 (emphasis added).

The Sixth Circuit Court of Appeals has agreed on several occasions that desegregation within the geographic limitations of *302 Detroit is "extremely difficult (if not impossible)." See, e. g., Bradley v. Milliken, 540 F.2d 229, 236 (6th Cir. 1976).

As we seek to determine whether further desegregation is possible, it has become apparent that the litigants have not retained the adversary posture that characterized the liability issues tried before Judge Roth.^[1] This changing adversary posture is only natural since the Detroit Board of Education has been willing to desegregate the school district without court prodding. Even before this action was filed in 1971, the Detroit Board planned the first stage for dismantling the dual system existing in the Detroit City Schools but was prevented from carrying it out by an act of the state legislature. In response to our April 1975 order, the Detroit Board submitted a plan for reassigning 51,000 students to desegregate the school district. At the hearings following the Court of Appeals remand, the Detroit Board produced demographic data to establish that it had accomplished all the desegregative pupil reassignments possible and that additional reassignments were not feasible. In addition, the Detroit Board offered extensive testimony to support a three-part plan for desegregating faculty and staff.

The plaintiffs are well aware that the Detroit Board of Education is striving to desegregate the school district. The plaintiffs persist, however, in their conclusion that a school district is desegregated only when the racial composition of each school mirrors the system-wide racial ratio within 15% in either direction. On the other hand, the defendant Detroit Board, charged with the primary responsibility for maintaining a viable school system, insists that a sound plan for desegregation must take into account the hard realities at hand. During the most recent evidentiary hearings, the plaintiffs did not produce any evidence to refute the Detroit Board's contention that inclusion of Regions 1, 5 and 8 in the desegregation plan is no longer feasible. The plaintiffs argued instead that the Detroit Board's "contained classroom" method of delivering bilingual instruction is resegregating Region 2 and is providing neighborhood schools for one ethnic group. Tr. Sept. 7, 1977 at 211. Thus, although the parties continue to disagree on an acceptable formula for desegregation, they are not and cannot be true adversaries.

Moreover, the state defendants are not in an adversary posture with either the Detroit Board of Education or the plaintiffs. They did not even participate in this last round of hearings. The state defendants have not really opposed the Detroit Board's efforts to desegregate; they only oppose paying the cost of desegregation. As a practical matter, the state's opposition to paying one half the cost for its own segregative acts is not always clear. After agreeing to pay 50% of the construction costs for five area vocational centers, the state appealed to the Supreme Court this court's order requiring them to pay for one half of the other remedial programs. At the same time, the state voices no opposition to this court's order requiring them to provide the budgetary requirements of the court-created Monitoring Commission and has not sought an annual review of that order.

No party has ever taken the initiative in the remedial phase of these proceedings. For example, although unfavorable Monitoring Commission Reports of lagging implementation of court-ordered remedial programs provided many opportunities for initiative, the plaintiffs have failed to take any action. In addition, following each appellate court mandate, it has been necessary for this court to order a status conference to determine how the parties intended to proceed. In each instance, the court then had to order one party or the other to proceed. Now that demographic projections predict that the school district will be 87% black in 1979, the prospect that the *303 parties will take adversary initiative in future proceedings is unlikely.

With the Court of Appeals remand in hand, we convened a status conference on October 27, 1976, to determine how the parties intended to proceed. Order of October 14, 1976. Five months later, with no party having moved for a hearing, we sua sponte ordered that hearings on the issue of faculty assignments begin on April 21, 1977. The plaintiffs then filed a motion to compel the Detroit Board to submit a further pupil reassignment plan and to adjourn the faculty assignment hearings pending resolution of the reassignment issue. We declined to adjourn the faculty hearings and subsequently ordered the Detroit Board to show cause why it should not be compelled to include Regions 1, 5 and 8 in the pupil reassignment plan. Order of April 20, 1977. The show cause hearing was subsequently adjourned at plaintiffs' request until September 6, 1977. The hearings on both issues have been completed and the parties have been afforded an opportunity to argue and submit proposed findings, briefs, and responsive briefs.

Four years ago, this cause was remanded by the Supreme Court for the "prompt formulation of a decree directed to eliminating the segregation found to exist in Detroit city schools . . ." Milliken v. Bradley, 418 U.S. 717, 753, 94 S. Ct. 3112, 3131, 41 L. Ed. 2d 1069 (1974) (Milliken I). Upon receipt of that mandate, we convened a pretrial conference to determine how the parties intended to proceed. We ordered the plaintiffs and the defendant Detroit Board of Education to submit plans for desegregating the Detroit school district, and we afforded each of the parties an opportunity to critique each of the plans submitted. Following lengthy hearings, we found both plans unacceptable. We then issued guidelines for desegregation and required the Detroit Board to submit plans which conformed to those guidelines.

Under the plaintiffs' plan, any school that varied more than 15% in either direction from the system-wide racial ratio was a racially identifiable school in need of desegregation. We found the plaintiffs' plan unacceptable because it steadfastly relied upon an arbitrary ratio for defining racially identifiable schools. We also rejected plaintiffs' plan because it involved extensive pupil transportation that achieved only negligible desegregative results. Black students would be frequently transported from one identifiably black school to another for the sole purpose of achieving a predetermined racial mix. 402 F. Supp. at 1122-25. The Court of Appeals also declined to implement plaintiffs' plan. 540 F.2d at 239.

We found that the Detroit Board's plan also impermissibly sought to achieve rigid racial ratios in each school. The Detroit Board believed that schools that were 40-60% black (50-50% plus or minus 10%) were less susceptible to demographic change. We rejected the Board's plan because it paired schools located in naturally integrated neighborhoods and transported white or black children for the sole purpose of accommodating a racial pattern. We agreed with the Detroit Board, however, that an acceptable desegregation plan within the meaning of Milliken I required the elimination of racially identifiable white schools in the City of Detroit. We reasoned:

[T]he evil of segregation lies in the devastating psychological impact upon black children of the knowledge that they are being excluded from white schools. . . . However, when blacks are represented in all schools throughout the system, i. e., when white identifiable schools are eliminated, this psychological effect no longer exists. 402 F. Supp. at 1132.

We also concluded, as had the defendant Detroit Board, that Regions 1, 5 and 8 could not be included in the pupil reassignment plan because to do so would preclude meaningful interaction between black and white students in the other regions, reasoning:

*304 Plaintiffs refuse to acknowledge that the racial composition of these three regions precludes their inclusion in a desegregation plan. In Region V, for example, there are 31,354 students of whom (excluding the Spanish surnamed) only 989 (3.1%) are white. In Region VIII, there are 29,725 students of whom (excluding Spanish surnamed) only 1,329 (4.5%) are white. In Region I, there are 24,907 students of whom (excluding Spanish surnamed) only 2,049 (8.2%) are white. Clearly, it would be futile to attempt desegregation within the boundaries of these regions; thus, a desegregation plan including these three regions would have to cross regional boundaries. But to include these regions in the Board's plan would bring about the same result that pertains after application of the plaintiffs' own plan. The plaintiffs' plan itself is sufficient proof that any attempt to include these regions produces only negligible results. Application of plaintiffs' plan "would make the Detroit school system more identifiably Black, and leave many of its schools 75 to 90 per cent Black. (Findings of FactMarch 28, 1972, 484 F.2d 215, 243-44 (1973.))

That inclusion of these three regions in a desegregation plan would produce only negligible desegregative results is inevitable because even excluding Regions I, V and VIII there are only 63,446 white students as compared to 103,007 black students. To attempt to disperse those white students throughout the eight regions, including the three overwhelmingly black regions, would produce such negligible desegregative benefits that the extraordinary remedy of such cross-regional bussing would be unwarranted. To do so would only serve to lessen the little community control blacks now enjoy in those regions and, therefore, injure the very class the remedy is intended to benefit. In the face of these "practicalities" there is no constitutional objection to leaving a number of one-race or predominantly one-race schools. 402 F. Supp. at 1129 (footnotes and citations omitted) (emphasis added).

The Court of Appeals, although "recognizing the absence of alternatives," disapproved our exclusion of Regions 1, 5, and 8 and affirmed the pupil reassignment plan with respect to the remaining regions. The court characterized our finding of futility as "perfunctory treatment" holding that we "did not subject the exclusion of these three regions to the close scrutiny required by Swann [Swann v. Charlotte-Mecklenburg Bd. of Education, 402 U.S. 1, 91 S. Ct. 1267, 28 L. Ed. 2d 554]." 540 F.2d at 238. The court held that the record did not support the conclusion that the school board carried the heavy burden of justifying the one race schools located in these regions, and that it appeared we impermissibly considered these three regions in isolation. What neither the record before the Court of Appeals, nor this court's prior opinions revealed was the painstaking care with which the court attempted, with the assistance of its court-appointed experts, to devise a plan that would provide meaningful pupil reassignments in these regions. Three subsequent submissions by the Detroit Board were thoroughly examined, and we subsequently approved with modifications a Detroit Board plan which conformed to our guidelines. Bradley v. Milliken, 411 F. Supp. 943, 944 (E.D. Mich.1975). Given the small number of white students in the school district, we found it impossible to devise a plan which could bring meaningful pupil reassignments in each and every region. The Court of Appeals shared the same frustration that this court experienced in attempting to reassign students within these regions. The court stated:

We recognize that it would be appropriate for us at this point to supply guidelines to the District Judge as to what he should do under this remand. Omission of such guidelines is not based on any failure to consider the problem in depth. It is based upon the conviction which this court had at the time of its en banc opinion in this case . . . that genuine constitutional desegregation can not be accomplished within the school district boundaries of the Detroit School District. 540 F.2d at 240.

*305 Our guidelines for desegregation had provided that schools should not have a black student population less than 30%, and we treated as desegregated a school whose student body was between 30% and 55% black. In order to preserve naturally integrated neighborhoods, we determined that schools already within these parameters should not be subjected to student reassignments. We required that all other available methods of pupil reassignments be exhausted prior to relying upon transportation. We further provided that the Detroit Board should achieve a racial mix of 50%-50% in zoned middle schools and should create open enrollment middle schools similar to the magnet middle schools then in existence, which would be acceptable when 55-70% black. 402 F. Supp. at 1134-38. We also ordered remedial programs which were designed to alleviate the effects of defendants' prior acts of segregation. 402 F. Supp. at 1118. These educational components were affirmed by the Court of Appeals, 542 F.2d at 241-42, and by the Supreme Court, 433 U.S. 267, 97 S. Ct. 2749, 53 L. Ed. 2d 745 (1977) (Milliken II).

Finally, incidental to and concurrent with the implementation of our guidelines, we ordered that teachers in the Detroit school system be reassigned to achieve a distribution of not more than 70% of teachers of one race in each school. Order of August 28, 1975. The events leading to the entry of this order have been detailed elsewhere. Bradley v. Milliken, 426 F. Supp. 929, 940 (E.D.Mich.1977). The Court of Appeals vacated that order but agreed that a district court has equitable power to order the reassignment of faculty despite Judge Roth's prior finding that there had been no de jure segregation in faculty assignments by either the Detroit Board or the state defendants. 540 F.2d at 246-47. Our guidelines had acknowledged that there were other criteria for measuring a unitary system in addition to pupil assignments, such as faculty assignments, staff assignments, extracurricular activities, and equality of facilities. 402 F. Supp. at 1133.

The population shift from white to black in the district has accelerated well beyond our 1975 demographic projections. See 402 F. Supp. at 1137; see also 338 F. Supp. at 585. The official October 1976 count for the school district disclosed that it was 79.4% black. Tr. Sept. 6, 1977 at 10. By the end of the 1978-79 school year, the school district will be 87.0% black. Furthermore, the defendant Detroit Board has produced evidence, which we fully credit, that this school district will be 89.6% black by 1980 and 91.8% black by 1981. BX-C.^[2]

The Detroit Board's chief demographer, Mr. Henrickson, prepared worksheets that depict the grade-to-grade survival method he utilized to arrive at the projections contained in BX-C. See BX-D. Mr. Henrickson obtained the actual number of births for the year 1966 as recorded by the Board of Health, corrected to Detroit residency. He then tracked the actual number of children, black and white, who appeared in kindergarten five years later and then followed that group of children as they progressed from grade to grade in their schools from 1971 through 1976. Tr. at 11-23; BX-D. Those computations establish that the grade retention rate for black students is much higher than for white students. For example, 97.9% of the black students entering kindergarten in 1971 remained in the school system through the sixth grade while only 78% of the white students remained. Mr. Henrickson testified that he used those retention rate percentages to project the composition of the student body through 1981 based on the trends that he observed for the period 1971-1976. Tr. at 11-12.^[3] He did this for each grade individually, repeating the process for black and white students separately. Tr. at 13-17.

*306 We find Mr. Henrickson's method of projection logical and reliably accurate. He testified that in his experience he has been able to predict future enrollment within 100 to 200 students. Tr. at 18. This is a standard procedure used in many school districts and one which he has used in the regular course of his employment to forecast future needs of the school system. Tr. at 31.

Mr. Henrickson further testified that in each year since 1971, approximately 10,000 fewer students have enrolled in the Detroit school system. Tr. at 24.^[4] This downward trend is due principally to a reduction in the number of white students, and is expected to continue for at least the next five years. Tr. at 30.^[5] The decline in the number of white students is attributable to the declining white birth rate, and in part to white parents who take their children out of the Detroit school system. Tr. at 22-24. During this period of decreasing enrollment, the total number of black students in the system has remained reasonably constant. Tr. at 24.^[6] Considering the declining birth rate, together with the declining retention rate for white students, Mr. Henrickson projects the racial composition in future years as follows:

     YEAR   TOTAL ENROLLMENT   NUMBER/% BLACK   NUMBER/% WHITE
     1978         218,110          184,230          33,880
                                    84.5%            15.5%
     1979         208,610          181,550          27,060
                                    87%              13%
     1980         198,000          177,330          20,670
                                    89.6%            10.4%
     1981         188,170          172,820          15,350
                                    91.8%             8.2%

Thus, we are confident that Detroit will have a school system that is 91.8% black in 1981. This trend is independent of student assignments, and any changes that might be made in the pupil reassignment plan will neither accelerate nor decelerate the downward trend. Tr. at 26, 30.

In addition, pronounced demographic shifts are occurring throughout the entire city. Detroit Board Exhibits H, I and J clearly depict the constant and increasing influx of black students into every region.^[7] These exhibits are maps which divide the city into elementary school attendance areas and show the racial composition of students by residence in each school area for the years 1974, 1975 and February 1977. In 1974, there were 60 school attendance areas which were less than 30% black and 10 which were 30-55% black. BX-H. In 1975 there were 57 school attendance areas less than 30% black and 9 that were 30-55% black. BX-I. In February 1977, the number of school attendance areas that were less than 30% black declined to 39 while the number between 30-55% doubled. BX-J. Thus, in a brief three-year period, there were 21 fewer school attendance areas that were under 30% black, and 13 fewer school attendance areas less than 55% black. Many of the school attendance areas that remain under 30% black continue to integrate and steadily approach that percentage:

*307
  SCHOOL AREA   1974 % BLACK   1977 % BLACK
  Burgess             4%            24.7%
  Lodge               0%             8.1%
  Houghten            4%            11.4%
  Healey              0%            15.0%
  Gompers             0%             6.4%
  McLean              2%            16.4%
  Pulaski             2%            16.0%
  Columbus            2%            14.6%
  Wilkens             6%            14.0%
  Marquette           6%            14.1%
  Hanstien            0%             8.0%

BX-H & J.

These changing residential patterns are occurring principally in schools that border the outer limits of the city, east and west. In just two years, many of those schools reflect black enrollment that is 30-55% and over:

  SCHOOL AREA   1975 % BLACK   1977 % BLACK
  Holcomb            15.0%          35.2%
  Burt               17.0%          37.0%
  Harding            29.0%          35.2%
  Vetal              25.0%          45.9%
  Mann               27.0%          52.8%
  McCall              8.0%          33.0%
  Wayne               9.0%          33.3%
  Stellwagon         15.0%          43.5%
  Clark              11.0%          35.8%
  Emerson            62.0%          73.1%
  Cooke              38.0%          57.8%

BX-I & J; Tr. at 48-50.

Thus, the outer regions of the city are experiencing a steady increase in residential black population and a steady decline in white population. Mr. Henrickson concluded that this is the result of black families moving from Regions 1, 5, and 8 into regions further out from the center of the city. Tr. at 37; BX-E. He discounted the possibility that the increase in black enrollment could result from black families taking their children out of private schools and enrolling them in public schools. He based this opinion on his own experience and knowledge that few black families transfer from private to public schools, and on discussions with regional superintendents. Tr. at 42-43. We find that Mr. Henrickson has fully documented his judgment on this point. When this changing residential pattern is taken together with our reassignment plan, the result is that out of approximately 185 elementary schools, 141 are at least 55% black, 37 are 30-55% black and only 7 are less than 30% black. Tr. at 50; BX-K.^[8]

We conclude that as a result of the present racial ratios in the vast majority of the schools, the predominance of black students in the school district and the accelerated population shifts, no more desegregation is feasible in Regions 1, 5 and 8 collectively. These pronounced demographic shifts, first predicted by Judge Roth,^[9] and the predominance of the black student population have obliterated the optional and gerrymandered attendance zones which permitted white students to avoid attending integrated schools, the practice of bussing black students from overcrowded schools beyond a closer white school and the practice of altering feeder patterns in racially changing neighborhoods. The Supreme Court in Milliken I was unable to find from the evidence in the record below that these de jure acts of segregation found by Judge Roth had a segregative effect beyond the boundaries of the Detroit City Schools. This was so because the actions which formed the basis for Judge Roth's liability findings were basically reactions by the Detroit Board of Education to changing residential patterns. The optional and gerrymandered attendance zones served the purpose of keeping white students out of *308 schools that, in the absence of such policies, would have become, and in fact have become, integrated naturally. Judge Roth did not find, however, that these policies in turn created additional residential segregation which in turn created additional school segregation. Had he made such findings, and had such findings been supported in the record, we assume the Supreme Court would have affirmed the propriety of an interdistrict remedy.

The demographic data presented by the Detroit Board compels us to further conclude that it is the demographic residential patterns, operating independently of school assignments, together with the predominance of blacks in the school system, and not the segregative acts of the Detroit Board and the state that have resulted in the one race schools that remain in Regions 1, 5 and 8. Although a school district that retains one race schools bears a heavy burden to satisfy the court that their composition is not the result of present or past discriminatory action, Swann v. Board of Education, 402 U.S. 1, 26, 91 S. Ct. 1284, 28 L. Ed. 2d 586 (1971), that burden is satisfied by the Detroit Board's showing that such schools are the result of population shifts, or the predominance of black school children in the district as a whole, or both. See Calhoun v. Cook, 522 F.2d 717 (5th Cir.), rehearing denied 525 F.2d 1203 (5th Cir. 1975); Hobson v. Hansen, 327 F. Supp. 844 (D.D.C.1971); Tasby v. Estes, 412 F. Supp. 1192 (N.D.Texas 1975); United States v. Choctaw County Board of Education, 339 F. Supp. 901 (S.D.Ala.1971); Carr v. Montgomery County Board of Education, 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). A school-by-school examination of student enrollment by residence and as affected by student assignment make it apparent that the school district no longer discriminates against black students. That examination discloses that the schools are increasing black enrollment through natural influxes of black families into the changing neighborhoods. The Detroit Board's desegregation plan brought schools within the parameters of our guidelines, and its present assignment pattern is free from the vestiges of prior discrimination. While there may be a presumption of system-wide impact of segregative policies, Keyes v. Denver School District No. 1, 521 F.2d 465 (10th Cir. 1975), cert. denied, 423 U.S. 1066, 96 S. Ct. 806, 46 L. Ed. 2d 657 (1976), we find that the Detroit Board has overcome such presumption and established conclusively that the existence of one race schools is not the product of their past or present discriminatory acts.

In Calhoun, the court was required to determine the amount of desegregation possible in the Atlanta School District, which had become 85% black. The district court had approved a plan which achieved 30% black enrollment in every school and which left 92 of 148 schools at least 90% black. The Fifth Circuit Court of Appeals acknowledged that an 85% black student enrollment would inevitably leave a number of one race schools:

The district court also found that Atlanta's remaining one-race schools are the product of its preponderant majority of black pupils rather than a vestige of past segregation. These findings are not clearly erroneous. The aim of the Fourteenth Amendment guarantee of equal protection on which this litigation is based is to assure that state supported educational opportunity is afforded without regard to race; it is not to achieve racial integration in public schools. 522 F.2d at 719.

The court also affirmed the district court's findings that the Atlanta School Board no longer discriminated against black students, that its assignment pattern was free from the vestiges of prior discrimination and that its large number of identifiably black schools was the product of demographic factors. The court elaborated in its denial of a rehearing:

It would blink reality and authority, however, to hold the Atlanta School System to be nonunitary because further racial integration is theoretically possible and we expressly decline to do so. 525 F.2d at 1203.

*309 Similarly, in Tasby v. Estes, 412 F. Supp. 1192 (N.D.Texas 1975), the court, in devising a remedy for the Dallas Independent School District (DISD) which was 41.1% white, 44.5% black and 13.4% Mexican-American, refused to disturb schools which had become naturally integrated through residential patterns. 412 F. Supp. at 1206. The court, after considering the geographic characteristics of the district and the small number of white students remaining, approved a plan which did not disturb the South Oak Cliffs, an area of the DISD which had become 98% black. The court emphasized that its ruling did not deny the students of South Oak Cliffs a remedy, finding that its superior facilities would make it a model for the district. 412 F. Supp. at 1204. See Hobson v. Hansen, 269 F. Supp. 401, 515 (D.D.C.1967) (refusing to disturb assignment patterns because of the ten to one ratio of black to white children and granting instead additional relief which dealt with equalizing the equality of education in the district).

In Brunson v. Board of Trustees of School District No. 1, 429 F.2d 820 (4th Cir. 1970), the court affirmed the district judge's order implementing a desegregation plan developed by the Department of Health, Education and Welfare. That plan provided for disbursing white students among the schools so that each school had a white population between 5 and 17% in a school district which contained 2,408 black students and 256 white students. In a separate opinion, three judges criticized the HEW plan as one which "threatens to swallow up, rather than vindicate, the constitutional mandate for integrated schools in Clarendon District No. 1." 429 F.2d at 822 (Craven, Cir. J., concurring and dissenting). Judge Sobeloff concurring in the judgment, disagreed with the dissenting opinion. He found the fear of white flight irrelevant and criticized the theory of the dissent that majority white schools provide an optimum racial mix. That theory was premised on an assumption of white superiority which he found unacceptable in fact or law. 429 F.2d at 823-27 (Sobeloff, Cir. J., concurring).

We agree with Judge Sobeloff's criticism of the Brunson dissent and note that Judge Craven later acknowledged doubts about the validity of the approach of his dissent. Bradley v. School Board of City of Richmond, Virginia, 462 F.2d 1058, 1063 n.5 (4th Cir. 1972), aff'd 412 U.S. 92, 93 S. Ct. 239, 34 L. Ed. 2d 171 (1973). We also agree with one basic premise of Brunson; the mere scattering of a few remaining white students does not provide a remedy consistent with the fourteenth amendment as interpreted in Brown v. Board of Education, 347 U.S. 483, 74 S. Ct. 686, 98 L. Ed. 873 (1954). This was also the conclusion of the court in the two cases which most closely approximate the demographic data in this district, the school districts for Washington, D.C., and Atlanta, Georgia. The Court of Appeals has recognized this to be so in Detroit as well, in that it rejected the alternative of reassigning the remaining white students "among the predominantly black schools on a percentage basis somewhat along the lines originally proposed by plaintiffs. It is obvious that such a requirement would accomplish nothing more than token integration, and that of uncertain duration." Bradley v. Milliken, 540 F.2d 229, 239 (6th Cir. 1976).

We therefore adhere to our prior conclusion that a school system cannot be desegregated by making all schools racially identifiably black. This result inevitably occurs when further desegregation is attempted for Regions 1, 5 and 8 collectively. When racial proportions are so extreme that adequate interaction between children of both races cannot be accomplished, further desegregation is not possible and it is unwise to disturb assignment patterns which effectively desegregate the schools in other regions. Moreover, in a school district which has become increasingly black while undergoing a dramatic decline in white student enrollment, it is very important that naturally integrated neighborhoods not be disturbed. Judge Wright recognized this in the District of Columbia. Indeed, our own efforts to preserve the few remaining racially stable school attendance areas in Detroit received favorable comment by the Supreme Court:

*310 [T]he District Judge took great pains to devise a workable plan with a minimum of pupil transportation. For example, he sought carefully to eliminate burdensome transportation of Negro children to predominant Negro schools and to prevent the disruption, by massive pupil reassignment, of racially mixed schools in stable neighborhoods which had successfully undergone residential and educational change. Milliken II, 433 U.S. at 288 n.19, 97 S. Ct. at 2761.

Perhaps, the most graphic evidence that Regions 1, 5 and 8, treated collectively, cannot be further desegregated is the insignificant number of white students in the city available for that purpose. To establish a pool of white students available to further desegregate Regions 1, 5 and 8, the defendant Detroit Board submitted an exhibit which summarizes, from the data contained in the 1977 fourth Friday count, the number of schools which are less than 50% black. See BX-O. We agree with the Detroit Board that this is the relevant starting point in determining whether additional pupil reassignments can occur in Regions 1, 5 and 8. We have already accepted as desegregated any school 30-55% black and further held that a school 55% or more black is a racially identifiable school. See 402 F. Supp. at 1134-35. We deemed that guideline essential to effectively utilize the white children then available to achieve the racial ratios now existing in schools throughout the city. See p. 307, supra. To use a significantly higher figure than the 50% suggested by the Detroit Board would cause all the schools to become identifiably black, a suggestion we have repeatedly rejected. Moreover, a greater or lesser percentage starting point would materially disturb present assignment patterns and tend to increase the number of black children being bussed from one identifiably black school to another. As we stated previously, and the Sixth Circuit and Supreme Court agreed, it casts a heavy burden on black students to bus them great distances to attend another majority black school. See 402 F. Supp. 1125.

BX-O shows that there are 31 elementary schools in the district which are less than 50% black. These schools contain a total of 7,860 non-black students and 4,463 black students. BX-1 attached to Board findings. In order to assure that these schools remain desegregated, we could only reassign the excess of non-black students over black students, or 3,397. The 1977 fourth Friday count, however, shows that there are 38,259 black students in the elementary schools in Regions 1, 5 and 8, or 11 times the number of non-black students available to be reassigned among them. BX-O; Tr. at 55-64; BX-2 attached to Board proposed findings. Thus, the evidence establishes that most of the white children in the school district are being utilized in the present reassignment plan for schools which are at least 55% black, and an insignificant number of white students remain to meaningfully desegregate Regions 1, 5 and 8. To disperse these few available white children among the schools in those regions would accomplish nothing more than token integration and transport black children to other majority black schools.

The situation in the middle schools is no different. There are six schools less than 50% black containing 2,305 non-black and 1,572 black students.^[10] The difference, or only 733 non-black students, is all that is available to integrate the middle schools in Regions 1, 5 and 8, or fewer than the number of black students in 10 of the 21 middle schools in those three regions. The 21 schools have a combined black student population of 17,015 or 23 times the number of available non-black students. This predominance of black students is even more pronounced at the high school level. The only high school that is not over 50% black is Western High School in Region 2, which has 1,069 non-black and 894 black students. This provides a pool of 175 students to integrate the high schools in Regions 1, 5 and 8 which have a total of 17,783 black students, approximately 100 times the number of available non-black students. Assuming *311 the Detroit Board transferred all 175 students to Northeastern High School, the school with the smallest black population, and assigned an equal number of students from Northeastern to Western, Northeastern would have a student population of 1,210 black and 244 non-black. The maximum desegregation possible at the high school level in Regions 1, 5 and 8 would yield a high school 83.2% black, still racially identifiably black.

Having concluded that no further desegregation is possible in Regions 1, 5 and 8 collectively, we now examine all regions separately. In doing so, we note that the demographic patterns which we have described above fairly characterize every region in the district except Region 2. Black student enrollment in Region 2 has remained virtually unchanged, increasing from 60 to only 62% from 1974 to 1977. The residential trends there greatly differ from every other region. Virtually every school attendance area of Region 2 that was less than 30% black in resident student population in 1974 has remained unchanged, and, unlike the other regions, has not integrated naturally. The most extreme examples are the Higgins, Bennett, Harms, Beard, Neinas, McKinstry and Logan school attendance areas. The largest black resident population is at Logan where it was only 3.6% in February 1977, a decrease from 9%. The most extreme example is at Hunter, which was 34% black in 1974 but 1.2% black as of February 1977. BX-H, I and J. Moreover, 5 of the 6 majority white middle schools that exist in the city are located in Region 2. Western High School, located in Region 2, has a lower black enrollment than any other high school in the city.^[11]

Thus, the residential patterns which are conspicuous in every other region are simply not occurring in Region 2. The impact of this demography on the racial composition of the schools is substantial and can clearly be seen when the analysis offered by the Detroit Board to prove that it is impossible to desegregate the schools of Regions 1, 5 and 8 by pupil reassignment is applied. There are fourteen (14) elementary schools in Region 2 which are less than 50% black. These schools contain the following numbers of students:

                REGION 2
  SCHOOL    NON-BLACK    BLACK
Beard          391        133
Bennett        444        189
Harms          329        152
Higgins        354        210
Holmes         280        190
Hunter          23          0
Jefferies      379        208
Logan          357        192
Maybury        379        166
McKinstry      252        160
Neinas         366        233
McMillan       258        114
Priest         613        505
Webster        465        293
               ___        ___
             4,890      2,745

In these schools above, there are 2,145 more non-black than black students. To this may be added the excess white student population of Burton and Franklin Schools in Region 1 yielding a total of 2,459 more white students than black.^[12] Exchanging students between the majority white schools in Regions 1 and 2 and the predominantly black schools in Region 1 would produce an elementary school population in Region 1 of 5,091 black and 3,248 white, or 61.1% black. Of course, we are not suggesting that this one-for-one exchange is feasible, or should be undertaken. We have already held that all pupil reassignments must be considered in the light of the practicalities of the situation. This mathematical exercise, however, does demonstrate that the numbers alone do not pose an obstacle to further desegregation of Region 1 using students from Region 2.

*312 The Detroit Board contends that most of these non-black students in Region 2 are not available for transfer, because they are Spanish-speaking students involved in bilingual-bicultural programs in their schools.^[13] There is a large number of Spanish-surnamed students residing in Region 2, and the Detroit Board argues that it cannot control where Spanish-speaking persons choose to live. That may well be. As plaintiffs contend, however, restricting bilingual programs to one region provides Spanish families with neighborhood schools not available to other Detroit children, and is an added inducement for them to move there. We cannot approve this wholesale exclusion of Spanish-speaking students from the pupil reassignment remedy. This case began as and remains a racial desegregation case, and for school assignment purposes Spanish-surnamed students cannot be treated differently than other white students. Bilingual education is not a justification for treating one ethnic group in isolation nor is it a substitute for desegregation. Keyes v. Denver School District No. 1, 521 F.2d 465, 480 (10th Cir. 1975), cert. denied, 423 U.S. 1066, 96 S. Ct. 806, 46 L. Ed. 2d 657 (1976).

In a supplemental brief, the Detroit Board contends that its bilingual-bicultural program has not resegregated the schools in Region 2. We cannot agree. The influx of black families into Region 2 cannot occur naturally when bilingual programs for Spanish-surnamed persons are confined to one region.^[14] The Detroit Board has produced testimony to the effect that adjustment will be made to the number of students transported to the schools we have singled out in Region 2, so that all schools will reflect at least 30% black student enrollment. Tr. at 51. This argument, however, misses one crucial point. The Court of Appeals remanded this cause with instructions that we re-examine the issue of pupil reassignments in Regions 1, 5 and 8. The bilingual-bicultural program cannot provide wholesale justification for excluding Spanish surnamed students along with other white students from reassignments which may further desegregate Regions 1, 5, or 8. Bilingual programs, although essential, must be subordinated to a desegregation plan. Keyes, 521 F.2d at 480. The evidence, on the whole, compels us to conclude that the Detroit Board has not shown cause why there should not be additional pupil reassignments in Region 1, using students from Region 2.

Our August 15, 1975, guidelines recognized that desegregation of faculty and staff was an essential indicia for measuring a unitary school system. 402 F. Supp. at 1133. We believed, however, that "it would be inappropriate for this court to order any reassignment of faculty at this time" because those changes should await the court's assessment of the effects of pupil reassignments and changed grade structures on faculty assignments.^[15] Thus, we directed "the Detroit Board and the Detroit Federation of Teachers [to] begin negotiations concerning reassignments necessitated by other components of this plan." 402 F. Supp. at 1144. Thereafter, the Detroit Board and the DFT stipulated to the entry of an order, which achieved "a distribution of not more than 70% of teachers of one race in each school." Order of August 28, 1975. We must now determine whether this assignment pattern adequately desegregates faculty and staff in this school district.

The plaintiffs argue that the court must go further than the present order provides. *313 They contend that the Detroit school system "still shows patterns of racial identifiability that have never been fully dismantled," that the school system should achieve a distribution of teachers by race that reflects, where possible, the system-wide racial ratios at each grade level, and that variances should not be greater than that required to accommodate specialization of teaching functions. Tr. April 24, 1977 at 6, 13-14. To support these contentions, the plaintiffs called as a witness Dr. Robert Green, Dean of the College of Urban Development at Michigan State University. Dr. Green defined racial identifiability:

A school can be viewed as being racially identifiable when any set of selective personnel that services that school, be they teachers or pupils, when their numbers in that particular school departs significantly, or unusually from their representation in that system on a system wide basis. That's the kind of a general rule of thumb that can be used. Tr. of April 29, 1977, at 543.

Doctor Green testified that the racial identifiability of the faculty has a significant educational impact upon students, that teachers serve as adult role models for students and hence it is important that children have teachers of both races, and that students whose environment is racially isolated will grow up with narrow, restricted views of the world. Tr. at 535-36.^[16] He later explained that racial identifiability of a school is readily perceived by the community, parents and students attending that school, that a student's perception of the composition of a school's faculty as identifiably black or white affects a child's motivation to achieve, affects the general performance and conduct of black students, and finally causes black students to have a lower perception of themselves, inhibiting their educational achievement. Tr. at 1384-85, 1390-94. It is his view that students perceive majority white schools with a majority white faculty as academically superior. Tr. at 1390. Although their perceptions are not statistically accurate, students are, according to Dr. Green, generally aware of the racial composition of a school system. Tr. at 580. He concluded that students would perceive any school with a faculty which departs radically from the system wide racial ratios as one that has been given special consideration for racial reasons. Tr. at 1387.

The defendant Detroit Board called as its expert Mr. Fred Martin, Executive Director of Personnel for the Detroit School System.^[17] He testified that students viewed teachers as role models and emphasized that in a system undergoing desegregation it is desirable that students have experiences with teachers of all races. Tr. at 284, 1028-33. Mr. Martin also believed that student perceptions are important; but unlike Dr. Green, Mr. Martin was concerned with the opportunities afforded students to interact with teachers of different races. Tr. at 1163-64. Thus, at the elementary school level, Mr. Martin would not only strive to achieve a faculty that reflected the racial distribution in the system as a whole, but he would also seek to insure individual student contact with teachers of both races. Tr. at 1175. In commenting upon the Detroit Board's proposal to replace four black teachers with four white teachers at Central High School, he conceded that changing Central from 64.4% black to 60.4% black was not significant in terms of percentage change, but he viewed the addition of four white teachers who have contact with 140 *314 students as significantly contributing to the desegregative experience of those students. Tr. at 1460-61, 1469. When the system-wide faculty approached 70% of one race, Mr. Martin would then inquire into the reasons behind that racial composition and seek to change hiring policies to reverse the trend. Tr. at 755-56. Dr. Green, on the other hand, would strive to achieve faculty assignments that mirror the system-wide racial composition regardless of what that racial proportion may be.

The Detroit Federation of Teachers (DFT) called as its expert witness, Dr. John Simpkins, Dean of the College of Education at Wayne State University and an eminently qualified educator. Dr. Simpkins also agrees that teachers serve as role models for students and that it is educationally desirable that students have teachers of both races. Tr. at 1666-70. Dr. Simpkins does not agree, however, that teacher assignments must reflect the system-wide racial composition. He testified that it is desirable to integrate faculty only to the extent necessary to avoid the appearance of what he termed tokenism, and that token integration, such as one white teacher in a school with ten black teachers, should be avoided because it places disproportionate pressure on the single white teacher. Tr. at 1677. Therefore, Dr. Simpkins would approve the present faculty distribution accomplished by our August 28, 1975 order. He testified that a faculty is sufficiently desegregated when a student has one chance in three to interact with a teacher of both races, and that additional reassignments would serve only to improve faculty statistics without necessarily having a beneficial effect upon students. Tr. at 1662-63, 1683.

Dr. Simpkins clearly understands the importance of insuring student contact with teachers of both races to eliminate the effects of prior state imposed racial isolation. Commenting upon the need for black students to interact with black teachers, he said:

[Y]oungsters need to believe that they can be whatever they aspire to be. And this belief is reinforced when they see people who look like them, who are, in fact, doing those things . . . Tr. at 1667.

He testified that it was important that students interact with black and white teachers because it enhances their concept of equality, and because it is important that black students see white people in a humanitarian role. Tr. at 1668-69.

We agree with these expert witnesses that students view teachers as role models, and therefore, it is educationally desirable that students interact with teachers of both races. We do not agree, however, with Dr. Green's theory of racial identifiability as it pertains to a school system such as Detroit. He is unduly concerned that a student attending a school with a faculty which departs significantly from the system-wide racial composition would perceive that school as one that has been afforded special treatment, and would have the impression that other schools in the system have been treated similarly. Tr. at 1565-66. The most serious objection we have to Dr. Green's theory is that it treats the racial mix of a faculty in the Detroit City Schools in a vacuum. We cannot accept the suggestion that a student's perception of racial identifiability of his or her school faculty is limited to the school district's boundaries. A student attending a school on the far east side of Detroit is more likely to compare his school to closer schools in adjacent suburbs rather than to schools in the far northwest section of the city. Dr. Green's testimony is also of doubtful assistance to the court because it was based on the incorrect assumption that the court had found the faculty of schools in the district de jure segregated. Tr. at 1424, 1428, 1440.^[18] We reject the rigid concept that students would receive a desegregated experience in a school with a 90% black faculty *315 simply because the faculty was 90% black system wide. A faculty is desegregated fully only when there is adequate opportunity for students to interact with teachers of both races and teacher assignments are made on a racially neutral basis.

Neither can we fully accept the conclusions of Dr. Simpkins. Although he agreed that the present 70-30 ratio created a high probability that a student would not have contact with teachers of both races, he nevertheless accepted the current faculty as desegregated, despite his acknowledgement of the need for interaction between students and faculty of both races. Tr. at 1676-78. In support of his position, Dr. Simpkins admitted that he assumed the Detroit district was moving toward a 50-50 faculty distribution in each school, and that his conclusions might well have been different had there not already been one large scale teacher reassignment. Tr. at 1881-83. Dr. Simpkins was overly concerned that additional teacher transfers would detrimentally impact the stability of the schools by disrupting other essential faculty characteristics such as sex and age. His analysis failed to place sufficient weight on the need to afford Detroit school children a complete desegregative educational experience. Although the current faculty distribution does provide something more than token integration, Detroit students should, if at all feasible, receive more than a remedy that provides only minimal desegregation. This is particularly important in a district where the students have been racially isolated for years.

Dr. Simpkins conceded that the faculty distribution in Regions 1, 5 and 8 left something to be desired. Tr. at 1677. In fact, it appears that faculty reassignments in those regions were not made on a completely racially neutral basis, despite our order of August 28, 1975. Schools which had a substantial white student enrollment retained a predominantly white faculty, while schools with a predominantly black student enrollment have predominantly black faculties. Tr. at 2005-2019; DX-3A, 6 and 11. Thus, Regions 1, 5 and 8, which have an overwhelmingly black student enrollment, have faculty ratios that approximate 70% black, while in Region 7, more than half the schools have faculty ratios that are close to 70% white. Indeed, in Region 7, the percentage of schools with a majority white faculty increased from 72% to 88% between 1974 and 1976. Tr. at 139, 672-73. Similarly, in Region 8, the percentage of schools with a majority black faculty increased from 76% to 85% between 1975 and 1977. We cannot agree that this distribution of faculty by race provides adequate interaction between students and teachers of both races.

Because it is the Detroit Board's primary obligation to desegregate, we have examined carefully its plan to desegregate faculty and staff. That plan requires implementation in three distinct stages. To implement the first stage the Detroit Board has requested the court to reinstate the August 28, 1975, order. The second phase of the Board's plan adopts the HEW regulation which requires that faculty assignments be made so that the proportion of full-time minority classroom teachers at each school is between 75% and 125% of the proportion of such minority teachers existing on the faculty as a whole. Since the current faculty is 52.8% black, each school in the district would have a faculty between 39% and 65% black. The third and final stage of the Detroit Board's plan is quite unique in that no immediate faculty transfers are anticipated. Instead, the Detroit Board would be authorized by court order to release or add teachers to the faculty in each school in such a way that would move that school closer to the racial composition of the faculty system wide.

The specific racial proportions sought by phase two of the Detroit Board's plan are arbitrary and do not withstand scrutiny.^[19]*316 In planning for the implementation of phase two, Mr. Martin did not confer with principals of the affected schools to determine whether specific transfers would adversely impact their school. He merely examined the racial composition in each school by subject area and made a judgment as to which department, in his view, had an excess of black or white teachers. Tr. at 1108-1114. Such inflexible adherence to specific racial ratios will result in teacher transfers which do nothing more than alter faculty statistics and which bear no relationship to the ultimate purpose of achieving full desegregation of faculty and staff. For example, the Detroit Board proposes to transfer teachers from Central, Cooley, Kettering, King, McKenzie, Mumford, Murray-Wright, Northeastern, Northwestern, and Southwestern despite the fact that each of these schools is already in compliance with the proposed parameters of phase two. Cooley High School, predominantly black by student enrollment, would undergo a faculty change from 45.8% to 43.7% black. Tr. at 1305. This would occur even though Cooley High School is now within the parameters of the third phase of the Detroit Board's proposal. Thus, to the extent that the goals of phase three of the Detroit Board's proposal are valid, Cooley students are impacted detrimentally by phase two.

Phase three of the Detroit Board's proposal, however, is even more objectionable. It provides that when teachers are to be released from or added to a particular school, the Board will have sole discretion to assign teachers so that the faculty in each school more closely reflects the racial composition of the system as a whole. Tr. at 911-12. If the Board's contention that phase three is essential to afford plaintiffs an adequate remedy is correct, plaintiffs may never receive their remedy. This is so because in regions where student enrollment is declining, a teacher is usually released and the position discontinued as no longer needed. Therefore, in Regions 1, 5 and 8 where there is a pronounced decline in enrollment,^[20] most of the faculty transactions effected by phase three will involve the release of teachers. Because additional releases will also occur from normal faculty attrition such as resignations, retirements and leaves of absence, black students in Regions 1, 5 and 8 could not expect to interact with a larger number of white teachers; they could only expect to interact with a smaller number of black teachers. Since phase three is not adverse to the faculty assignment goals advanced by the plaintiffs, they contend that the court may not prevent the Board from implementing it. This may be true, but we decline to enter an order sanctioning a proposal that we think makes little or no sense.

Furthermore, the ramifications of phase three upon the current collective bargaining agreement between the Board and the DFT are incalculable. Under that agreement, releases from a particular school are based on seniority, subject to the provision that no release will occur if it causes a school faculty to exceed 70% of one race. BX-2. Phase three of the Board's proposal would effectively obliterate all seniority rights of DFT members and replace one provision of the collective bargaining agreement with Board discretion. This provision of the collective bargaining agreement has never been adjudged to have caused a segregated faculty, so that when the DFT intervened in this litigation, it did so as an innocent third party. Although we would not permit the DFT's collective bargaining agreement to stand in the way of providing plaintiffs' class with an adequate *317 remedy, neither will we unnecessarily enter a court order providing for the wholesale eradication of a provision in its collective bargaining agreement. We conclude that the Detroit Board's proposal is arbitrary, self-serving and not designed to maximize interaction between students and teachers of both races, and we reject it.

Based upon the factual record made by the parties, we find that it is necessary for the court to set guidelines to further desegregate faculty and staff for several reasons. First, while it is true that no school has a distribution of more than 70% of teachers of one race, one chance in three for students to interact with teachers of both races is not adequate. The school system can do better. The exhibits offered by the Detroit Board show particularly that in Regions 1, 5 and 8, there has been little improvement in the desegregative experience children are receiving from their teachers.^[21] The problem with the present faculty distribution is that schools with a predominance of black students also have a predominantly black faculty, while schools which were traditionally white by student enrollment have a predominantly white faculty.^[22]

Second, the decline in student enrollment and a lack of coordination in hiring policies between the central and regional boards affects the rate of change in the composition of faculty in the majority of the schools. Mr. Martin explained the method utilized by the central board when releasing teachers because of declining enrollment. The principal of the affected school makes the initial determination that a position is expendable. The teacher with the least seniority in that position is then released, unless the release of that teacher would cause the faculty composition to exceed 70% of one race, in which case the teacher of the opposite race with the least seniority is released. Tr. at 59-60. If the enrollment in a region remains stable, vacancies that arise by resignations, retirement and leaves of absence present the only opportunity for the Detroit Board to implement phase three. In those regions where enrollment is declining, however, there is a tendency to close out the position when a vacancy occurs because it is no longer needed. Thus, the Detroit Board's ability to increase the desegregative experience of Detroit students and "balance staff" as to qualifications, sex, age and experience is somewhat limited in regions experiencing a declining enrollment. Tr. at 97-98. Although the DFT contends that the long range trend is toward a more "balanced staff" in all schools, plaintiffs should not be required to wait for their remedy.

Finally, regional hiring practices are not uniform and tend to impede efforts to achieve further faculty desegregation. The central board prepares a list of qualified and eligible applicants, but each regional board determines who will be hired from the list. Although this decision is subject to approval by the central board, approval is almost always received. Tr. at 819-24. Thus, if a majority black region hires more black teachers than white, it would further impede the central board's efforts to balance staff. Tr. at 819.^[23] The hiring of more black than white teachers appears to be occurring in all regions except Region 2.

Thus, we conclude that plaintiffs are entitled to an order that will afford a greater degree of interaction between Detroit students and teachers of both races, that provides *318 adequate flexibility to account for qualifications of the teacher to teach the subject and grade level, and further permits the balancing of faculty in each school by race, teaching experience, and sex. We believe that the guidelines which we order implemented can be accomplished with minimal disruption of the present composition of faculty and staff.

The DFT argues, as it has throughout this litigation, that Judge Roth's holding that neither the Detroit Board nor the DFT engaged in de jure acts of segregation in faculty assignments precludes a remedial order reassigning faculty. The Court of Appeals rejected this argument, reasoning that faculty assignments help "to mitigate the fact that the majority of Detroit's children are left in schools that are overwhelmingly one race." 540 F.2d at 247. In Milliken II, making the same analogy, the Supreme Court said:

[I]n United States v. Montgomery County Bd. of Educ., 395 U.S. 225 [89 S. Ct. 1670, 23 L. Ed. 2d 263] (1969), the Court concerned itself not with pupil assignment, but with the desegregation of faculty and staff as part of the process of dismantling a dual system. In doing so, the Court, there speaking through Mr. Justice Black, focused on the reason for judicial concerns going beyond pupil assignment: "The dispute . . . deals with faculty and staff desegregation, a goal that we have recognized to be an important aspect of the basic task of achieving a public school system wholly free from racial discrimination." Id. at 231-232 [89 S. Ct. at 1674]. (Emphasis supplied.)

Montgomery County, therefore, stands firmly for the proposition that matters other than pupil assignment must on occasion be addressed by federal courts to eliminate the effects of prior segregation. 433 U.S. at 282-83, 97 S. Ct. at 2758 (emphasis original).

Finally, the DFT argues that the interim decision of Teamsters v. United States, 431 U.S. 324, 97 S. Ct. 1843, 52 L. Ed. 396 (1977), precludes this court from making any additional faculty assignments. In Teamsters, the Court held that seniority systems which are otherwise "bonafide" are entitled to enforcement according to their terms even if they "perpetuate" or "lock in" the effects of prior discrimination. The DFT argues that "by virtue of the decision in Teamsters, any remedy which might otherwise be directed by this court is proscribed if it impacts on this otherwise lawful seniority system." Brief of DFT at 30, 32. We think this argument lacks merit. Teamsters dealt with congressional intent in enacting Title VII of the 1964 Civil Rights Act. It did not consider constitutional violations or the power of a district court when correcting prior de jure acts of segregation. In any event, the order which we enter does not impact on otherwise lawful seniority rights, since it directs transfers which take into account seniority rights of teachers affected by the order.

The necessity to reconsider the pupil and faculty reassignment portions of our remedial decree has occasioned a reevaluation of the remaining aspect of that decree, the quality education components. The task of monitoring the implementation of court-ordered remedial programs was assigned to a court-created Monitoring Commission, a group of citizens intended to reflect the city's racial and ethnic composition. 402 F. Supp. at 1145. The Monitoring Commission has performed its mission admirably. Each participant has served faithfully and is well deserving of the respect and undying gratitude of the community and the court. They have served as the "eyes and ears" of the community and court, observing and reporting on the progress of the defendant Detroit Board's implementation of the remedial programs included in the plan. Notwithstanding the numerous reports submitted by the Commission, however, no party to this litigation has filed any pleadings to challenge in an adversary proceeding the reason for the Detroit Board's lagging implementation of the remedial programs. The silence of the parties, particularly the plaintiffs, leads the court to believe that the *319 interest of all parties in this component has been reasonably satisfied.^[24]

Although the Monitoring Commission has been a great benefit to the community and the court, its prolonged existence can have several seriously detrimental community effects as well. First, the Monitoring Commission may solidify unrealistic community aspirations that it can or will solve the difficult problems plaguing the school district. For example, we have, on another occasion, stated that this school board's principal difficulty concerns financing the cost of delivering quality desegregated educational services. 402 F. Supp. at 1119. The school board's financial crisis is as critical today as it was in 1975. It continues to face escalating educational costs at a time when student enrollment, upon which its financial resources depend, is decreasing at an alarming rate. A monitoring commission cannot solve this problem, yet it may unwittingly impede the school board's ability to achieve a solution.

Second, we are concerned that the prolonged existence of the Monitoring Commission will continue to stifle the growth of the School-Community Relations (parental involvement) component, a program we emphasized was essential to the continued vitality of a school district:

We agree with plaintiffs that an acceptable community relations plan should include provisions for school-community liaison and parental involvement. The school-community relations program should give real meaning to the decentralization anticipated by the Michigan Legislature and provide an effective vehicle for true community involvement in all the schools. To achieve maximum community participation the program should depend upon parental support; . . .. An effective community relations program must develop a partnership between the community and the schools and must cooperate with traditional groups such as parent-teacher organizations and local school advisory boards. There should be a cooperative flow of information from the school to the community and from the community to the school . . .. The school-community relations organization should receive complete encouragement, budgetary support, direct assistance and a free flow of information from school authorities. 402 F. Supp. at 1143.

A fully decentralized school-community group can, in partnership with the school board, better solve the problems facing this school district. Court-sanctioned monitoring commissions cannot actually solve problems since they are neither a policy-making body nor a super school board. If the community believes that an independent group of citizens is required to monitor the business of the school administration, it is better that such a group spring from the community rather than from a court decree. Moreover, a viable community relations program can more economically perform its own monitoring functions.

Third, prolonged court-sponsored monitoring creates the danger that the group will take on the role of advisors to elected officials charged with the responsibility of managing the affairs of the school district. For example, some Monitoring Commission reports contain recommendations based upon information collected from the very source being monitored. This is a function which is a step beyond monitoring. If recommendations and advice are to be given by a monitoring group, it is best that it come from a group that is created and supported by the community.

Fourth, court-sponsored monitoring commissions may well prompt the state defendants to relax their superintending control over the affairs of the local school board. It is the responsibility of the state defendants to monitor the steps taken to ascertain that Detroit children are receiving quality desegregated educational services. The obstacles which impede the delivery of quality education must be overcome by the state and local boards and broad based community *320 groups. We are also concerned that the Monitoring Commission's budgetary requirements now being furnished by the state will continue to increase. Those requirements now exceed $200,000 annually. This amount can be used to better educational opportunities for Detroit school children. Surely, there are more economical methods for performing the monitoring function.^[25]

Finally, we are deeply concerned that the members of the Monitoring Commission do not represent a fair cross section of the school community. Although the school district is predominantly black, the Monitoring Commission as presently constituted is predominantly white. We must be sensitive to unfavorable impressions likely to occur when a court-ordered white monitoring commission continues indefinitely to audit a predominantly black school administration, especially one that contains a remarkable number of highly competent staff members, totally sensitive to the needs of Detroit school children, black and white. Moreover, Detroit has a group of black leaders that is competent and interested in the needs of the school system. Prolonged court-sponsored monitoring could very well stifle the developing interest of black leaders. We must bear in mind that the present school administration is burdened with the task of remedying the de jure acts of segregation committed by a white board some 20 years ago, the individual members of which have long ago abandoned the city. Court-sponsored monitoring cannot aid in that task. In light of the above, we question the wisdom and necessity of retaining court-sponsored monitoring services.

Accordingly, an order will be entered consistent with this opinion.

IT IS SO ORDERED.

[1] This observation did not escape the scrutiny of the Supreme Court. See Milliken v. Bradley, 433 U.S. 267, 292, 97 S. Ct. 2749, 53 L. Ed. 2d 745 (1977) (Milliken II) (Powell, J., concurring).

[2] Detroit Board of Education Exhibits will be designated BX followed by a letter or number.

[3] If Mr. Henrickson was unable to depict a clear trend, he would use an average. Tr. at 27.

[4] The city and school population has been declining steadily since 1950. 338 F. Supp. at 585.

[5] For example, in 1971, 53.7% of the white children born five years earlier in Detroit entered kindergarten in the Detroit school system. This retention rate decreased to 51.2% in 1972, to 48.1% in 1973, to 46.9% in 1974, to 42.2% in 1975, to 40.2% in 1976 and to 39.0% in 1977. Tr. at 23.

[6] Thus, during the period from 1971-1976, black students as a percent of total enrollment increased from 65.1% to 79.4%.

[7] This pronounced demographic shift of black families is not occurring in Region 2, where there is a high concentration of Spanish surname families.

[8] Five of the seven schools under 30% black are located in Region 2. Greenfield Union in Region 6 is only 26.1% black because of the large Chaldean population attending that school. The percentage of white students at Greenfield Union has actually increased somewhat apparently as a result of further movement of Chaldean children into this area who are counted as white. Tr. at 51.

[9] Judge Roth traced the demographic history of the City of Detroit and its school district describing its change from a city 91% white in 1920 to a metropolitan area in which whites have migrated to the suburbs leaving the city composed of predominantly poor black and poor aged and white families. White families continued to leave the outer portions of the city by either moving to the suburbs or placing their children in private schools, while black families continue to move into their place from the core city.

[10] Five of these six schools are in Region 2.

[11] Western is significant because Judge Roth singled it out in his findings on liability as the school to which white students transferred during the early 1960's to escape integration.

[12] In Region 1, Burton School has 239 white students and 77 black students. Franklin School has 249 white students and 97 black students. The remaining elementary schools in Region 1 have 7,062 black students and 615 white students.

[13] The Detroit Board has always prepared its demographic data so as to reflect student enrollment, black, white, Spanish surname, and other. The Detroit Board has argued that these separate categories are essential to comply with regulations of the Department of Health, Education and Welfare.

[14] The Detroit Board is required by state law to provide bilingual education for all students requiring such services in all regions. M.C.L.A. § 380.1153(1) (1978 Supp.).

[15] Grade restructuring and other aspects of the pupil reassignment plan necessitated the transfer of 1200 teachers.

[16] Dr. Green relied upon the authoritative work of Albert Bandura, a leading authority on psychological modeling. Mr. Bandura discusses in a special chapter the impact that role models play in the personality development of children.

[17] Mr. Martin is a highly qualified staff member of the Detroit Board. His vast experience has given him an expertise on the relationship of faculty assignments to the desegregative experience of Detroit students. He has a Bachelor's Degree in Social Science, Psychology and English, a Master's Degree in Administration, Sociology and Social Science. He is presently a doctoral candidate in Education Administration and Psychology. He has been an assistant principal and teacher in the Detroit school system. He also taught courses in sociology of urban schools at Wayne State University and the University of Michigan. Tr. at 278-79.

[18] His testimony became less credible when, on cross-examination he admitted that under the theory he was advancing, the faculty at the college where he is dean would be perceived as racially identifiably black. Tr. at 1370-71.

[19] Mr. Martin testified that the HEW regulations will provide sufficient latitude to account for qualifications to teach the grade level and subject, age, sex and other criteria for a balanced faculty, but he admitted that one motivation for choosing these parameters was his belief that such faculty racial proportions were necessary under applicable regulations of HEW to secure funding under the Emergency School Aid Act. Tr. at 1456.

We have already held that in a school district undergoing desegregation, HEW regulations are not controlling. A school district is bound by constitutional precepts and not governmental regulations. Bradley v. Milliken, 432 F. Supp. 885 (E.D.Mich.1977). Arbitrary racial percentages are not designed to fit all school districts under all circumstances and they are no less undesirable merely because they have been predetermined by a governmental agency.

[20] Regions 1 and 8, for example, have experienced a 6% annual decline in student enrollment from 1974-1977, while Region 7 now 31.9% white, has experienced a decline of 9% annually. Tr. Sept. 6, 1977 at 23-28.

[21] For example, in Region 1, 33 of 43 schools had a faculty more than 60% black in 1975; in April 1977, the comparable figure was 30 of 34. In Region 5, the faculty in 27 of 34 schools was over 60% black in 1975, while the comparable figure in 1977 was 26 of 34 schools. In Region 8, 22 of 38 schools had faculties that were more than 60% black in 1975 while in 1977, 16 of 35 schools had a faculty with similar composition. See BX-10B, 4, 4A.

[22] This is true not only with faculty but also with principals and department heads as well. For example, in Region 1 when 88% of the schools had a predominantly black faculty, 57% of the schools had black principals and 66% black department heads. We agree with the plaintiffs that this cannot be the result of random assignment patterns.

[23] There is no evidence that the central board's "balance of staff" goal has been followed by regional boards in their hiring practices.

[24] Generally, monitoring commissions serve year to year and after three years of court-ordered monitoring services it is only natural that we re-evaluate its mission.

[25] For example, we could appoint a master with authority to conduct annual audits of school district records and submit a report to the parties; or the school administration could be required by court order to submit annual reports, supported by affidavit, to all interested parties.