This is a school desegregation case involving the public schools of the city of Youngstown, Ohio. It was certified as a class action on behalf of all children attending those schools and their parents or guardians. The complaint alleges that the Youngstown Board of Education, its individual members and Superintendent (the "Youngstown defendants"), along with the Ohio State Board of Education, its individual members, and the Superintendent of Public Instruction of the State of Ohio (the "State defendants"),1 created, maintained, and were presently operating an illegally segregated system of public schools, thereby depriving the plaintiffs of their right to equal protection of the laws as guaranteed by the Fourteenth Amendment to the United States Constitution. The segregation was allegedly accomplished by the use of various techniques, including the assignment of students to schools, the manipulation of attendance zones and feeder patterns, schoolsite selection, and the utilization of existing racially discriminatory patterns in public and private housing. The complaint sought equitable relief in the form of an injunction requiring the defendants to develop and implement a systemwide plan of desegregation. Trial of the case commenced on January 3, 1977 and continued for 25 days, during which 31 witnesses testified and more than 5,000 pages of testimony were transcribed. After the trial was completed but before the District Court handed down its decision, the United States Supreme Court rendered its opinion in Dayton Board of Education v. Brinkman (Dayton I), 433 U.S. 406, 97 S. Ct. 2766, 53 L. Ed. 2d 851 (1977). The District Court directed the parties to file additional briefs in light of this decision. On April 12, 1978, the District Court issued an exhaustive Memorandum Opinion and Order, published at 454 F. Supp. 985 (N.D.Ohio 1977). In that opinion, it concluded that the defendants had not at any time referred to in the complaint or at the time of trial operated a dual or intentionally segregated school system. The court did find, however, that the Youngstown defendants had violated plaintiffs' rights under the Equal Protection Clause by the disproportionate assignment of black teachers and administrators to predominantly black schools.2 The court expressly found that those impermissible assignment practices had not, alone or in combination with other practices, created a segregated or dual school system. The court further held that the plaintiffs failed to establish any liability on the part of the State defendants.

The District Court certified its decision for immediate interlocutory appeal pursuant to 28 U.S.C. § 1292(b) and plaintiff's Petition for Leave to Appeal was granted by this court.3

The Youngstown City School District is substantially coterminous with the city of Youngstown. At the time of trial, it operated 41 schools to service a student population of approximately 20,400. The percentage of black students in the district rose steadily in the years preceding the trial. In 1952-53, the first year for which racial composition data is available, blacks comprised less than 20% of the student population; at the time of trial in 1977, the student population was 48.9% black. The district's schools are characterized by a substantial degree of racial identifiability. The question in this case, as in most school desegregation cases, is whether that segregation is the result of a violation by the defendants of the plaintiffs' constitutional rights. The District Court held that it was not, and we affirm.

The fact that Ohio law does not mandate segregated schooling does not preclude a finding of unlawful segregation.4 "(T)he Equal Protection Clause was aimed at all official actions, not just those of state legislatures." Columbus Board of Education v. Penick, 443 U.S. 449, 457 n.5, 99 S. Ct. 2941, 2946 n.5, 61 L. Ed. 2d 666 (1979). If a school system was intentionally segregated at the time of Brown I, (Board of Education of Topeka, Shawnee County, Kansas ), 347 U.S. 483, 74 S. Ct. 686, 98 L. Ed. 873 (1954), the school officials are under a continuous constitutional obligation to disestablish the dual system as of the date of the Supreme Court's decision in Brown II (Board of Education of Topeka, Shawnee County, Kansas ), 349 U.S. 294, 75 S. Ct. 753, 99 L. Ed. 1083 (1955). Columbus Board of Education v. Penick, supra, at 458, 99 S.Ct. at 2946. This obligation imposes an "affirmative duty on school boards to take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch." Green v. County School Board, 391 U.S. 430, 437, 88 S. Ct. 1689, 1693, 20 L. Ed. 2d 716 (1968). "Each instance of a failure or refusal to fulfill this affirmative duty continues the violation of the Fourteenth Amendment." Penick, supra at 459, 99 S. Ct. at 2947.

On the other hand, the mere existence of unintegrated schools does not establish a constitutional violation; "(A) school district has no affirmative obligation to achieve a balance of the races in the schools when the existing imbalance is not attributable to school policies or practices and is the result of housing patterns and other forces over which the school administration had no control." Davis v. School District of City of Pontiac, Inc., 443 F.2d 573, 575 (6th Cir.), cert. denied, 404 U.S. 913, 92 S. Ct. 233, 30 L. Ed. 2d 186 (1971). As the Supreme Court wrote in Washington v. Davis, 426 U.S. 229, 239, 96 S. Ct. 2040, 2047, 48 L. Ed. 2d 597 (1976):

(O)ur cases have not embraced the proposition that a law or other official act, without regard to whether it reflects a racially discriminatory purpose, is unconstitutional solely because it has a racially disproportionate impact.

....

The school desegregation cases have also adhered to the basic equal protection principle that the invidious quality of a law claimed to be racially discriminatory must ultimately be traced to a racially discriminatory purpose. That there are both predominantly black and predominantly white schools in a community is not alone violative of the Equal Protection Clause. The essential element of de jure segregation is "a current condition of segregation resulting from intentional state action." Keyes v. School Dist. No. 1, 413 U.S. 189, 205 (93 S. Ct. 2686, 2696, 37 L. Ed. 2d 548) (1973). "The differentiating factor between de jure segregation and so-called de facto segregation ... is purpose or intent to segregate." Id., at 208 (93 S.Ct. at 2697). See also id., at 199, 211, 213 (93 S.Ct. at 2699.)

It is now well established that there are three prerequisites to a finding of de jure segregation: (1) school board action; (2) a segregative intent or purpose; and (3) a causal relationship between the official conduct and the segregation in the schools. See Keyes v. School District No. 1, Denver, Colorado, 413 U.S. 189, 93 S. Ct. 2686, 37 L. Ed. 2d 548 (1973). Plaintiffs are required to bear the burden of establishing discriminatory intent in a meaningful portion of the district.5 As a practical matter, intent can only be proven circumstantially. This places upon the trial court the critical responsibility of deciding whether or not such intent has been established. Trial courts are not without guidance in this effort. Although Washington v. Davis, supra, makes it clear that official action will not be held unconstitutional solely because it results in a racially discriminatory impact, "Davis does not require a plaintiff to prove that the challenged action rested solely on racially discriminatory purposes." Arlington Heights v. Metropolitan Housing Corp., 429 U.S. 252, 265, 97 S. Ct. 555, 563, 50 L. Ed. 2d 450 (1977). Thus plaintiffs need only prove that racial discrimination has been a motivating factor in the decision. Id. at 265-66, 97 S.Ct. at 563. In addition, while actions resulting in foreseeable and anticipated disparate impact do not necessarily establish a constitutional violation, such actions are relevant evidence to prove the forbidden purpose: "Adherence to a particular policy or practice, with full knowledge of the predictable effects of such adherence upon racial imbalance in a school system is one factor among many others which may be considered by a court in determining whether an inference of segregative intent should be drawn." Columbus Board of Education v. Penick, 433 U.S. 449, 465, 99 S. Ct. 2941, 2950, 61 L. Ed. 2d 666 (1979), quoting from Penick v. Columbus Board of Education, 429 F. Supp. 229, 255 (S.D.Ohio 1977).

It is thus the duty of the District Court to determine whether, on the facts of a particular case, an inference of segregative intent should be drawn. The impact of the challenged official conduct is an important starting point. Other evidentiary sources available to the fact-finder include the historical background of that conduct, the specific sequence of events leading up to it, and the administrative record, particularly where there are contemporaneous statements by members of the decision-making body, minutes of its meetings, or reports. Arlington Heights, supra, at 267-68, 97 S.Ct. at 564-65.

The task of fact-finding in a case such as this is substantially more complex than in a typical lawsuit. See Dayton Board of Education v. Brinkman (Dayton I), 433 U.S. 406, 414, 97 S. Ct. 2766, 2772, 53 L. Ed. 2d 851 (1977). It is an inherently difficult task to ascertain the motivations of multimembered public bodies. The question whether a racial imbalance in a school system resulted from purely neutral public actions or is instead the intended result of actions which appeared neutral on their face but were in fact invidiously discriminatory is not an easy one to resolve. It requires careful consideration by the District Court of a multitude of official actions, their effects, and the circumstance in which they were undertaken.

Appellants advance a number of arguments on appeal. They contend that the District Court failed to apply the correct legal principles to the facts in its decision of this case. With respect to the issue of segregative intent, they claim that the court erroneously required plaintiffs to demonstrate that racial animus or bad faith was the sole motivating factor behind a myriad of administrative decisions. Appellants further contend that the trial court failed to consider the legal relevance of the defendants' pre-Brown conduct. They argue that the court was "duty-bound" to hold the defendants to an obligation, commencing in 1954 and continuing thereafter, to effectuate a transition to a racially non-discriminatory school system. Appellants also claim that the District Court erred in failing to invoke the burden shifting principles announced in Keyes, and misallocated the burden of proof to appellants' detriment. Finally, they argue that the District Court erroneously treated Youngstown's six high school zones as separate and impenetrable subsystems in its examination of the evidence. We disagree with each of these contentions, and find that the District Court was sufficiently cognizant of the controlling cases and legal principles.

The District Court's opinion evidences a full awareness of the principles governing the issue of segregative intent. Contrary to appellants' contention, it did not require a demonstration that racial animus or bad faith was the sole motivating factor behind the allegedly segregative actions of the school officials. Indeed, it stated:

Only in exceptional circumstances is a decision of such authorities motivated by a single consideration or dominated by one particular purpose. Arlington Heights v. Metropolitan Housing Development, 429 U.S. 252, 265, 97 S. Ct. 555, 563 (50 L. Ed. 2d 450) (1977). However, a racially discriminatory purpose need not be the primary motivating factor in a decision to establish the element of intent; it need only be a factor. Id. at 265-66, 97 S.Ct. at 563.

454 F.Supp. at 988. The court acknowledged its authority to infer discriminatory intent from acts or policies which had the natural, probable, or foreseeable result of increasing or perpetuating school segregation. It was also aware, however, that such an inference is permissible rather than mandatory. See Higgins v. Board of Education, 508 F.2d 779, 793 (6th Cir. 1974). This is obviously consistent with the Supreme Court's decision in Columbus : "(Keyes, Washington v. Davis, and Arlington Heights ) do not forbid 'the foreseeable effects standard from being utilized as one of the several kinds of proofs from which an inference of segregative intent may properly be drawn.' " 443 U.S. at 464-65, 99 S.Ct. at 2949-50, quoting from Penick v. Columbus Board of Education, 429 F. Supp. 229, 255 (S.D.Ohio 1977), affirmed in part and vacated in part, 583 F.2d 787 (6th Cir. 1978), affirmed, 443 U.S. 449, 99 S. Ct. 2941, 61 L. Ed. 2d 666 (1979) (emphasis added). Throughout its opinion, the District Court discusses school board actions and surrounding circumstances which support an inference of segregative intent. In every instance, however, the court declined to draw that inference, setting forth its reasons in its opinion.6 Nowhere in that opinion do we find evidence that the court failed to apply the correct legal rules governing the issue of intent.

We also disagree with appellants' claim that the District Court failed to consider the legal relevance of the defendants' pre-Brown conduct. The court was required to hold the defendants to a continuing obligation to dismantle a segregated school system only if it found that an intentionally segregated school system had been established at the time of Brown I. Having specifically found that no such intentional segregation had been proved, the trial court was under no duty to impose on the defendants a constitutional obligation to establish a unitary system.

Nor was the trial court obligated to invoke the burden-shifting principle set out in Keyes, supra. That principle is the allocation to the school board of the burden of proving either that its actions were "not taken in effectuation of a policy to create or maintain segregation ..., or, if unsuccessful in that effort, were not factors in causing the existing condition of segregation in these schools." 413 U.S. at 214, 93 S.Ct. at 2700. However, this burden can only be shifted when the court finds that school authorities have practiced purposeful segregation in part of the school system. Id. at 208, 93 S.Ct. at 2697. See also Penick v. Columbus Board of Education, supra, 429 F.Supp. at 252. No such finding was made in this case. Thus, unless the trial court's findings with regard to segregative intent are clearly erroneous, neither the Keyes burden-shifting principle nor the constitutional obligation to establish a unitary school system plays a part in this case.

We do observe that the District Court misstated the Keyes burden-shifting principle. That principle is properly invoked upon a finding of intentionally segregative school board actions; the District Court, however, stated that the burden of proof shifts to the defendants where the evidence supports such a finding. 454 F. Supp. at 989. Whether the court actually placed the burden on the defendants is not clear from its opinion. Neither is it significant to this appeal in light of our decision. If the court in fact shifted the burden, the error could only have operated in appellants' favor.

With reference to the burden of proof, the District Court also cited this court's opinion in Oliver v. Michigan Board of Education, 508 F.2d 178 (6th Cir. 1974), cert. denied, 421 U.S. 963, 95 S. Ct. 1950, 44 L. Ed. 2d 449 (1975). In that case, we held that a presumption of segregative purpose arises when plaintiffs establish that the natural, probable, and foreseeable result of public officials' action or inaction was an increase or maintenance of segregation. According to Oliver, the presumption becomes proof unless defendants affirmatively establish that their conduct was a consistent and resolute application of racially neutral policies. Id. at 182. In other words, the presumption shifted the burden of proof to the defendant to demonstrate that its policies have been racially nondiscriminatory. The validity of this presumption was cast in doubt by the Supreme Court in Dayton II :

We have never held that as a general proposition the foreseeability of segregative consequences makes out a prima facie case of purposeful racial discrimination and shifts the burden of producing evidence to the defendants if they are to escape judgment; and even more clearly there is no warrant in our cases for holding that such foreseeability routinely shifts the burden of persuasion to the defendants. Of course, as we hold in Columbus today, ... proof of foreseeable consequences is one type of quite relevant evidence of racially discriminatory purpose, and it may itself show a failure to fulfill the duty to eradicate the consequences of prior purposefully discriminatory conduct.

443 U.S. at 536, n.9, 99 S. Ct. at 2978, n.9. See also Columbus Board of Education v. Penick, 443 U.S. at 465, 99 S.Ct. at 2950. Again, however, any erroneous invocation of the Oliver presumption by the District Court would have aided appellants below by divesting them of their burden of proof. Such an error, if it in fact occurred, can therefore have no bearing on our decision.

Finally, we disagree with appellants' contention that the District Court erroneously treated the six high school zones as discrete subsystems. Throughout its opinion, the District Court makes reference to a document, published in 1921, entitled "A Survey and Building Program for the Youngstown City Schools 1920-1940." Authorized by the Board of Education, this study was designed to provide "a thorough-going survey on which the building program of the next twenty years might be based." The survey, under the heading "Natural Division and School Zones", makes the following statement:

Several strong natural features divide the city area into definite and separate school zones, each of which must attain to a complete system of schools for itself in order to conserve the best interests of all the people and the travel time and safety of the children. First of all, the Mahoning River divides the territory east and west into two large units which we call the North and South sides of the city. Then the North side is again divided by Crab Creek and its many railroad tracks into two distinct sections. The North-west section we shall call School Zone No. 1, and the North-east No. 2. The South side is cut into four well defined parts by Mill Creek and its Park, by Pine Hollow, and the Southern Railroad and its ravine, thus creating four distinct sections. The East one will be Zone No. 3; the next, No. 4; the next, No. 5; and the Western one, No. 6. All of these secondary but marked features of division in the territory connect with the Mahoning River as the primary dividing line and thus set forth six clear and clear-cut zones of area and population to be treated as such in a constructive program of schools for the future. This survey will so recognize these natural divisions of the city territory and provide a complete system of schools for each of them.

As the District Court pointed out, the school system did not develop precisely according to the survey. Although six high school zones were in fact established, only three were situated below the Mahoning River. One reason for this was the city's unanticipated failure to expand on its south-east side outside the 1921 city limits. The other deviation from the survey was the development of a third zone north of the river. Nevertheless, there is substantial support for the District Court's observation that the school zones in Youngstown were created essentially in conformity with the survey.

Appellants contend that the District Court relied on the survey to treat Youngstown's high school zones as separate and impenetrable subsystems of the school district. They argue that the size of the districts and the numerous instances of student assignments across high school attendance zones preclude such treatment. Moreover, appellants assert that the Supreme Court's opinion in Keyes prohibits an analysis of the system in compartmentalized units even if these units have a rational basis in fact.

Our review of the District Court's opinion convinces us that it did not treat the six high school zones as impregnable subsystems. It is true that the court's opinion addresses the factual allegations within the framework of the six high school zones and that it examines the challenged actions and failures to act on an individual basis. However, the enormous complexity of the factual presentations in these cases requires a trial court to choose an orderly framework for its discussion of the facts. We will not assume from the format chosen here that the District Court, as appellants contend, "failed to see the forest for the trees."7 The opinion thoroughly analyzes the inter-zonal and systemwide effects at all educational levels of various official actions.

We find no error in the District Court's resort to the 1921 survey as evidence to support its findings of fact. As we pointed out above, a showing of disparate impact provides no shortcut to the ultimate fact of discriminatory intent. "Determining whether invidious discriminatory purpose was a motivating factor demands a sensitive inquiry into such circumstantial and direct evidence of intent as may be available." Arlington Heights, supra, 429 U.S. at 266, 97 S. Ct. at 563. The 1921 survey, along with similar studies and internal memoranda referred to by the District Court, provides the sort of historical background to the challenged official conduct which, under Arlington Heights, is relevant to the issue of discriminatory intent.

We conclude that the District Court's opinion reveals an adequate grasp of the legal principles applicable to this case. Any misapprehension or misapplication of those principles necessarily inured to the benefit of appellants. The central issue in this appeal, therefore, is whether the District Court's factual findings must be set aside. Before examining the merits of that contention, we deem it appropriate to articulate what we believe to be the appropriate standard of review.

In civil cases, Rule 52(a) of the Federal Rules of Civil Procedure provides that an appellate court must not set aside findings of fact of the District Court unless those findings are clearly erroneous. Findings of fact are "clearly erroneous" only when the reviewing court "is left with the definite and firm conviction that a mistake has been committed." United States v. United States Gypsum Co., 333 U.S. 364, 68 S. Ct. 525, 92 L. Ed. 746 (1958); Johnson v. United States, 600 F.2d 1218, 1222 (6th Cir. 1979).

Appellants argue that this standard is inapplicable to the present case for two reasons. First, they claim that a reading of the District Court's opinion reveals that the findings of the court were primarily based on documentary, rather than testimonial, evidence. Appellants claim that because we have before us the same "paper" evidence upon which the trial court based its decision, the clearly erroneous rule plays only a limited role in our review of those findings.8 This is not the rule in this circuit. The clearly erroneous standard applies notwithstanding the fact that the appellate record may consist entirely of documentary evidence. United States Steel Corporation v. Fuhrman, 407 F.2d 1143 (6th Cir. 1969), cert. denied, 398 U.S. 958, 90 S. Ct. 2162, 26 L. Ed. 2d 542; Gartrell v. United States, 619 F.2d 1150 (6th Cir. 1980). "The 'clearly erroneous' test does not derive solely from the trial judge's superior opportunity to assess the credibility of witnesses; it also reflects and preserves the proper relationship between trial courts and courts of appeal." United States v. Jabara, 644 F.2d 574, 577 (6th Cir. 1981).

The second reason cited by appellants for the proposition that the clearly erroneous rule plays only a limited role in this appeal is that the contested issues involve mixed questions of law and fact. As we observed above, the District Court's opinion contains no errors of law upon which a reversal can be based. For that reason, the critical question in this case is the propriety of the District Court's determination that the defendants did not act with segregative intent. That finding is the crux of the decision below, and the standard to be employed in reviewing it is of paramount importance. This court has always treated district courts' resolutions of this question as findings of fact, to be set aside on appeal only if clearly erroneous. See, e.g., Reed v. Rhodes, 607 F.2d 714, 717, 732 (6th Cir. 1979), cert. denied, 445 U.S. 935, 100 S. Ct. 1329, 63 L. Ed. 2d 770 (1980); Penick v. Columbus Board of Education, 583 F.2d 787, 789, 798 (6th Cir. 1978), aff'd, 443 U.S. 449, 99 S. Ct. 2941, 61 L. Ed. 2d 666 (1979); Brinkman v. Gilligan, 583 F.2d 243 (6th Cir. 1978), aff'd, 443 U.S. 526, 99 S. Ct. 2971, 61 L. Ed. 2d 720 (1979). Moreover, it is clear from the decisions of the Supreme Court that such findings are subject on appeal to the clearly erroneous rule. Referring to our determination in the Dayton case that the district court's finding of no intentional segregation was clearly erroneous, the Supreme Court stated: (T)here is great value in appellate courts showing deference to the fact-finding of local trial judges.... The clearly erroneous standard serves that purpose well. But under that standard, the role and duty of the Court of Appeals are clear: it must determine whether the trial court's findings are clearly erroneous, sustain them if they are not, but set them aside if they are. The Court of Appeals performed its unavoidable duty in this case and concluded that the District Court had erred." Dayton Board of Education v. Brinkman (Dayton II), supra, 443 U.S. 526, 534 n.8, 99 S. Ct. 2971, 2977 n.8, 61 L. Ed. 2d 720 (1979). We think it clear, therefore, that the District Court's finding in this case of no intentional segregation must be accorded the same deference usually given to findings of fact in civil cases. See Columbus, supra, at 468, 99 S.Ct. at 2952 (Burger, C. J., concurring). Indeed, there is a persuasive argument in support of even greater deference to such findings in school desegregation cases. Justice Stewart stated that argument in his separate opinion in the Columbus and Dayton II cases:

It seems to me that the Court of Appeals in both of these cases ignored the crucial role of the federal district courts in school desegregation litigation-a role repeatedly emphasized by this Court throughout the course of school desegregation controversies, from (Brown II to Dayton I ). The development of the law concerning school segregation has not reduced the need for sound factfinding by the district courts, nor lessened the appropriateness of deference to their findings of fact. To the contrary, the elimination of the more conspicuous forms of governmentally ordained racial segregation over the last 25 years counsels undiminished deference to the factual adjudications of the federal trial judges in cases such as these, uniquely situated as those judges are to appraise the societal forces at work in the communities where they sit.

Whether actions that produce racial separation are intentional within the meaning of (Keyes, Washington v. Davis, and Arlington Heights ) is an issue that can present very difficult and subtle factual questions. Similarly intricate may be factual inquiries into the breadth of any constitutional violation, and hence of any permissible remedy. See Millikin v. Bradley, 418 U.S. 717 (94 S. Ct. 3112, 41 L. Ed. 2d 1069) (Millikin I ); Dayton I, supra. Those tasks are difficult enough for a trial judge. The coldness and impersonality of a printed record, containing the only evidence available to an appellate court in any case, can hardly make the answers any clearer. I doubt neither the diligence nor the perseverance of the judges of the courts of appeals, or of my Brethren, but I suspect that it is impossible for a reviewing court factually to know a case from a 6,600-page printed record as well as the trial judges knew it. In assessing the facts in lawsuits like these, therefore, I think appellate courts should accept even more readily than in most cases the factual findings of the courts of first instance.

443 U.S. at 469-71, 99 S. Ct. at 2983 (footnotes omitted). With these principles in mind, we turn to an examination of the District Court's finding that the defendants did not intentionally act to segregate Youngstown's public schools.

We shall not endeavor in this opinion to discuss all the challenged boundary changes, feeder patterns, schoolsite selections, and other actions and failures to act addressed by the trial court.9 With respect to many of those actions, appellants' brief contains only a conclusory statement that the action was intentionally segregative, without discussing why the District Court's findings to the contrary are clearly erroneous. Other findings by the District Court, although challenged in appellants' brief, are sufficiently supported by facts cited in the District Court's opinion. Upon careful review of the record, briefs, and oral arguments presented to this court, we think further discussion of many of these findings would be unnecessarily duplicative. Indeed, we feel that the trial court's opinion represents an admirable analysis of the prodigious body of facts presented by the parties. The opinion articulates adequate bases for all of its factual and legal conclusions. Nevertheless, we deem it appropriate to address specifically some of the challenged acts and series of acts undertaken by the school officials. These actions are among those which provide the strongest support for the inference that Youngstown's school system was intentionally segregated.

Appellants contend that it was deliberately segregative to open Covington Elementary School in 1940 to serve an area which was predominantly black. The District Court found that the facility actually replaced the Covington Elementary School structure in existence since at least 1900. It further noted that the 1921 Survey described the structure as perhaps the worst school building in Youngstown and recommended its replacement because of its ideal location. 454 F. Supp. at 993. In our view, these findings obviously represent a determination by the court that the construction and siting of the Covington school in 1940 was not the result of racial discrimination. Because they are adequately supported by the record, we do not find them to be clearly erroneous.

Appellants assert that Youngstown officials engaged in an effort to separate the western part of the district, a predominantly white residential area, from the area where most of Youngstown's blacks lived. As part of this effort, a series of actions was allegedly taken to erect a barrier between the South and Chaney High School areas. In 1940, Grant Junior High School was converted into an elementary school, and Hillman Junior High School was opened. Grant had been a feeder school for Chaney High School, but after the conversion it was reassigned to the South High School zone where it became a feeder school for Hillman Junior High. Hillman's attendance area included the majority of the black population living within the South High School area, and the majority of the black students within Hillman's area resided within the attendance area for Grant elementary school. To accommodate the inclusion of the new Grant elementary school within its attendance area, South High's northern boundary was expanded to include Grant's area. According to the 1940 census data, South's new attendance area now contained virtually all of the black population south of the Mahoning River.

The District Court found that these actions were not motivated by racial considerations. The conversion of Grant to an elementary school and the opening of a junior high school at the Hillman site had been specifically recommended by the 1921 Building Survey, which the court found was concluded prior to the development of the racial residential pattern in that area of Youngstown. The court also noted that, based on racial composition data for the 1952-53 school year, Grant and Hillman in 1940 were probably at or only slightly above the system-wide average in their black student populations.10 According to the District Court, the actual intent of the school officials was to provide modern school plants within each of the six separate population zones of Youngstown. The new South High zone conformed to the natural boundaries of the Mahoning River and Mill Creek Park. The court concluded that the preponderance of the evidence does not demonstrate any segregative intent on the part of the officials in undertaking these changes, but rather supports the finding that they were made in response to the population growth south of the Mahoning River which was recognized as early as the 1921 Survey. 454 F. Supp. at 1059. Again, the District Court's findings are supported by facts in the record, and cannot, in our view, be set aside as clearly erroneous.

One form of "acts and omissions" which the District Court found demonstrative of a "reasonable possibility of establishing a pattern of intentionally segregative actions" was the series of assignments of students from the old Harrison Elementary School. Prior to 1949, the Harrison school zone included areas on both the north and south sides of Thorn Hill Avenue. The students residing north of Thorn Hill road were predominantly, if not exclusively, white. Those living south of Thorn Hill were racially mixed and from predominantly low-income homes. In 1949 Harrison was closed. The black and lower income students south of Thorn Hill Avenue were assigned to the adjacent Madison Elementary School, which in 1952 was 21.0% black. The students north of Thorn Hill Road were temporarily assigned to discontiguous Richey Elementary School. In 1951 those students were reassigned to Elm.

In 1957 a new Harrison Elementary School was opened. It was located south of Thorn Hill Road, and was not large enough to accommodate all the students in the old Harrison attendance area. Among the reasons for this was the development in the preceding years of two housing projects in the area south of Thorn Hill Road. Students residing north of Thorn Hill Road continued to be assigned to Elm after the new Harrison school was built.

In 1965 Elm Elementary was closed. The white students from the area north of Thorn Hill were reassigned to Harding, which at that time had a black population of less than 10%. Finally, in 1973, these students were reassigned to McKinley, and McKinley's attendance area was altered to include the area north of Thorn Hill Avenue.11 McKinley also had a relatively low percentage of black students at the time.

Appellants claim that this series of actions demonstrates a determined effort on the part of the school officials to separate the white and black students in the Harrison attendance area. They claim that the new Harrison school was deliberately "underbuilt" so the white students in the area would never have to attend it.

The District Court, however, viewing these assignments in their historical and chronological context, found that there was no reasonable basis for inferring discriminatory effect or intent at least prior to 1965. It found that the 1949 assignment of the white students north of Thorn Hill to Richey was the result of their parents' requests that the students stay in the Rayen High School feeder pattern.12 As for the 1951 reassignment of these students to Elm, the District Court found this action to be integrative in effect, in light of Elm's above average percentage of black students at that time. 454 F. Supp. at 1002.

The District Court also rejected the assertion that the construction of the new Harrison school and the concomitant assignments of students supported an inference of intentional discrimination. It noted that the Harrison school was quite sizable by local standards. The court apparently believed that the lack of space was due to a lack of foresight rather than to a deliberate plan.13 The continued assignment of the white students north of Thorn Hill to Elm after the opening of Harrison was the "probabl(e) result", according to the court, of the lack of room at Harrison in conjunction with the 1949 promise to families residing north of Thorn Hill Avenue that their children would remain in the Rayen feeder pattern. Moreover, the court found it significant that Elm had a higher proportion of black students than Madison both before and after the 1957 opening of Harrison.14 Finally, the District Court's conclusion as to the pre-1965 actions in this matter is supported by the fact that Elm was also highly integrated; it was 49% black in 1958, as compared to Harrison's 52%, and was 71.6% black in 1963, as compared to 64.2% at Harrison. See 454 F. Supp. at 1003, 1007.

The District Court found that the post-1965 actions concerning the white students residing north of Thorn Hill Avenue supported an inference of segregative intent. The 1965 reassignment to Harding, a predominantly white school, was segregative in effect. In addition, by 1971, Harrison Elementary, the "neighborhood" school for these children, was sufficiently underutilized to accommodate them. The reassignment of these students to McKinley in 1973 was also segregative in effect; McKinley's black enrollment was significantly below the elementary school average, while Harrison, the neighborhood school, was a racially identifiable black school. The court nevertheless declined to draw the inference of segregative intent. It concluded that the 1973 boundary change bringing these students within the McKinley area was not the result of a segregative purpose. It based this conclusion on several findings. First, Superintendent Robert Pegues, who ordered the reassignment, had expressly committed himself to improve racial balance in Youngstown's schools. Second, it found that the school officials were not motivated by racial considerations when they based their decision not to enlarge the new Harrison school on "socio-economic" and "sociological" differences between the students north of Thorn Hill and those south of Thorn Hill.15 The court apparently agreed with defendants that the differences referred to in the memorandum concerned relative income and social positions and not to the students' race. Finally, the court relied on other contemporaneous correspondence indicating that the school board's motives in making the assignments were not racially discriminatory, but rather were to remedy perceived utilization problems and keep the promise made to the parents north of Thorn Hill when the old Harrison School was closed in 1949.

It is not for us to decide whether it is appropriate for school officials to consider "socio-economic" differences among school children. We must determine only whether there is an adequate evidentiary foundation for the trial court's determination that none of the actions with respect to the Harrison area were motivated by a racially segregative intent. In this instance, we conclude that the trial court's finding was not clearly erroneous.

Another series of acts which might support an inference of segregative intent occurred in the South and Wilson High School zones. Williamson Elementary School, located south of the Mahoning River and opposite the central city area, was constructed in 1949. A 1956 boundary change ceded nearly one-third of Williamson's students to Bennett Elementary, which had a lower black enrollment. Although the elementary students from Bennett were fed into Wilson Junior and Senior High Schools, the former Williamson area students at Bennett were fed into Hillman Junior and South High Schools, the receiver schools for Williamson. It is not clear from the record what the racial composition of the transferred area was; the parties' stipulation regarding this boundary change does not include this information. Appellants assert that it was predominantly white, thereby increasing segregation at the elementary school level. However, this District Court noted that the 1960 census data indicate that the affected area was predominantly black. Moreover, plaintiffs apparently argued below that the segregative effect of the boundary change was the failure to carry it through at the secondary level. See 454 F. Supp. at 1034. Such a segregative effect could only occur if the students ceded by Williamson were predominantly black. In any event, the District Court concluded that the boundary change was necessitated by overcrowded conditions at Williamson which did not persist at the schools into which it fed, and that therefore no general reorganization of the feeder pattern structure was necessary. 454 F. Supp. at 1034. That conclusion is based on facts in the record, and the court's refusal to draw the inference of segregative intent cannot be termed clearly erroneous.

In 1958 Williamson's boundary was contracted again, transferring an area stipulated to be predominantly white in residential population to Garfield, which was a predominantly white school. The District Court noted that the transfer may have been integrative, since the few black students involved may have increased Garfield's proportion of black students. This observation is questionable; it appears to ignore the effects of a transfer of white students from Williamson, which in 1958-59 was 30% black. Nevertheless, the District Court found that the boundary change was obviously logical from a utilization standpoint. The figures cited by the court indicate that the change reduced the overcrowding at Williamson while adding students to the contiguous, underutilized Garfield.

In 1960 Williamson's boundary was expanded to take in a part of the Monroe Elementary attendance zone. At the same time, Monroe was also transferring part of its area to Cleveland Elementary. Because the area transferred to Williamson was predominantly white and Williamson was slightly above the system-wide average in black enrollment, the District Court concluded that the transfer was racially neutral in effect. This assessment disregards the apparent segregative impact of the boundary change on Monroe, which was 49.4% black in 1958-59 and which lost white students. Again, however, the District Court found that the boundary change was a reasonable response to facility utilization problems, a finding which is supported by the record.

In 1961 there began another series of boundary changes involving Williamson, Bennett, Garfield, and Monroe Elementary schools. Those changes, along with the inconsistencies and deficiencies in the data supplied to the court below, were discussed at length in the District Court's opinion. 454 F. Supp. at 1036-49. Related to the boundary changes was the leasing of the Glenwood facility to relieve overcrowding at Monroe Elementary.

Appellant alleges that these changes were discriminatory in and of themselves, and were also part of an overall plan to protect the white Wilson High School zone from black encroachment. Wilson was the key to this alleged plan; appellants argue that it had been made a "racial traffic cop", redistributing black students among Garfield, Bennett, and Monroe so as to divert black students away from Wilson's zone. The District Court declined to draw those conclusions. It is clear from the court's opinion that it found the evidence supportive of a finding of segregative intent. However, the court also cites reasons, such as facility utilization and a racially neutral neighborhood school policy, which support its conclusion that the challenged boundary changes were not motivated by segregative intent. The court's characterization of the school officials' conduct as too spontaneous and unplanned to warrant a finding of systematic segregative gerrymandering is also supported by the evidence. We agree that the record portrays a school board which "ran a 'stop gap' operation," 454 F. Supp. at 1960, and which continually made administrative errors and miscalculations. The District Court concluded that the board was incapable of engaging in the sophisticated population manipulation which the plaintiffs alleged had occurred. Again, the court's conclusions find adequate support in the record, and cannot, in our view, be deemed clearly erroneous.

In conclusion, it is clear from a reading of the District Court's opinion and the voluminous record before us that the facts of this case could support a judgment in favor of either side. Our function is not to determine whether we would have decided that defendants created or maintained an intentionally segregated school system. Rather, it is to review the court's opinion, the record, the briefs, and oral arguments, and to decide whether they leave us with the definite and firm conviction that a mistake has been committed. After careful review, we entertain no such conviction, and therefore affirm the judgment of the District Court with respect to the Youngstown defendants.

We also affirm the District Court's finding of no liability on the part of the state defendants. We have considered this question carefully in light of our decisions in Penick v. Columbus Board of Education, supra, and Reed v. Rhodes, supra, and conclude that neither case indicates a contrary result here.

In Penick and Reed, the school districts involved had practiced extensive de jure segregation; the state's liability was predicated on its failure to prevent or remedy this unlawful course of conduct. In the present case, by contrast, we have upheld the District Court's ruling that the Youngstown school officials acted unlawfully only in the narrowly circumscribed area of teacher assignments. It is true, of course, that the state failed to investigate the Youngstown schools; had it done so, however, it presumably would have reached the same conclusions as the District Court. Under these circumstances, we do not believe any useful purpose would be served by remanding the case for further inquiry into this issue.

The judgment of the District Court is affirmed.