This school desegregation case was initiated in 1970 as a class action by black students and their parents against the Grand Rapids Board of Education.1 Upon motion of the defendant, the court ordered the joinder of 11 suburban school districts to afford them a- complete hearing on all issues in the event inter-district bussing was to be decreed.2 Certain state defendants were added by a subsequent amended complaint.
A trial was held on all issues except remedy, consuming 27 trial days of hearings. Evidence was heard on the plaintiffs’ charges that the Grand Rapids School District (GRSD) was operating the Grand Rapids schools in a manner so as to perpetuate a segregated system. Alleged were specific instances of action and inaction on the part of both the GRSD and the state defendants. In an exhaustive opinion, the district judge carefully analyzed the evidence presented and concluded that there existed substantial de facto residential segregation in Grand Rapids. He further found that the imbalance existing in a portion of the city’s schools was the direct result of residential segregation which in turn resulted from “factors other than any policies or practices of the Grand Rapids Board of Education or other defendants in this lawsuit.” More specifically, the district court found: (1) that inner-city schools had been overcrowded during most of the time in question but that these schools were currently underutilized to some measurable extent as a result of the school board’s attempt to enable students of low socio-economic status to attain a broader experience through a lower student-teacher ratio, and also as a result of the reluctance by the Board to take the risk that exchanging students from the outer areas would induce “white flight” from the district, thus jeopardizing the success of the Board’s voluntary desegregation plan; (2) that the school construction program was based on a desire to build the schools “where the children were” and that decisions in this area were not motivated by any segregative intent; (3) that the actions taken by the GRSD since 1965 to achieve greater racial balance were impressive although far from complete; (4) that the closing of South High School was not improper since it had no effect whatever on the racial balance of the remaining South Middle School; (5) that apprehension of white flight was of such character that “it could not be rationally ignored by a school board charged with the constitutional duty of maintaining a unitary system” or by a school board endeavoring of its own volition to correct racial imbalance within its jurisdiction; (6) that the Creston-Cen-tral optional attendance area (the sole significant optional attendance area) was based on completely neutral criteria, no credible evidence of racial overtones or bias being presented; (7) that all special schools are integrated; (8) that there was no discriminatory use of student transfer policies and procedures; (9) that predominately black schools were of a uniform quality with respect to facilities, teachers, and equipment as compared with other schools in the system; (10) that State Senate Bill 1100 which provided that school district boundaries do not automatically expand to take in areas annexed to the city was based on a legitimate concern for a city’s tax base and *783involved no racial bias; (11) that discriminatory hiring practices in early years against blacks had been attenuated over the years, recent recruitment efforts giving preference to black applicants; (12) that faculty placement was on a discriminatory basis until 1969, a condition not yet fully corrected despite significant efforts in that regard; (13) that the discriminatory assignment of teachers did not substantially contribute to the creation of a dual school system; (14) that imbalance in the school system did not significantly contribute to segregated housing conditions; and (15) that the distinction between urban and rural areas in the reimbursement of transportation costs by the state was not related to racial concerns but was a rational distinction related to physical and geographic differences.
Based on these findings, the district court concluded that there had been no constitutional violation in the GRSD’s history of student assignments although the practices of teacher assignments had violated plaintiffs’ constitutional rights. The court went on to find that there was sufficient evidence to rebut the presumption under the doctrine of Keyes v. School District No. 1, Denver, 413 U.S. 189, 93 S.Ct. 2688, 37 L.Ed.2d 548 (1973), that by reason of the discriminatory practices concerning faculty assignments other actions of the school board were motivated by segregative intent.3
With respect to charges that the burdens involved in the Board’s voluntary master plan for desegregation fell more heavily on black students,4 the court found that any inequality was “motivated by desire to achieve greater racial balance without inducing self-defeating exodus of whites from the district.” It was further concluded that this fear was reasonable under the facts of the case and not beyond the broad discretion available to a board seeking voluntarily to achieve greater racial balance. The court finally concluded that the Board’s failure to carry through completely with its “master plan” was not a violation in itself in that uncertainty regarding the pending lawsuit justified the Board’s reluctance to invest heavily in construction that might be found to be inconsistent with a court-ordered plan.
By its order of July 18, 1973, the court enjoined further teacher and administrative assignments according to race and ordered submission of a plan to eliminate racial identifiability of schools by teaching and administrative staffs. The court retained jurisdiction of the subject matter of the lawsuit as to GRSD, but dismissed the action as to all other defendants. It granted the motion of the Michigan Education Association and of the Grand Rapids Education Association to intervene. Plaintiffs then perfected their appeal from the findings of the district court insofar as they were adverse to their contentions.5 As GRSD did not appeal, the finding of discrimination in teacher assignments is not questioned here.
While some of the issues decided by the court below are issues of law and fact, the pivotal determinations are primarily factual not to be set aside by this court unless they are found to be clearly erroneous. It is thus necessary to review the evidence and the factual findings of the court in some detail. A good portion of the plaintiffs’ evidence was *784focused upon comparatively few actions of GRSD. We deal with these below but first a general discussion of the historical background of Grand Rapids and its school system is in order.
Grand Rapids is a city lying in the southwest portion of Michigan having a 1970 population of 197,649. It is the county seat of Kent County with a population of 411,044. The Grand Rapids School District is generally coterminous with the city of Grand Rapids.
The racial composition of the Grand Rapids population has changed materially in recent years, growing from a 1% black population in 1940 to 11.3% in 1970. In the twenty years from 1950 to 1970 the black population tripled, with the vast majority concentrating within an area that began in and expanded from what is generally referred to as the central city.
Historically, the Grand Rapids school population has reflected a higher percentage of blacks than that found in the total population of the city. Thus, the non-white6 school population grew from 2.58% in 1941-42, to 6.52% in 1950-51, to 9.80% in 1954-55, to 15.20% in 1960-61, and to 21.24% in 1968-69. Finally in 1972-73 the black enrollment was 8,459 out of 32,864, representing a percentage of 25.74%. At the elementary level where the black population is greater as a result of recent birth rates the percentage of non-white enrollment increased from 7.2% in 1950 to 18% in 1960 and to 31.2% in 1972. As the district judge assessed these statistics, non-white enrollment in elementary schools increased 351% from 1950 to 1965 and increased another 31.7% from 1965 to 1970. White enrollment increased only 38% from 1950 to 1965. It actually decreased 9% from 1965 to 1970.
By 1965, 97% of the total non-white elementary school population in Grand Rapids was in attendance at the 11 inner-city schools whose non-white enrollment averaged 83%. By 1970, the percentage of black elementary school children attending these inner-city schools was down from 97% to 85% but by then these schools were on the average 89% black. The figures continued to shift until at the time of trial only 65% of all non-white students were attending the inner-city schools then having an average enrollment of 95% black students.
In the earlier years there was a heavy concentration of the black junior and senior high school population at South Junior and Senior High Schools. However, as the result of the voluntary efforts of the school board in the late 1960’s to improve the racial composition of the system7 the high school level has now been completely integrated and the racial imbalances at the junior high and middle school levels have been much improved. Even so, in 1970 South Middle School continued to be attended by more than 50% of the black students of that grade level and remained approximately 95% black.
*785The general allegations of the plaintiffs with respect to GRSD were summarized by the district court as follows:
. [B]oth by inaction and by action, the Grand Rapids Board throughout the years in question, confined black students to largely black schools, through the use, in the first instance, of a neighborhood school system in the elementary grades, through failure to avail themselves of opportunities to increase racial balance by changing boundary lines to favor that balance, by building and constructing schools or adding onto existing schools in such a way as to confine or lock in black students. Further, it is the claim of the plaintiffs that what efforts the Board has made and plans to make toward achieving greater racial balance are at the expense of the black students without a corresponding sharing of the burden by the whites, and finally that the justifications and explanations that are given by the Board are nothing but window dressing to obscure an invidious discriminatory intent on the part of the Board.
The specific actions relied on by plaintiffs to support these allegations must be examined seriatim. Plaintiffs attempted to show at trial that undercapacity in the inner-city schools was indicative of the school board’s intent to discriminate by isolating black children in these schools. There was, however, more than adequate evidence to support the district judge’s finding that there was a valid educational reason for keeping a low student-teacher ratio in the inner-city schools. In view of the age of the buildings and the special needs of the children' with a low socio-economic status, a decision to use the same “standard capacity”8 figures in the inner-city schools would probably have been improper. The questions concerning the period after the introduction of the Board’s master plan during which there was bussing out of the inner-city for desegregation, will be dealt with below.
Plaintiffs complain that school construction policy and decisions under that policy on new constructions and additions tended to lock in and confine the black school population to racially identifiable schools and to prevent their even distribution throughout the school district. The court below acknowledged that this charge was one requiring close scrutiny since such policies and actions have often been relied on to establish a segregative intent on the part of school officials. In particular, the construction of Campau school was viewed with a critical eye because the selection of that site had been controversial at the time of construction and was heavily attacked by plaintiffs at trial as having been made directly contrary to objections from a study committee reporting to the school board.
We are unable to agree with plaintiffs that the district court erred in refusing to find improper the school board’s decision to retain the concept of a neighborhood school system in meeting the needs of a growing population. As pointed out by the court, the construction site of Campau school was chosen in accordance with a University of Chicago study.9 In addition, site selection and construction for 8 of the other 10 predominantly black schools was completed before 1930 when the schools were virtually all white. Of the remaining two imbalanced schools, Madison Park was built at a time when the school was only 26% non-white and Maplewood, while constructed with a majority black enrollment in 1955, was built on a site acquired in 1870 and is now used as a federally funded “follow-through school.” 10
*786This evidence compels us to conclude that it was not error for the trial judge to find that GRSD’s construction decisions were not motivated by segregative intent. It is evident from the data before the court that schools now predominantly black are imbalanced as a result of the displacement of whites by blacks in an already established neighborhood attendance grid.11
A few instances of changes of boundary lines and feeder patterns were challenged by plaintiffs. These we feel were properly dismissed by the district court as too isolated to support charges of gerrymandering to achieve a forbidden racial discrimination. The sole instance of boundary changes to which a ' significant complaint was lodged was that of a shift of a white area from South School to Burton School, thus making South somewhat “blacker.” This instance, standing alone is not sufficient to give rise to an inference of segregative intent where there exists, as in this case, a rational explanation for the shift in terms of capacity problems at South.
Evidence of optional attendance zones was also carefully analyzed by the trial court. It was pointed out that all such options except one were merely temporary options to allow children already enrolled in a school to complete their education at that school despite a change in the attendance zone for future students. The sole exception was an option which allowed students normally assigned to Central High School to attend Crestón High School. The district court in its opinion carefully weighed the evidence relating to this option and concluded that the creation of this zone was racially neutral. In making this finding, it was observed that both schools involved were substantially racially balanced and that the students who exercised the option (about 35-40) were able to walk to Crestón and that their choice made no significant difference in the degree of racial balance at either school.
Strong criticism was made in this case because of the closing of South High School, as a part of the Board’s master plan to desegregate the schools, leaving the remaining grades as a middle school overwhelmingly black. It was argued that this was an act of de jure segregation, in that it removed a number of white high school students and thus made the composition of South Middle more black than had been the over-all composition of South Junior High and South High together. As was recognized by the district judge, however, the closing of South High School had no effect whatsoever on the composition of South Middle which remained in existence. The only effect of the closing of South High School on the racial composition of the Grand Rapids schools was to aid in completely integrating the high school level.
A review of the evidence and statistics in this case makes it clear not only that Grand Rapids was not guilty of acts of intentional segregation but that much progress has been made toward elimination of the de facto segregation resulting from housing patterns. Initially, it is noteworthy that all students in the city receive an integrated education at the high school level. • At the junior high and middle school levels, 4 of the 9 schools have a black enrollment within 15% of the overall black enrollment at the secondary level. While almost exactly 50% of the black students on the junior high and middle school levels attend South Middle School, which is 97% black, the other 50% attend majority white schools throughout the period of their secondary education. In only 3 of these *787schools (Secondary) is the black enrollment under 5%.12
On the elementary level where 26.7% of the total enrollment in 43 schools is black, there are 11 schools where black enrollment is within 15% of the overall average. Another 9 schools, while beyond this range, have black enrollments from 5% to 11.7%. Only 10 have majority black enrollment, these being the schools which, except for one, were built when they served white neighborhoods. From these figures it can be seen that 46% of the elementary age children attend schools that are from 5% to 36% black. Since all black students who do not go to elementary schools that feed into South Middle School on the 6 — 9 level will attend majority white schools from Kindergarten to grade 12, it can be seen that 24% of the black elementary school students will attend majority white schools for their entire public school experience. When it is considered that 50% of all middle school aged blacks will attend majority white schools for the rest of their public education, and that all high schools and special schools are completely integrated, the progress which the Grand Rapids system has made despite the de facto residential segregation is notably substantial. These figures demonstrate that there now exist more racially balanced schools in the Grand Rapids system than existed in the system approved by this Court in Goss v. Board of Education, 482 F.2d 1044 (6th Cir. 1973), cert. denied 414 U.S. 1171, 94 S.Ct. 933, 39 L.Ed.2d 120 (1974). It was there held that the Knoxville schools constituted a unitary system after earlier court-ordered desegregation. In Knoxville, a city with almost the same number of students as Grand Rapids but with only a 16.5% black enrollment, there were only 8 out of 64 schools meeting the standards for racial balance. In Grand Rapids, there are 19 such schools in a system of 66 as well as 10 completely integrated special schools. See Goss v. Board of Education, 340 F.Supp. 711 (E.D.Tenn.1972).
It is clear, therefore, that affirmative and voluntary action by the GRSD has effectively eliminated a not insubstantial number of the racially identifiable schools in the city.
Another relevant factor to be considered in assessing the finding below that segregation in the Grand Rapids school system is not the result of intentional acts by the school board, is the testimony of the plaintiffs’ own witness that many of the more commonly used or classic segregative techniques found in other cases were absent in Grand Rapids. These devices included intact bus-sing, bussing blacks past white schools having extra capacity to more distant black schools, widespread use of optional attendance zones, use of multiple and overlapping attendance zones, disparity between physical quality of black and white schools, constant gerrymandering of attendance zones, and discriminatory use of transfer policies. The absence of these classic techniques, coupled with the existence of sound educational reasons for other Board decisions complained of compels us to conclude on the present record that the district court’s determinations were not in error. Although it is true that this enumeration of segregative techniques is not exhaustive, it is also true that most any school board decision could be subject to criticism because of some incidental racial effect, especially when viewed with the benefit *788of hindsight. The mere fact that there is some incidental racial effect is alone insufficient to support a finding of a constitutional violation.
Our conclusion that segregation in the Grand Rapids schools is the result of housing patterns does not end the inquiry. Plaintiffs also attack the district court’s findings concerning the causes of the residential segregation and the legal effect of such causes.
First, it is argued that segregated schools themselves had a reciprocal effect, causing (or helping to cause) segregated housing patterns. There is some evidence in the record concerning the general effect that choices of schools have on housing patterns. That evidence, however, does not compel a finding that the GRSD was responsible for the residential segregation in the city of Grand Rapids as no specific testimony was given touching on such effects. Furthermore, much of the testimony centered around the fact that parents will avoid inferior schools. Yet, there was no evidence that the quality of facilities was different for black and white schools. From the evidence in the case it cannot be said that the district court’s failure to find a meaningful reciprocal effect was clearly erroneous. Although it is certainly possible for segregated schools caused by segregated housing in turn to cause a worsening of residential segregation, a stronger showing of a significant effect than that presented in this case would be required to upset the finding of the trial court.
Neither was it clearly erroneous for the district judge to find that discrimination in teacher assignments was not a substantial or significant contributing cause of school segregation. As observed by the court below, there were few black teachers in the system as a whole and the schools of predominantly black enrollment did not have predominantly black faculties. Furthermore, the schools complained of were already racially identifiable as a result of residential patterns and the resulting student enrollment and any further effects of teacher assignments were not shown to be substantial. There were always white teachers in black schools and, since 1954, at least some black teachers in white schools.
Plaintiffs also assert that discriminatory acts of other individuals and governmental agencies were the cause of Grand Rapids segregated housing and therefore of segregated schools. It is argued that these acts are sufficient to support a finding of de jure segregation. The actions relied upon include the enforcement before Shelley v. Kraemer, 334 U.S. 1, 68 S.Ct. 836, 92 L.Ed. 1161 (1947), of racially restrictive covenants, the practice of title insurance companies to except such covenants from their policies even after they become unenforceable, and the consideration in the Federal Underwriting Manual of racial segregation as a factor in keeping high real estate ratings.13
We are unable to agree with plaintiffs that these actions require a finding that GRSD is operating a dual school system and is therefore legally required to take affirmative action to improve the racial balance in the system. There is no indication that the school officials had any complicity in these allegedly segregative acts. Nor is there a basis for finding any causative relationship between the acts and residential segregation in Grand Rapids.
In Deal v. Cincinnati Board of Education, 369 F.2d 55 (6th Cir. 1966) cert. denied 389 U.S. 847, 88 S.Ct. 39, 19 L.Ed.2d 114 (1967), this Court decided that discrimination by other than school authorities cannot be relied upon as the sole basis for showing a violation by the. school board. The proper remedy in *789 Deal was recognized to be a suit against the parties so discriminating since any such discrimination is and has been vio-lative of the Constitution and applicable statutes and can be remedied in the proper forum. Those persons or entities that are said to have engaged in such illegal practices are not made parties to the present action. Nor are all of them governmental bodies whose activities could be classified as state action.14 As the Supreme Court pointed out in Swann v. Board of Education, 402 U.S. 1 at 23, 91 S.Ct. 1267, at 1279, 28 L.Ed.2d 554 (1971):
Our objective in dealing with the issues presented by these cases is to see that school authorities exclude no pupil of a racial minority from any school, directly or indirectly, on account of race; it does not and cannot embrace all the problems of racial prejudice, even when those problems contribute to disproportionate racial concentrations in some schools.
Neither does it appear that the district court’s finding that the GRSD had not improperly rescinded its master plan was erroneous. While plaintiffs submit that the Board’s failure to follow through with the plan in its entirety was a sufficient showing of segregation in and of itself, the district judge had sufficient evidence before him to find that the Board’s hesitation was justifiable. Not only was there expressed dissatisfaction from portions of the black community and a continually changing factual situation, but also there existed a real possibility that a court order in this litigation would result in the Board’s actions under the plan becoming wasted effort. In' light of the uncertainty present in the case, we are not prepared to hold that the postponement of the effectiveness of the remainder of the Board’s voluntary plan pending the outcome of the litigation was illegal. Of course, this is not to say that a complete rescission of any such plan after the termination of this suit would be unobjectionable.
It is recognized that Keyes v. School District No. 1, Denver, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973), required that a finding of segregative intent in a meaningful portion of a school district results in the shifting of the burden of proof to the school board to show that it has not had such improper intent in other areas of the district. Keyes involved findings as to various geographical areas. It did not deal with separate aspects or features of a total .system, such as teacher assignments as opposed to student assignments. Here, while there was a finding of illegal discrimination in teacher assignments, it is not clear that the placement of teachers according to race, largely corresponding to teacher preferences and under the educational rationale considered to be valid at the time, should give rise to the Keyes presumption.15 Nevertheless, the lower court found that the defendants had adequately rebutted any such presumption. We do not consider that this finding is erroneous.
It is argued by plaintiffs that the school board had sufficient knowledge and forewarning that continued operation of a neighborhood school system would result in imbalanced schools and that this alone was enough to place on *790them the affirmative duty to take action to eliminate such imbalance. Specifically, plaintiffs object to the fact that the site for Campau School was chosen according to a recommendation by the University of Chicago study over the objection of those who predicted that keeping the neighborhood system in that instance would lead to racially imbalanced schools. Also objected to was the refusal to build schools on the “fringe” areas in an effort to counteract the effect of residential housing patterns.
These objections, however, do not appear to be critical when considered in the light of other factors. Despite the warnings received, the phenomenal increase in the black population in Grand Rapids was not clearly foreseeable until the imbalances were so advanced that a far-reaching reshuffling would have been required to correct them. When it first appeared that the black population at the secondary level was becoming concentrated in a few schools, the black enrollment of those schools was still only 10-20%. No stigma would attach to this type of attendance pattern. In fact, reassigning according to race in such a situation would most likely have involved more of a stigma. By the middle sixties when the scope of the problem was recognizable, a complete “fruit basket upset” would have been required to balance the system entirely.16 Indeed it is remarkable that GRSD has accomplished as much in the way of desegregation as it has with so little turmoil. It is by no means certain that a few changes in building sites would have solved the problem before it developed. A glance at the history of a racial breakdown of the various schools shows that the schools that came to be on the “fringe” of the black population in fact often became predominantly black in the years that followed. This trend does not support plaintiffs’ theory that GRSD could have easily prevented the segregation that developed if it had not “blindly” followed a neighborhood school policy. We are convinced that there has not been an acceptable showing that such a theory of guilt by acquiescence is applicable here where the changes came so rapidly and the school board adhered to a “pure” neighborhood system without gerrymandering boundary lines to increase the segregative effect.
Plaintiffs contend that Deal v. Cincinnati Board of Education, 369 F.2d 55 (6th Cir. 1966), cert. denied 389 U.S. 847, 88 S.Ct. 39, 19 L.Ed.2d 114 (1967), is no longer good law for the proposition that “[w]hen no discrimination is shown, racial imbalance alone is no warrant for relief.” Deal followed both the 10th Circuit, Downs v. Board of Education of Kansas City, 336 F.2d 988 (10th Cir. 1964), cert. denied, 380 U.S. 914, 85 S.Ct. 898, 13 L.Ed.2d 800 (1965), and the 7th Circuit Bell v. School City of Gary, 324 F.2d 209 (7th Cir. 1963), cert. denied 377 U.S. 924, 84 S.Ct. 1223, 12 L.Ed.2d 216 (1964), in upholding the constitutionality of retaining neighborhood schools where racial imbalance has not been caused by any discrimination on the part of school officials. While there are cases in which purported neighborhood school systems have been found to be discriminatory, these cases are factually distinguishable from Deal which we deem to be controlling here. Decisions contra in other circuits we regard as being inconsistent with the Supreme Court’s rationale in Keyes.
In United States v. School District of Cook County, Illinois, 404 F.2d 1125 (7th Cir. 1968), the same court that decided Bell, upheld a finding by the trial court that the actions of the school officials were not innocent but rather included acts of purposeful discrimination. Similarly United States v. Board of School Commissioners of Indianapolis, 474 F.2d 81 (7th Cir.), cert. denied, 413 U.S. 920, 93 S.Ct. 3066, 37 L.Ed.2d 1041 (1973), *791merely involved upholding the findings of the district court as substantially supported by the evidence. It was there held that the district court had correctly stated the law and that the inference it drew concerning intent was proper and not clearly erroneous. Cisneros v. Corpus Christi Independent School District, 467 F.2d 142 (5th Cir. 1972), cert. denied 413 U.S. 920, 922, 93 S.Ct. 3052, 37 L.Ed.2d 1044 (1973), is relied on heavily by plaintiffs to support their argument that the continuation of a neighborhood school system after the appearance of racial segregation on the residential level is illegal. The problem with Cisneros is that the court “discarded] the anodyne dichotomy of classical de facto and de jure segregation,” subsequently reasserted and reaffirmed by the Supreme Court in its decision in Keyes.
The Sixth Circuit cases relied on by plaintiffs are also distinguishable. Davis v. School District of Pontiac, 443 F.2d 573 (6th Cir.), cert. denied 402 U.S. 913, 92 S.Ct. 233, 30 L.Ed.2d 186 (1971), also involved the affirmance of a lower court’s factual findings as supported by the record. While the lower court in Davis, 309 F.Supp. 734 (1970), speaks of acts of omission as being equally as serious as acts of commission, the Court of Appeals relied upon findings of an active role on the part of school officials resulting in a segregated situation. The Court cited Deal with approval. Bradley v. Milliken, 484 F.2d 215 (6th Cir. 1973), rev. in part on other grounds, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974), unlike the case at bar, involved a finding by the district court of segregative intent which this Court found not to be clearly erroneous. The same distinction applies to the case of Oliver v. Kalamazoo Board of Education, 508 F.2d 178 (6th Cir. 1974).
Brinkman v. Gilligan, 503 F.2d 684 (6th Cir. 1974), also involved a district court finding of constitutional violations. That case was not one in which racial imbalance alone was used to predicate a finding of de jure segregation. It also involved optional attendance zones and the rescission of resolutions calling for implementation of a desegregation plan. These three findings were held in th'eir cumulative effect to be enough to reflect de jure segregation.
In granting a preliminary injunction, the district court in Oliver v. School District of Kalamazoo, 346 F.Supp. 766 (W.D.Mich.) aff’d 448 F.2d 635 (6th Cir. 1971) also spoke of the equally serious effects-of de jure and de facto segregation. The case involved, however, problems of rescission of a desegregation plan. To rely upon it as rejecting the de jure-de facto distinction would in any event be futile in view of the continued recognition of the distinction by the Supreme Court in Keyes. It is to be noted that this Court in affirming the district court in Oliver, expressly refrained from approving all of its language, and that on remand that court accepted the distinction between de facto and de jure segregation. 368 F.Supp. 143 (1972) aff’d 508 F.2d 178 (6th Cir. 1974).
It is thus seen that the law imposes no affirmative duty upon school officials to correct the effects of segregation resulting from factors over which they have no control. Neither are they operating a dual system when they fail accurately to anticipate the full effect of their racially neutral retention of a neighborhood school system, absent a finding of segregative intent.
Here, we are convinced that defendant’s neighborhood school policy was in reality racially neutral and that the affirmative action taken by school officials in the late and middle sixties to improve racial balance was not an “abandonment” of the neighborhood system so as to preclude them from relying upon it as a valid justification for their position.17
*792The critical issue in this case, after distilling from it related factual questions and incidental legal issues, is that of the proper interpretation of Keyes regarding the intent required for a finding of de jure segregation. It must be determined whether the district court applied the correct legal standard in concluding that the Grand Rapids School District was not in violation of constitutional mandates in the operation of its school system.
The early desegration cases for the most part dealt with school systems where statutes had mandated segregation of students by race. In consequence, it was not necessary to deal with the quantum of state action required to establish a constitutional violation. In Swann, the criteria for finding de jure segregation were specifically left open. It was recognized that school boards were not responsible for all forms and results of racial segregation. As there stated:
The elimination of racial discrimination in public schools is a large task and one that should not be retarded by efforts to achieve broader purposes lying beyond the jurisdiction of school authorities. One vehicle can carry only a limited amount of baggage. It would not serve the important objective of Brown I to seek to use school desegration cases for purposes beyond their scope, although desegregation of schools ultimately will have impact on other forms of discrimination. We do not reach in this case the question whether a showing that school segregation is a consequence of other types of state action, without any discriminatory action by the school authorities, is a constitutional violation requiring remedial action by a school desegregation decree. This case does not present that question and we therefore do not decide it.
Our objective in dealing with the issues presented by these cases is to see that school authorities exclude no pupil of a racial minority from any school, directly or indirectly, on account of race; it does not and cannot embrace all the problems of racial prejudice, even when those problems contribute to disproportionate racial concentrations in some schools. 402 U.S. 1, at 22-23, 91 S.Ct. 1267, at 1279, 28 L.Ed.2d 554.
Whatever the formula for finding a violation, it is clear that the courts’ remedial power “may be exercised only on the basis of a constitutional violation.” Swann at 16, 91 S.Ct. at 1276. ' “Absent a constitutional violation, there would be no basis for judicially ordering assignment of students on a racial basis.” Swann at 28, 91 S.Ct. at 1282.
In Keyes, the Supreme Court again recognized the distinction between de jure and de facto segregation, de jure segregation requiring a showing that it was brought about or maintained by “intentional state action.” Keyes, 413 U.S. at 198, 93 S.Ct. 2686. Although the Keyes case, in which certain intentional acts were found to exist, did not require a holding concerning a “pure” neighborhood school system; the Court outlined two ways that a school district can show itself not in violation even if it had the burden of proof to rebut a presumption against it: A school board can rebut a prima facie case by either showing that “segregative intent was not among the factors that motivated [its] actions,” or that “its past segregative acts did not create or contribute to the current segregated condition.” Keyes at 210 — 211, 93 S.Ct. at 2698-2699. So, it is clear that the Supreme Court requires at least a finding of segregative intent. It has gone even further to suggest that there is no violation even if there was segregative intent where intentional segregative action has become so attenuated as to be insufficient to support a finding of de jure segregation. • The standard expressed is “a systematic program of segregation affecting a substantial portion of the students, schools, teachers, and facilities within the school system.” Keyes at 201, 93 S.Ct. at 2694.
The Supreme Court in Keyes and Swann has not expressed a standard in*793consistent with that enunciated by this Court in Deal:
Appellants’ right to relief depends on a showing of more than mere statistical imbalance in the Cincinnati schools. They must also expose that added quantum of discriminatory state action which deprives them of their constitutional right to freedom of choice. If the school officials, through overt practice or by subterfuge, have treated students differently solely because of race, then they not only must cease doing so, but also must take affirmative action to remedy the condition which they have caused. Thus, even if the Negro students were distributed uniformly in the schools, if other forms of discrimination were used against them they would still be entitled to the aid of the law. When no discrimination is shown, racial imbalance alone is no warrant for relief. 369 F.2d at 63.
We are convinced that the district court did not misapply this standard. It did not require, as plaintiffs charge, a showing of an “evil, malevolent, malicious and subjective intent of the particular person in charge of public schooling.” Rather it merely required that the evidence support a finding of purposeful segregation. While it is true that a court may infer such an intent from the circumstances there is no authority for the proposition that such an intent must be inferred in all cases where segregated patterns exist in fact. The inference is permissible, not mandatory. Upon review of the evidence, we conclude that the district court applied the proper legal standard to its findings which were supported by substantial evidence. It follows that its conclusion that the school officials of Grand Rapids were not guilty of purposeful segregation of students and therefore were not guilty of de jure segregation by operating a dual school system must be upheld.
We must go a step further. Plaintiffs insist that even if it is found that there was nothing illegal done by the school board prior to the time it adopted an affirmative desegregation plan, the implementation of the plan was itself illegally discriminatory. It is argued that by utilizing a general scheme of bussing one-way from the central city to the peripheral areas, the school board has unjustifiably placed a greater burden on black children than white children since the greatest density of black population is in the core city. Plaintiffs submit that the school board by considering the fact that massive bussing might result in significant “white flight” to the suburbs, was merely catering to the fears and prejudices of whites. The fact that such an accommodation of white fears is necessary is said to be an indication that black children bear a stigma of inferiority regarding their schools.
Such a charge is to be closely scrutinized. The burdens and inconveniences of integration should not be placed discriminatorily. Unequal burdens should not be borne by different groups unless a compelling justification presents itself. Yet, the authority of school officials to formulate plans for achieving an improved racial balance should not be as restrictive in the case of a school system which has not been found to have engaged in purposeful segregation as for a system which has practiced de jure segregation. The Supreme Court in Swann recognized the broad discretion which lies with school authorities in implementing educational policies of increased integration:
School authorities are traditionally charged with broad power to formulate and implement educational policy and might well conclude, for example, that in order to prepare students to live in a pluralistic society each school should have a prescribed ratio of Negro to white students reflecting the proportion for the district as a whole. To do this as an educational policy is within the broad discretionary powers of school authorities; absent a finding of a constitutional violation, however, that would not be within the authority of a federal court. 402 U.S. at 16, 91 S.Ct. at 1276.
*794It is settled that the possibility of white flight from the school district will not justify a failure to comply with a court decree ordering an integration plan nor will it justify actions which would make such a decree unworkable. Wright v. Council of City of Emporia, 407 U.S. 451, 92 S.Ct. 2196, 33 L.Ed.2d 51 (1972); United States v. Scotland Neck Board of Education, 407 U.S. 484, 92 S.Ct. 2214, 33 L.Ed.2d 75 (1972); Monroe v. Board of Commissioners of the City of Jackson, 391 U.S. 450, 459, 88 S.Ct. 1700, 20 L.Ed.2d 733 (1968). However, it does not follow that a board must ignore the probability of white flight in attempting to formulate a voluntary plan which would improve the racial balance in the schools without at the same time losing the support and acceptance of the public. As the district court recognized, there is a valid distinction between using the defense of white flight as a smokescreen to avoid integration and realistically considering and dealing with the practical problems involved in making voluntary efforts to achieve integration.
The district judge, after reviewing the evidence before him, found that the school board’s concern over white flight was justified, and that the actions taken which do not impose precisely equal burdens on all portions of the population, were “motivated by a desire to achieve greater racial balance without inducing self-defeating exodus of whites from the district. . . . ”
This is not to say that threats of whites leaving the district would justify catering to white fears by placing the entire burden of a desegregation plan on black students. Certainly all reasonable attempts must be made to equalize the inconveniences among various segments of the population. However, such concerns over the equal sharing of burdens should not overshadow the primary consideration of devising a desegregation plan that is effective in improving racial balance while retaining the necessary cooperation from the community. The GRSD officials who voluntarily took upon themselves the affirmative burden of improving the educational system by increasing the racial balance in the schools are in a much better position than the courts to ascertain the best method to eliminate de facto segregation without otherwise jeopardizing the educational quality of the schools. The record in this case does not show that school officials in Grand Rapids abused their discretion as they might have if they had placed all burdens and inconveniences on a single portion of the population. Rather, the evidence shows that the course adopted by GRSD was a reasonable and equitable attempt to improve the quality of education for all students.
Although the general concept of the GRSD Master Plan involved bussing from the core city to^the periphery, the implementation of this plan has not involved a singularly one-way burden as it has been applied to date. The testimony showed that only around 4000 of the over 33,000 students in the system were being bussed for any purpose and that a majority of these were white. White children have been bussed into what has been perceived as the inner city in the cases of Iroquois and Morris Primary Schools. In the closing of South High School in order to desegregate the high schools of the city, the majority of those reassigned to other schools were white. In addition, there were a good number of white students residing in the inner city who were thus also subject to being bussed to outlying schools. Consequently, while the impact of the desegregation plan is admittedly greater on black children, it can by no means be said that the entire burden has fallen on the blacks or that the inconveniences of that burden have been unreasonable.
The inequality of burdens that does exist has not been adequately demonstrated to be outside the school board’s sound discretion. Its decisions can be supported by valid reasons other than a concern over possible white flight. The evidence suggests that at least a part of the reason for moving more of the school population toward the periphery is a de*795sire to broaden the education of those children of low socio-economic status, regardless of race. Furthermore, it is only natural that a program of bussing for desegregation purposes will affect those in the compacted inner-city more than those in the suburbs. The latter are in more sparsely populated areas with larger “neighborhoods” and are thus more likely to have been bussed already for simple transportation reasons. Conceding that it is possible that a discriminatory distribution of hardships may rise to the scope of a constitutional violation, we are unable to agree that the GRSD has yet exceeded the limits of its permissible discretion. Neither do we anticipate that the total master plan as originally proposed would be so inequitable as to result in such a violation when fully implemented. The district court’s retention of jurisdiction would allow it to pass upon such a question in the future if the occasion should arise.
The district court properly observed: On the whole, there is too much evidence of good faith effort and too much potential for ultimate benefit to plaintiffs’ class to warrant the court’s finding that the efforts of the Grand Rapids Board to achieve racial balance have been constitutionally violative of plaintiffs’ rights.
An integrated school experience is too important to the nation’s children for this Court to jeopardize the opportunity for such an experience by constructing obstacles that would discourage school officials from voluntarily undertaking creative programs.
In their brief, the plaintiffs argue that the State of Michigan was involved in and ultimately responsible for the alleged de jure segregation practiced by the GRSD. In an effort to prove this charge, the plaintiffs catalogued several factors indicating state involvement. First, they listed the various obligations of the state to its educational system as imposed by the Michigan legislature. Secondly, the plaintiffs noted the duties that adhered to the state as a disburser of federal funds under Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d. Next the plaintiffs referred to the state’s obligation to uphold the Michigan and United States Constitutions. In addition to these obligations of general involvement, the plaintiffs claimed two statutes had the effect of increasing segregation in the city of Grand Rapids. One statute makes the annexation and property transfer provisions of the school code applicable to second class school districts like GRSD. M.C.L.A. § 340.146. The other statute provides for reimbursement by the state to rural but not urban school districts for the expenses of student transportation. M.C.L.A. § 388.-1171. Also the plaintiffs relied on previous findings of state involvement made by courts in other desegregation suits in Michigan.18
It is maintained that findings of state involvement in other Michigan cities should serve to shift the burden to the state to refute charges of illegal segregative involvement in the Grand Rapids System.19 In effect, the plaintiffs urge this Court to find that the state and its officers have been found to practice an extensive policy of segregation and therefore must carry the burden of proof concerning specific charges raised in this particular suit. This argument relies on the burden-shifting principle expressed by the Supreme Court in Keyes where the court held that:
a finding of intentionally segregative school board actions in a meaningful portion of a school system, as in this case, creates a presumption that other segregated schooling within the system is not adventitious. It establishes, *796in 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.
However, in the Keyes case, this burden-shifting technique was applied only to a single school district and the different geographic areas within it. This Court is not persuaded that the Keyes decision can or should be construed so expansively as the plaintiffs contend.
If the plaintiffs’ theory were upheld, a finding that the state was involved in creating a system of de jure segregation in a single city would shift the burden to the state to show that it was innocent of such practices in every other city of the state. This would completely ignore the local autonomy and individuality of the. various school districts as recognized in Milliken. The responsibility of a school district to desegregate should not be made to depend on the complicity of the state board of education in the de jure segregation of another city.
Nevertheless, we must consider the effect in Grand Rapids itself of the plaintiffs’ specific allegations of state involvement resulting from the two statutes dealing with annexation and reimbursement. Under the prior law, any area annexed by a city comprising a second class school district automatically became part of that school district. This procedure was changed by Senate Bill 1100 which required that the territory to be annexed approve the proposed consolidation.20 The plaintiffs contend that this was a deliberate effort by the state and some of the suburban school district defendants to prevent the desegregation of the white suburban schools and to lock the blacks into the Grand Rapids inner-city schools. The district court studied the transcript of the legislative hearing on the Bill, and it concluded that the real motivation for the change was the concern of the suburban districts that they would lose “their tax base by piecemeal annexation under existing municipal annexation procedures.”
The second statute alleged to be part of the overall state action provides for the reimbursement of transportation funds. The law draws a distinction between the needs of the rural and urban areas. The rural districts are reimbursed for the school transportation because the areas are more sparsely populated, encompass greater distances and lack sidewalks and public conveyances. According to the district court, the statute is a valid exercise of legislative discretion, and the plaintiffs direct us to no substantial evidence to the contrary. In summary, the plaintiffs have not introduced convincing evidence to support a finding that the State of Michigan contributed to and is responsible for any act of de jure segregation in Grand Rapids.
We are not bound to conclude otherwise by the finding of this Court in Mil-liken that the State of Michigan was involved in creating an illegally segregated system in Detroit. Similar actions by the state can be found to be a violation in one city and innocent in another city, depending on whether the local board was at the same time using the state’s policies to further a program of purposeful segregation.
Plaintiffs’ allegations of state involvement also are pertinent to the question of inter-district remedy. In the recent decision of Milliken v. Bradley, 418 U.S. 717, 94 S.Ct. 3112, 41 L.Ed.2d 1069 (1974), the Supreme Court considered the grounds for imposition of an inter-district remedy, and it concluded that there must be a causal connection among the districts involved before inter-district relief could be employed.
[I]t must first be shown that there has been a constitutional violation within one district that produces a significant segregative effect in another district. Specifically it must be shown that racially discriminatory acts of the state *797or local school districts, or of a single school district have been a substantial cause of inter-district segregation. [Without an inter-district effect, there is no constitutional wrong calling for an inter-district remedy. [Emphasis added] 418 U.S. 719, 94 S.Ct. 3115.
Our finding that the Grand Rapids district did not engage in purposeful segregation of students renders to a large extent moot the question of inter-district or metropolitan remedy. We further hold that the violation concerning teacher assignments does not in any sense require an inter-district remedy. We conclude that there has been no showing of a constitutional violation of any kind in the outlying districts and that a metropolitan remedy would therefore be completely inappropriate under the standards of Milliken.
We affirm the order and judgment of the district court of July 18, 1973, but remand the action to that court for any further proceedings that may be necessary pursuant to its retention of jurisdiction.