United States v. Gregory-Portland Independent School District, Plaintiff-Intervenor-Appellant v. State of Texas

GEE, Circuit Judge: The appellant school district comprises two discrete communities of differing ethnic composition, surrounded and separated by miles of farmland. It is the product of an innocently motivated consolidation of two former districts, each centered on one of the communities. Neither the establishment of the former districts nor their consolidation into the present one was the result of discriminatory state or local action. This appeal presents questions of whether and, if so, in what circumstances intercommunity busing to produce a racial balance in each of a district’s schools that corresponds to that of the district at large is required by the Constitution of the United States. Subsidiary but significant procedural questions are raised, among them those presented by the district court’s application of burden-shifting presumptions derived by it from the Supreme Court’s opinion in Keyes v. Denver School District No. 1, 413 U.S. 189, 93 S.Ct. 2686, 37 L.Ed.2d 548 (1973), and of its reliance on a great body of stipulations between two allied parties only,'-filed in a connected case. As the factual and legal setting is complex, we are obliged to write at some length. The Gregory-Portland Independent School District Gregory and Portland are two essentially rural communities lying some miles north and east of Corpus Christi, Texas, separated from its urban environs by a wide bay of the Gulf of Mexico but connected to it by a bridge at all times relevant here. For decades before 1950, each town and its surrounding rural area was served by its own local school district. Gregory, with a population in 1950 of about 1,000, was and is the smaller town, lying inland about four miles north of Corpus Christi Bay. Four miles southward on the bay shore, separated from Gregory by unpopulated farmland, lay Portland, with about 1,300 residents. In that year, motivated by ethnically neutral purposes, the voters of each district, voting in separate ballots, elected to merge the two school districts. At that time the children of both of the two significant ethnic groups in the area attended local schools located in each town. Thus, the proportions of Mexican-American and Anglo students in each local school necessarily mirrored the ethnic composition of each local school district. 1 *992 Neither school reflected the then overall 68/42 percent Mexican-American/Anglo ethnic composition of the new district, however, since, for neutral, historical reasons, almost eighty percent of the students at the Gregory elementary school were Mexican-American, by contrast to only slightly over thirty percent at Portland. Thus in 1950, innocently and by a stroke of the voters’ pens, were created what the trial court termed “ethnically identifiable schools” within one now-unitary district. At this time in Texas, black and white students attended schools strictly segregated de jure. As for the nation, Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), still lay four years in the future. At no time, however, had Texas segregated Anglo students from Mexican-American ones by law. Then, as now, local school boards like appellant Gregory-Portland Independent School District (“GPISD”) were by law autonomous bodies exercising exclusive control of the public schools within their local jurisdictions. 2 Then as now, independent of control by the State of Texas and its departments (except now of one constituted as its agent by the court below), these local governmental corporations constructed schools, assigned students and faculty, established attendance zones, and in all such allied respects saw to the business of the local schools. Another state-created entity, the Texas Education Agency, figures in these proceedings as the agency empowered to administer state school funds and distribute them to such school districts as GPISD. Pursuant to these powers, derived from state sovereignty, for the school year immediately following the consolidation election the newly constituted GPISD Board (sometimes hereafter “Board”) voted to continue assignment of elementary students residing within the former Portland school district to the local Portland school and to do likewise as to residents of the former Gregory district and the local Gregory facility. All junior high and high school students were assigned to a single, unitary school each, as they are today; and thus the Board’s policies as to them are not under scrutiny here. In that same year, 1951, a major national corporation located a bauxite plant in the area of the former Gregory school district. 3 Undisputed testimony in the record suggests, though the trial court made no finding on the subject, that bauxite pollution borne on prevailing winds thereafter inhibited the population growth of Gregory. At all events and for whatever reasons, during the thirty years intervening between the *993 merger of the districts and trial below, the population of Gregory has inched up from about 1,000 to only 2,500, while that of Portland has increased tenfold to 12,000. During this same period the ethnic composition of the student bodies in elementary schools also shifted, so that by the time of trial the Gregory school was 94.5 percent Mexican-American, while the student body at Portland became 81 percent Anglo. Nothing in the record suggests, however, nor did the trial court find, that these differential growth patterns and ethnic shifts had anything to do with state or GPISD action, let alone action of a discriminatory character by either. 4 The response of the Board to these changes was to build to the population growth, leaving the southern boundary of the attendance zone for the elementary school at Gregory where it was at the time of the district’s consolidation and where it remained at the time of trial. Two new elementary schools were constructed in the Portland area, one in 1962 and one in 1979. Thus, at the time of trial Portland, with a population of 12,000, had three such schools and Gregory, with a population of 2,500, had one. 5 Since population distribution within the district has always been representable, schematically, by the figure of an hourglass or barbell, with two relatively small foci connected by a four-mile roadway, it is obvious that no configuration of attendance zone short of one about four miles in length could have diluted the concentration of young Mexican-American school children at Gregory. Equally apparent is that a daily four-mile walk or bicycle ride, both ways down a busy highway, is not reasonable to expect of kindergarteners or even of sixth graders. Thus, the Board’s solution, which utilized existing facilities, avoided unnecessary transportation of students, and assigned younger children to schools near their homes, was an eminently reasonable and practical one. That it was does not, of course, dispose of the question before us: whether that solution violated the Constitution. It does, however, conclusively remove the Board’s scheme from that category of tortured arrangements which, in and of themselves, suggest suspect motives. Bearing this thumbnail sketch of the District’s genesis in mind, we turn now to developments leading up to the trial below, to that trial, and to the district court’s decision. 6 The Case Below 1. Procedural Background. The roots of this litigation lie far from Gregory-Portland or South Texas. In course of the process of dismantling Texas’ dual black/white school system, as required by Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), the United States brought a suit in the Eastern District of Texas relating to various counties lying mostly in the northeastern quadrant of the state. One result of that proceeding was a determination by the court that the Texas Education Agency (“TEA”) had failed properly to discharge its duty to assist in eliminating the vestiges of racially segregated public education. That court ordered various corrective measures, including that the TEA monitor and report on the number of state schools having minority enrollments exceeding 66 percent and devise and submit to the court a general plan to carry out its duties under Title VI and the fourteenth amendment, to be enforced on school districts by the sanctions of with *994 drawal of accreditation and funds. See United States v. State of Texas, 321 F.Supp. 1043 (E.D.Tex.1970), 330 F.Supp. 235 (E.D.Tex.1971), aff’d and modified, 447 F.2d 441 (5th Cir. 1971), cert. denied, 404 U.S. 1016 (1972). The effect of this and succeeding orders was to draw to this one, local district court much of the responsibility for desegregation throughout the extent of the State of Texas, e. g., United States v. State of Texas, 342 F.Supp. 24 (E.D.Tex.1971), aff’d, 466 F.2d 518 (5th Cir. 1972) (concerning San Felipe Del Rio CISD, located on the Mexican border). All parties agree that the more proximate cause of this suit was an order of that court entered on August 9, 1973, at a time when GPISD was not a party to the ongoing proceedings, requiring the TEA to locate all more than 66 percent minority schools throughout the state, make findings whether the concerned student assignment plan of the district complied with the court’s order, notify any found to be out of compliance of that fact, provide it with a detailed written plan for compliance, and suspend its funds and accreditation. Any school aggrieved by such TEA action might petition the District Court for the Eastern District of Texas for such relief as that court might deem proper. Acting pursuant to the court’s above order, but in exclusive reliance on the mere percentage figures of minority attendance at the Gregory elementary school, three months later, in November of 1973, TEA notified GPISD that because of the concentration of Mexican-American students at the Gregory school and of correspondingly high Anglo percentages at two of the Portland schools, its funds and accreditation would be cut off unless it adopted one of two TEA-confected plans to “eliminate Austin [at Gregory] Elementary School as a minority identifiable school campus.” TEA took this action without inquiring into the reasons for the minority concentration at Gregory; nor did it hold or offer to hold any hearing on the subject. GPISD responded by filing suit in the federal district court for the Southern District of Texas, seeking an injunction against TEA’s threatened sanctions. Mandamus to transfer the case to the Eastern District, sought by the United States, was denied. United States v. United States District Court, Southern District of Texas, 506 F.2d 383 (5th Cir. 1974). Thereafter a trial on the merits was held, at which that court determined that GPISD had been guilty of no discrimination and enjoined the threatened sanctions. On appeal, however, we held that the matter should have been tried in the Eastern District and vacated the judgment in favor of GPISD. Gregory-Portland Independent School District v. Texas Education Agency, 576 F.2d 81 (5th Cir. 1978), cert. denied, 440 U.S. 946, 99 S.Ct. 1423, 59 L.Ed.2d 634 (1979). The case was then transferred to the Eastern District and tried on the merits in May of 1980. 2. Trial Procedures. The trial court recognized that, since there had been no de jure segregation of Mexican-American students anywhere in Texas, the fourteenth amendment violation requisite to the court’s intervention in student assignments by GPISD turned on whether the school authorities had been guilty of intentional discrimination. Keyes v. Denver School District No. 1, 413 U.S. 189, 208, 93 S.Ct. 2694, 2697, 37 L.Ed.2d 548 (1973). It concluded, however, that two presumptions, which it derived from Keyes and from later expressions by us, applied to the case in such a manner as to place the burden on the GPISD Board to prove that it had not been guilty of discriminatory purposes or intent in its assignment policies, rather than on TEA and the United States to establish that it had. One such presumption recognized and applied by the court was that stated by us in United States v. Texas Education Agency, 579 F.2d at 910, 913 (5th Cir. 1978), cert. denied, 443 U.S. 915, 99 S.Ct. 3106, 61 L.Ed.2d 879 (1979): “In a tri-ethnic setting, Keyes means that a finding of intentional segregation against one minority group raises the presumption that any segregation suffered by the second minority group was intentional.” Noting that the State of Texas had plainly discriminated against black *995 students in the past, the court concluded that a presumption arose that the minority concentration at Gregory elementary was the product of discriminatory intent by GPISD. United States v. State of Texas, 498 F.Supp. 1356, 1363 (E.D.Tex.1980). If more were needed to shift the burden of proof, the court discerned a second presumption to like effect arising from the existence of intentional actions by TEA, confessed by it in stipulations with the United States, leading to segregation of Mexican-American students in other school districts throughout Texas. This presumption as well was to be traced to the opinion in Keyes: Plainly, a finding of intentional segregation as to a portion of a school system is not devoid of probative value in assessing the school authorities’ intent with respect to other parts of the same school system. On the contrary, where, as here, the case involves one school board, a finding of intentional segregation on its part in one portion of a school system is highly relevant to the issue of the board’s intent with respect to other segregated schools in the system. 413 U.S. at 207, 93 S.Ct. at 2697. The trial therefore commenced with the court’s announcement that a burden of proof to disprove discriminatory intent would be placed on GPISD. By reference to two-party stipulations between the United States and TEA, the court determined that the State of Texas had been guilty of intentionally segregative acts by failing to enforce neutral student assignment policies on local school districts, by approving construction financing for “Mexican” schools, and in other, similar respects. None of this matter referred specifically to GPISD, however, and all actions admitted by TEA were in the nature of either failures by it to prevent a local district board from taking some segregative action or of taking some action indirectly supporting (as by approval of construction bonds) some such action. TEA’s admissions, then, took the form of concessions that it had passively aided and abetted local boards already resolved, as principal actors, on doing wrongful things; nowhere did it admit that it had forced discriminatory policies on local boards who wished to act properly, and indeed under state law it lacked power to do so. In the course of the trial, the district court, by reference to the Keyes presumptions noted and to the TEA stipulations, specifically approved the action of TEA in proceeding against GPISD on the sole basis of numbers and percentages at the Gregory school, without inquiry into reasons and without holding or offering to hold a hearing. In so doing the court observed, “[i]f TEA determines the existence of ethnically identifiable schools, it need show nothing further to implement the sanctions ....” 7 United States v. Texas, 498 F.Supp. at 1366. In addition, the court deemed significant evidence of the operation in the GPISD’s area of “Mexican schools” during the 1940’s and earlier, of the use by GPISD during the 1950’s and 1960’s of classroom ability groupings, of certain teacher assignments, of what the court termed a “No-Spanish” rule, and of three or four specific instances of what the court viewed as overt discrimination, instances such as the use by the high school coach of racial epithets. In the end the court concluded, on the basis of this and other matter, that “the evidence overwhelmingly supports the conclusion that GPISD has intentionally discriminated against the Mexican-American students in the district . . .. ” Id. at 1368. On appeal, the United States urges that this is so, regardless of where the burden of proof lay and even without reference to the TEA-United States stipulations. Based on this ultimate finding, the court entered an order assigning all kindergarten and first-grade students in the district to the school at Gregory, grades 2 and 3 to one *996 of the Portland elementary schools, grade 4 to another, and grades 5 and 6 to the third. The consequence has been intercommunity busing of almost 60 percent of the district’s five- to twelve-year-old children, at an estimated annual cost of over $600,000. Because of serious legal and factual errors made by the district court and discussed by us below, we must reverse. Erroneous Allocation of the Burden of Proof The Keyes presumptions, by means of which the district court placed the burden of proof on GPISD to establish that it had not intentionally discriminated in student assignments, do not apply to this case. One of these rebuttably presumes intentional discrimination in all of a unitary system if it be found as to any substantial portion of that system. Since TEA admitted that it had been guilty of permitting, even indirectly encouraging, segregative pupil assignments by some local boards, and since it further admitted that some such boards had intentionally engaged in such practices, the court concluded that it was presented with a Keyes situation. For two cognate reasons, it was not. The first is that TEA has no authority over either the location of schools or the assignment of pupils; under Texas law, these are exclusive prerogatives of local school boards. Tex.Educ.Code § 23.26 (1972). Second, Texas’ school system is not a unitary one as was the local district in Keyes. In Texas, different bodies — the various local boards — make pupil assignment policy in myriad districts throughout the state. This Keyes presumption is grounded in the logical proposition that one may infer the motive upon which a given decision-maker acted in one situation from its known motive for acting in a similar one. 8 Here no such decision-maker exists: TEA has no such powers, being in these things able at most to second the decision of a local board, and the actions and motives of any given local board tell us little or nothing about those of another distinct and different one. Thus the district court’s reliance upon this presumption was misplaced. The second Keyes presumption upon which the district court mistakenly relied was that presuming intentional discrimination, in a tri-ethnic situation, against one minority group from the fact that it has been established as to another. This presumption, like the first, assumes one given decision-maker and fails of application here for want of any. It fails as to the GPISD Board for an additional reason. That Board, if anyone, is the “given decision-maker” here, and the record is devoid of evidence that this Board has ever discriminated against Negro students. Indeed, had it been disposed to do so it probably never had the opportunity; even at the time of trial there were only eleven black students among its over four thousand scholastics, and the GPISD system is essentially a bi-ethnic, not a tri-ethnic, one. There can be no logical deduction from what the state or other local boards have done to black students elsewhere in the past to what this Board has done regarding Mexican-American students. It is thus apparent that no logical basis existed for employing either of the Keyes presumptions to require the GPISD Board to prove its innocence of discriminatory intent. The court’s action in doing so therefore resulted in a proceeding erroneously and fundamentally skewed against GPISD. Other Legal and Procedural Errors An additional error stemming from the district court’s mistaken application of Keyes is its approval of TEA’s arbitrary *997 suspension of GPISD funds and accreditation. Among the “undisputed facts” agreed to by all three parties (to be distinguished from the United States-Texas two-party stipulations) is that this action was taken on nothing more substantial than the observation by TEA that there was a very high percentage (over 90 percent) of Mexican-American students at the Gregory elementary school, with much lower ones at two of the Portland schools. Undisputed Fact No. 45. Without further inquiry and with neither hearing nor the offer of one, TEA acted in obedience to the court’s 1973 order to impose a “remedial” pupil assignment policy on GPISD on pain of withdrawing both funds and accreditation from the district. Id. Whatever relief and due process GPISD might hope for, were to be had only by petitioning a faraway federal court and, as the event showed, proving the innocence of its motives. In its opinion, the district court declared that such a procedure on the part of TEA was precisely what the court’s 1973 order contemplated and required, specifically approving it. We cannot do so. In the first place, as we rioted above, the Keyes presumptions, relied on by the court to justify TEA’s action as well as its own, 9 have no application here. In the second, under Keyes a finding of intentional discrimination against Mexican-Americans is a prerequisite to court intervention in the student assignment policies of Texas school boards. Here, as all three parties agree, TEA acted merely as the agent of the court and in obedience to its order, exercising no independent policy judgment or initiative of its own except to obey the court’s order. Undisputed Fact No. 44. For TEA’s application, as we noted, the court laid down a presumption that intentional discrimination against Mexican-American students is established by their mere presence in percentages over 66 percent in any given student body. From the application of this presumption flows immediate action by the court’s agent, TEA, to impose final requirements upon any such district that the court itself could have imposed only after a hearing on the merits, requirements backed by sanctions competent to destroy the district and its functions root and branch. The district attacks this procedure directly, and we find it invalid as applied here. To reiterate how it actually operated against GPISD is to demonstrate why we do: by the order of August 1973, the district court established TEA as its agent for purposes of combing the Texas school districts in search of ethnically disproportionate campuses. These the court defined as ones comprising student bodies more than 66 percent minority. On finding one, and without further hearing or inquiry into causation, TEA was to do as it did here: offer the district a choice between accepting a student assignment plan or plans devised by TEA to dilute minority attendance, or suffering the imposition of sanctions calculated to disestablish it. Relief from either could be had only by the district’s proving its innocence of discriminatory intent in a single and faraway federal court. These things were done here and explicitly approved by the court. At least two fatal defects are apparent. First, the factual premise upon which rests the presumption’s inference of causation by discriminatory intent is inadequate. It is matter of general knowledge, shown in Census figures that we may notice, that many Texas counties have a percentage of Mexican-American population far greater than the 66 percent trigger-figure fixed by the court in its order. To cite only one example, in 1973 Hidalgo County’s percentage was 79.1. See Castaneda v. Partida, 430 U.S. 482, 486, 97 S.Ct. 1272, 1276, 51 *998 L.Ed.2d 498 (1977). 10 All of its schools may therefore fail the district court’s test. 11 To infer discriminatory intent from such slender factual data and act decisively upon that inference is to run a high risk of acting unjustly. Presumptions must bear a reasonable relationship to the facts from which they are made to arise in order to pass constitutional — even logical — muster. Stanley v. Illinois, 405 U.S. 645, 92 S.Ct. 1208, 31 L.Ed.2d 551 (1972) (unwed father presumed unfit to rear his child; invalidated); Bell v. Burson, 402 U.S. 535, 91 S.Ct. 1586, 29 L.Ed.2d 90 (1971) (presumption of fault from mere fact of accident); Carrington v. Rash, 380 U.S. 89, 85 S.Ct. 775, 13 L.Ed.2d 675 (1965) (presumption of nonresidence from mere military status); Tot v. United States, 319 U.S. 463, 63 S.Ct. 1241, 87 L.Ed. 1519 (1942) (presumption from mere possession of firearm by convictee of receipt of firearm in violation of statute). The mere numbers that found this one, while doubtless sufficient to prompt further inquiry, are an insufficient foundation, especially in view of Texas demography, for the drastic action that the court ordered and approved on their basis alone. Second, further inquiry should, to comport with elementary due process, incorporate at least the safeguards of notice and the offer of a hearing before such mortal sanctions are applied. Stanley v. Illinois, supra. It may be that GPISD’s status as a local governmental body does not, strictly speaking, implicate the constitutional guarantees invoked in the authorities cited. Interests of federalism, however, dictate that contests between the United States and such entities be as rationally conducted as any others; and in the exercise of 'our supervisory powers we direct the district court to discontinue the use of the procedures discussed above where Mexican-American students in Texas are concerned. A further awkward aspect of the case is presented by the district court’s broad reference to and reliance on extensive stipulations entered into between two allied parties in the case, TEA and the United States. By the time of trial, the state had joined the United States in its attack on GPISD. 12 In an earlier segment of the case and before GPISD was joined, Texas had confessed extensively to discriminating against Mexican-American students statewide, thus providing one essential foundation for the United States’ Keyes -based case, as Keyes was then understood by the district court. Since our understanding of Keyes is different, we pretermit extensive discussion of this issue beyond observing that many considerations may prompt factual concessions by a party litigant. Parties often put others to proof of matters that they believe to be true. Likewise, in hopes of avoiding, for instance, the damaging side effects of evidence that will be offered if the matter be disputed, they sometimes concede things that they do not believe to be so. For these and other reasons, two-party stipulations— especially those between aligned parties— offered in a three-party case should be re *999 ceived, if at all, with great caution and should never be regarded as conclusive. See James v. Wallace at 533 F.2d 963-967 n.7 (5th Cir. 1976). But since the state concession of its own wrongful intent is simply irrelevant to this case, where the ultimate and dispositive issue is that of the intent with which an autonomous body, GPISD, acted in making student assignments, we need inquire into these matters no further here. Clearly, the state lacked power to “concede” anything about the intent of GPISD. Nor did it attempt to do so, or any fact about the circumstances in GPISD. Other Evidentiary Matter Thus, we conclude that much of the evidence discussed and relied on by the district court was irrelevant to the case before it. In view, however, of the court’s characterization of the evidence against GPISD as “overwhelming,” we feel obliged to examine and discuss several categories of that evidence which the district court thought significant. 1. The “No-Spanish” statute. The court viewed the existence of a Texas statute requiring that instruction in all subjects except foreign language be done in English, and its enforcement by GPISD rule, as “facially discriminatory against Mexican-American children” on the part of both the state and GPISD. The GPISD rule added that “English should be spoken in conversation while at school.” It is an interesting question, though one more directly presented in another decision of the court below and only obliquely involved here, whether the Constitution implies anything, either way, about whether ours is to be a society officially monolingual or officially multilingual. See United States v. State of Texas, 506 F.Supp. 405 (E.D.Tex.1981). An appeal from the decision to which we refer is presently pending on our docket and should and will be in no degree prejudged today. Even so, we cannot avoid discussing some considerations that, in our present and relatively uninformed view, suggest that the position espoused by the district court may not be so self-evident as that court believed it. In the first place, what the district court characterized as the state’s “No-Spanish” statute is nothing of the kind. In fact, it was enacted as a “No-German” statute. A glance at the date of its original passage— 1918 — suggests this, and brief research confirms the suggestion: the passage of these statutes was indeed motivated by a xenophobic impulse, but by one entirely different from that assumed by the district court. They were, state archives indicate, directed at Fredericksburg and New Braunfels, not at Brownsville or Laredo. Former Texas Penal Code article 288, repealed in 1969, and Texas Education Code section 21.109, referred to by the district court in its opinion as the “No-Spanish” statutes and both relating to the use of English as the basic language of instruction in Texas schools, find their origin in Texas House Bill No. 128, passed by the Texas House on March 23, 1918, concurred in by the Senate on March 27, and signed into law on April 3 of that year. Acts, 4th Called Sess., 35th Legislature, at 170 (1918). 13 The sponsor of this bill was Rep *1000 resentative Poage of Waco, whose “Full Report” on it, House Journal at 351, appears in the State Archives. 14 The records of the Fourth Called Session are peppered with legislation and other matter pertaining to the European war. But if there be any doubt remaining as to what “foreign” tongue Representative Poage had in mind in introducing H.B. 128 on March 13, 1918, it is laid to rest by the terms of his resolution “Relating to Teaching German in Public Schools,” offered only nine days before he introduced H.B. 128. We set out excerpts' from it in the margin. 15 Texas House Journals of the time record other pathetic matter relating to German-Americans in Texas 16 but none suggesting the faintest intent to enact a “No-Spanish” law. Nor does our research reveal any later legislative history in connection with the *1001 few and minor amendments to the penal article derived by the codifiers from House Bill 128 that lend support to the notion that an anti-Spanish bias motivated any of these; rather the contrary. By an amendatory act in 1927, the legislature added a proviso specifically permitting the teaching of all modern languages in high schools and of Spanish to elementary students in border counties. 17 By a 1933 amendment, the permission granted was broadened to allow teaching of all modern languages, including Spanish, throughout the state in all grades above the second and, in border counties, in all grades. 18 The same is true of the corresponding Education Code article. As for the parallel rule of GPISD, this did little more than restate a binding provision of the state laws that we have discussed, one that already applied of its own force in GPISD as it did everywhere in the state. Nor is it accurate to speak, as does the brief of the United States, of the retention and enforcement of the rule by GPISD “despite the 1969 repeal of the state statute requiring such a rule.” The statute referred to was not repealed, it was merely amended to permit a degree of bilingual education. Ever since then, and even at the time of trial, it commenced with the sentence: “English shall be the basic language of instruction in all schools.” 19 And so, in closing this topic, we find the district court’s characterization of these statutes as “No-Spanish” ones, and its derivation of an anti-Mexican-American bias from them, overdrawn. Doing so requires the assumption that what commenced as a “pro-English” provision, deriving its initial impetus from a desire on the part of the state legislature to inhibit the conducting of general school instruction in the German language, was somehow transmuted over the years into an anti-Spanish one. Neither the evidence in the record of this case nor the historical public record supports that assumption. All are at least as consonant with a conclusion that both the state and the school authorities thought it best that, in a society predominantly English-speaking, it was in the best interest of all students to learn to function in English. Such a policy is not necessarily invidious, and it is certainly not irrational. And though we would be foolish to maintain that, in the context of GPISD, the policy did not have its main impact on students whose cradle *1002 language was Spanish, it is precisely these who might benefit most from it. The district court’s reasoning on this topic assumes its conclusion and does not bear scrutiny. Finally, the policy, having been applied generally throughout GPISD, can scarcely have had any effect on residential patterns there or on the student assignments that are the subject matter of this case. Mexican Schools in GPISD Thirty years ago the former school districts that became GPISD operated schools called “Mexican schools.” These seem to have been chiefly first- through third-grade facilities designed to teach younger Spanish-speaking students to function in English at school. As with the “Speak-English” rule, since the record shows that these schools existed in both former districts, it is unlikely that they had any effect on residential patterns or to produce the ethnic disproportion at Gregory. The evidence indicates none, and the court made no such finding. Little or nothing appears in the record regarding the long-lost intent with which they were established in this school area. For all the record reveals, they were early attempts to pursue the policies underlying modern-day bilingual education. After the third grade, usually, students from them entered freely into the regular local system. GPISD has never operated any such school or any on the site of a former one. 20 The circumstance that they once existed in this area, operated by the District’s forerunners, proves little or nothing about the motives with v/hich they were established or about those upon which GPISD acted thirty or forty years later. Hiring and Assignment of Faculty The district court found that GPISD pursued prejudiced policies in faculty recruitment and assignment. United States v. State of Texas, 498 F.Supp. at 1372. The record contains little or no evidence supporting such a conclusion and overwhelming evidence to the contrary. GPISD has twenty percent Mexican-American teachers. This represents a ratio twice that of the ratio of Mexican-American teachers throughout Texas. One elementary school (Austin) principal and the assistant principals at both the junior and senior high schools are Mexican-American. Undisputed record evidence demonstrates that GPISD looks with far greater favor on Mexican-American applicants for teaching positions than on Anglo applicants. Some ratios of Mexican-American applicants employed to Anglo applicants employed are as follows: Percent of Mexican-American Applicants Percent of Anglo Year Employed Applicants Employed 1974 33 14 1975 19 11 1976 12.5 5 1977 12 10 1978 44 18 1979 32.5 19 Whatever conclusion about the effect of ethnicity on GPISD faculty hiring such figures may support, it is certainly not one of prejudice against Mexican-American applicants. 21 Indeed, Dr. (former State Senator) Joe Bernal, an expert witness presented by GPISD’s adversaries at trial, confirmed GPISD’s evidence that in Texas competition to hire Mexican-American teachers is keen and, apparently surprised, commented that to have 20 percent Mexican-American teachers with a 34 percent Mexican-American student body, as does GPISD, is “very good.” As for faculty assignment policies, it is true, as the district court observed, that in the early years of GPISD its small numbers of Mexican-American teachers were assigned in disproportionate percentages to *1003 the Gregory school. Since then the percentage assigned there has steadily declined to a present one of about one-quarter of the total GPISD faculty. For several reasons, we do not attach to this circumstance the significance found in it by the trial court. In the first place, the proportion of such teachers employed by GPISD to total faculty in the earlier years was, for reasons noted above, not large. Hence, percentage figures derived from it do not show much. For example, to say, as can truthfully be said, that in 1957 one hundred percent of the Mexican-American faculty was assigned to Gregory sounds significant until one reflects that this represents one teacher out of fifteen assigned to Gregory, from a total of fifty-seven teachers in the whole system. It is equally true to say that in 1965 only 25 percent were assigned there and in 1969 only 18 percent; however, these more favorable percentages, based as they are on absolute numbers, respectively, of 1:4 and 2:11 signify little more. As larger numbers of Mexican-American faculty entered the GPISD system, rising steadily from a total of 13 in 1970 to 51 in the 1979-80 school year,, the percentage of those assigned to teach at Gregory to total faculty varied from a low of 17 percent in the 1972-73 academic year to a high of 28 percent in that of 1977-78, averaging 23 percent and stabilizing at 25 percent for the two most recent academic years at the time of trial. In those same last two years, almost equal percentages of total Mexican-American faculty were assigned to teach in the junior high school, the figures being respectively 21 percent and 20 percent. As for the ethnic mix of local faculty at Gregory, the highest percentage of Mexican-American faculty to total local faculty that taught at the Gregory school in any earlier year was in 1959-60, when among a total of 17 teachers assigned there were 4 Mexican-Americans, for a faculty percentage of 23.5. At present, the faculty at Gregory totals 39, of whom 13 are Mexican-American. We are unable to see that such figures make out a case for the notion of a segregated general faculty in GPISD. Nor do they support the notion of a “racially identifiable” one at any facility, at any time. Finally, even granting, as we must, a somewhat disproportionate assignment of Mexican-American teachers to the Gregory facility, especially in early years, we see nothing malignant presented. It is not necessarily true, in Texas today, that all Mexican-American school teachers speak Spanish or that all Anglo teachers do not, though the assumption that all are fluent in English is a reasonable one. But even the first assumption was probably generally correct in' 1957, when GPISD assigned its first Mexican-American teacher to the Gregory elementary school. We see nothing inherently unreasonable or invidious in assigning to a facility having a large percentage of students striving to adapt to instruction in English, and perhaps currently more fluent in Spanish, a teacher able to assist in both languages. 22 Nor did the state’s expert, Dr. Bernal, a distinguished Texan of Mexican-American ancestry. In his view, as he testified as an expert witness adverse to GPISD: Q You said that one of the problems was that the Mexican-American children were given the wrong teachers who didn’t speak the language. Do you recall that? A Yes, that the teacher needed to be compatible with the children in the community in which they serve. Q What does that mean, to put the Spanish-speaking teacher in the school with the most Spanish-speaking children? A No, if you’re talking about Mexican-American children, that the teacher understand the language and that the teacher would have taken courses say in bilingual courses, so that she could understand the culture of the commu- ■ nity. The ethnic question takes a secondary position in that regard. *1004 Q So it is the language that is most important? A Understanding of language and culture and being able to use that in teaching. Q So you would find no fault with putting a Mexican-American teacher at a school that had a large number of Mexican-American children? A No, unless you found all the Mexican-American teachers in the Mexican school and all the Anglo teachers in the Anglo school, or if you had a great imbalance there, I would surely find you suspect. As we have observed, no such suspect situation existed anywhere in GPISD at any time later than the 1959-60 school year, when GPISD’s then four Mexican-American teachers were all assigned to Gregory, along with thirteen Anglo ones. Indeed, looking at the evidence of what GPISD has done in faculty hiring and assignment, we find it clear that GPISD has done about what Dr. Bernal thought right. We conclude that the record entirely fails to support the district court’s view of these matters and compels the opposite conclusion. But even were the district court’s conclusions correct ones in this respect, here, as in the case of the earlier categories discussed- — the English language rule and the long-vanished Mexican schools — no connection is shown between GPISD policies as regards faculty hiring or assignment and the concentration of Mexican-American students at Gregory. Ability Grouping in Classrooms; Miscellaneous Other Evidence In the 1950’s and early 1960’s at the Gregory school and elsewhere in the system, grades were ability grouped, with the result that most of the Anglo children were in high sections along with proportionately fewer Mexican-American ones, while second and third sections were largely, sometimes even exclusively, Mexican-American. Another factor contributed to the occurrence of such all-Mexican-American sections: in those times, far more so than at present, Mexican-American families from the area followed the grain and cotton harvests, returning home near Thanksgiving or Christmas. When these -children re-entered school, special sections were created to attempt to compensate for the instruction lost in their three- to four-month absence. An expert witness appointed by the court testified that such ability grouping was “all the rage” during that time, a general fad among educators. There was no testimony that ethnic grouping in the sections resulted from any discriminatory purpose. It was done by the local faculty and principal at each school, and if their intent was to discriminate, it is difficult to explain the presence of mixed sections. Since the groupings were done locally, by school faculty, and generally across GPISD, it is plain that they could have had no effect on residential patterns or student ethnic percentages as between schools. The same is true of another related item of evidence that the district court found significant: that students transferring into the district from other districts and from parochial schools did not always receive full credit for work done there. Lacking evidence that Anglo transferees from such schools were treated differently from Mexican-Americans, or that such treatment was unwarranted, we find this of little significance. Finally, the district court referred to several incidents that, as it noted, “taken individually or in isolation, might prove little.” United States v. State of Texas, 498 F.Supp. at 1373. One was a question asked by an Anglo Board member of a Mexican-American one, Mr. Leonel Rios, before the construction of the new school at Gregory in 1953: “ ‘Leonel, do you believe or do you think we might run into a problem here if we kind of ignore or leave the elementary school here without air-conditioning?’ But first he asked me, ‘How many people in Gregory have air-conditioned homes?’ ” To the court’s ear, this colloquy, quoted from Mr. Rios’ testimony, became a suggestion “that Austin Elementary School need not be air conditioned because the Mexican-American children attending Austin Ele *1005 mentary School were not used to air conditioning in their homes.” Id. Given the fact that the Board members were at the time engaged in budgeting bond election funds for Austin and other school needs, that Mr. Rios was Mexican-American, and that at the time 21 percent of the students at Gregory were Anglo, this seems to us a somewhat severe reading of a casual inquiry almost 20 years gone by. Mr. Rios disagreed, and the school was air conditioned. Another incident was the refusal of the Board, at a meeting held in 1953, to recognize a 14-year-old Mexican-American student who wished to speak, while recognizing an Anglo adult later on in the meeting. Finally, it was shown, and the record makes clear, that on more than one occasion the athletic director at Gregory-Portland High School indulged in offensive ethnic characterizations — mainly of players on an opposing team or teams, but sometimes in goading on his own players as well. The record does not indicate what, if anything, his superiors did to correct this highly distasteful behavior, nor do we in anywise condone it. The same witness, a Mexican-American youth who played as first-string quarterback in his last two years there, complained as well that ten or eleven of the defensive unit were Mexican-American but only three on the offensive unit. The record does not explain the significance of this and, as those who in our long-past football experiences played both ways, it eludes us. We agree with the district court that those incidents, spread over a 15-year period, do not amount to much. Except for the coach’s repeated and inexcusable use of ethnic slurs, it is at worst ambiguous whether they evidence ethnic prejudice at all. Finally, none are of such a nature as to have had any impact on ethnic distribution of students in GPISD. Discussion and Applicable Law Having thus sifted the record evidence and reviewed the trial procedures adopted by the district court, we turn now to the dispositive issues in the ease and to the law that governs them. As we shall see, despite our necessarily lengthy preamble, by this point they have become surprisingly simple. It is apparent from the record evidence— indeed it is without significant dispute there — that the existence of an ethnically identifiable student body at Gregory elementary school is the natural consequence of a corresponding Mexican-American population concentration in this somewhat distinct and isolated part of the district. This condition, the record makes clear, was not caused and has not been contributed to by any governmental action whatever, state or local, but is rather the product of indifferent historic and demographic forces. In 1950, at the time each district voted to consolidate, Gregory and Portland, the district’s two population centers, simply happened to have different ethnic mixes in their populations. There is nothing intrinsically sinister in that circumstance, and nothing in the record suggests that it was the product of any government action, well or badly intended, whatever. This being so, the GPISD Board lay under no constitutional duty to “integrate” either school, and its decision to continue each as a community school for younger students was not a violation of the Constitution. As the Supreme Court wrote in Dayton I: The finding that the pupil population in the various Dayton schools is not homogeneous, standing by itself, is not a violation of the Fourteenth Amendment in the absence of a showing that this condition resulted from intentionally segregative actions on the part of the Board. 433 U.S. 406, 413 (1977). 23 At the time of merger, student populations at Gregory and Portland were, roughly speaking, mirror images of each other as to ethnic mix. Since then this ethnic dis *1006 proportion has increased, but again, the record is devoid of evidence that anything done by the Board or other governmental authority caused or contributed to these developments. Indeed, the court-appointed expert testified that the Gregory school was the second-best one in the district, “but only by a nose.” He elaborated: Clark [in Portland] wins first place, Austin [at Gregory] wins second place and we were tremendously impressed by what we saw going on at Austin and by what we heard from Mr. Barrera. Austin, in our opinion, under Mr. Barrera could easily become first place next year. So, again, it’s just a nose difference between Clark and Austin. In short, all the Board did was leave the attendance zone line between Gregory and Portland where it was in the face of a widening ethnic percentage difference between them. Whether doing so was wise or unwise as a matter of policy is not for us: In seeking to define even in broad and general terms how far this remedial power extends it is important to remember that judicial powers may be exercised only on the basis of a constitutional violation. Remedial judicial authority does not put judges automatically in the shoes of school authorities whose powers are plenary. Judicial authority enters only when local authority defaults. 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. Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 16, 91 S.Ct. 1267, 1281, 28 L.Ed.2d 554. Nor was it for the district court. But wise or unwise, in the circumstances presented by this record, there can be no doubt whatever that the Board’s inaction was not a violation of the fourteenth amendment. The ethnic imbalance having arisen without segregative intent or even governmental involvement, the Board was under no duty to alter it. Thus, its failure to do so furnishes no basis for intervention. To find violation of a duty, one must first find a duty; none existed here. Nor was the Board’s decision to build new elementary schools in Portland, where the increasing numbers of young students there could walk to class, rather than in the vacant farmlands stretching northward for four miles toward Gregory, a constitutional violation. Having done nothing to cause or influence the disproportionate population increases and ethnic trends, having done no gerrymandering, and there being obvious and neutral reasons for building schools where it did, the Board did nothing constitutionally infirm in this respect. As for the other evidence that we have discussed and that the district court thought significant — the English language law and rule, the one-time Mexican schools, the coach’s ethnic slurs, and the like — none of it could have had any causative effect to isolate Mexican-Americans at Gregory, since all of it, including the ethnic slurs at the district’s unitary high school, operated equally in all parts of the district. Since whatever effect they may have had was of such a broadcast nature, we are unable to see how that effect could have had any causal connection or nexus with ethnic concentrations in areas of the district. Having none, it cannot serve as a basis for altering them artificially by busing. Milliken v. Bradley, 418 U.S. 717, 750, 94 S.Ct. 3112, 3130, 41 L.Ed.2d 1069 (1974). And while the existence of such things justifies a close inspection of the Board’s actions, that inspection reveals nothing upon which to base a finding of unconstitutionality in student assignments. And so we come to the end of our inquiry. “As with any equity case, the nature of the *1007 violation determines the scope of the remedy.” 24 Since no action of the Board or other governmental authority caused or contributed either to the existence or the intensification of ethnic disproportion in GPISD, the Board had no duty to alter it. Its actions, at most, recognized local population trends and dealt with them, after the fact. There having been no segregative violation, no integrative remedy was justified. The evidence being clear and virtually undisputed to this effect, we reverse the judgment below except as noted in the margin 25 and render judgment here in favor of Gregory-Portland Independent School District and against the United States and the State of Texas. REVERSED and RENDERED. Footnotes [1] In 1973, out of a total enrollment of 3,656, there were only two Negro students in the *992 system. Twice as many students — four — were of Oriental origin. Even in 1980, at the time of trial below, there were only eleven black students in the consolidated district out of a total enrollment of about 4,400 scholars, a numerical percentage of less than one quarter of one percent. The corresponding percentage of Mexican-American students was approximately 34 percent, that of Anglo students about 66 percent. Numerically, then, our context is essentially a bi-ethnic one. As the court’s appointed experts reported: “the only identifiable minority resident in the district in numbers of consequence is the Mexican-American.” [2] § 23.26. In General (a) The trustees shall constitute a body corporate and in the name of the school district may acquire and hold real and personal property, sue and be sued, and receive bequests and donations or other moneys or funds coming legally into their hands. (b) The trustees shall have the exclusive power to manage and govern the public free schools of the district. (c) All rights and titles to the school property of the district, whether real or personal, shall be vested in the trustees and their successors in office. (d) The trustees may adopt such rules, regulations, and by-laws as they may deem proper. Tex.Educ. Code Ann. § 23.26 (1972). [3] The trial court found that it was a desire to share in ad valorem tax revenues deriving from the location of this plant that caused Portland officials to favor merging the districts. This does not account, however, for the parallel vote to consolidate in the Gregory district. Record evidence suggests that a general desire to insure enough students to retain an accredited local high school was at least an additional motive. At all events, segregative intent on the part of the predominantly Anglo Portland district cannot have been a factor, since the consolidation of the districts necessarily produced an integrative, not a segregative, effect. [4] There was evidence that Gregory remained more of a farming-oriented community, while Portland tended toward becoming a “bedroom suburb” of Corpus Christi. [5] This disproportion in schools per number of residents is not reflected in the number of students per school. At the time of trial, the number of elementary students at the Gregory school was 538, while numbers at the Portland schools were 491, 608, and 716, or an average of 605 per school. An explanation may be suggested by 1970 country-wide census figures showing that the median age of Hispanic citizens was then 18, while that of Anglos was 34. [6] Other discrete facts of state and district background that the court below thought significant will be raised and discussed hereafter: principally the “Mexican” schools, the so-called “No-Spanish rule,” ability grouping, and faculty assignment policy. [7] We are uncertain as to the manner in which the TEA would proceed in some Texas locales, since the court’s triggering figure is 66 percent and some Texas counties, such as Hidalgo, even-in 1973 had Mexican-American population percentages as high as 79.1 percent. See Castaneda v. Partida, 430 U.S. 482, 486, 97 S.Ct. 1272, 1276, 51 L.Ed.2d 498 (1977). [8] The Keyes Court says so: [W]here . .. the case involves one school board, a finding of intentional segregation on its part in one portion of a school system is highly relevant to the issue of the board’s intent with respect to other segregated schools in the system. This is merely an application of the well-settled evidentiary principle that “the prior doing of similar acts, whether clearly a part of a scheme or not, is useful as reducing the possibility that the act in question was done with innocent intent.” 413 U.S. at 207, 93 S.Ct. at 2697 (emphasis added). [9] That court observed, in its opinion: By failing to make explicit findings concerning the causes of the segregation of GPISD, TEA did not violate the rights of GPISD or any provisions of the 1973 Order. Under the Keyes presumption .. . the presence of segregated schools in the midst of Texas’ statewide de jure segregation is enough to make out a prima facie constitutional violation. TEA’S finding of segregation within GPISD was sufficient; no further inquiry into the cause of the segregation was required of the agency. United States v. State of Texas, 498 F.Supp. at 1367. [10] Preliminary 1980 Census figures indicate that Texas’ entire 20th Congressional District, consisting of central Bexar County, is 68% Mexican-American (Spanish Origin). The 15th District, comprising far South Texas, is at least 77% Mexican-American; and the 14th, where GPISD is located, about 39%. The exact Mexican-American percentages cannot be discerned, since the Census was self-assessing as to five ethnic categories, White, Black, American Indian, Asian, and Other. It is apparent from the figures that most citizens of “Spanish Origin” —a separate category overlapping the five basic ones — classified themselves as White. In the 15th District, for example, out of a total population of 659,265, 542,181 people classified themselves as White and 509,550 as Spanish Origin. Thus, the resulting 77% Mexican-American population figure is doubtful and probably a minimum one, since some Spanish Origin citizens may have classified themselves as Other or even, conceivably, as American Indian. Advance Report, 1980 Census of Population and Housing, U.S. Bureau of the Census (April 1981). [11] The court’s opinion recognizes this, United States v. State of Texas, 498 F.Supp. at 1365, but offers no guidance about what to do in such circumstances. [12] Since then, the state has withdrawn from support of the judgment below and has sought to retract the stipulations. [13] Section 1. Every teacher, principal, and superintendent employed in the public free schools of this state shall use the English language exclusively in the conduct of the work of the schools, and all recitations and exercises of the school shall be conducted in the English language, and the trustees shall not prescribe any texts for elementary grades not printed in the English language; provided, that this provision shall not prevent the teaching of Latin, Greek, French, German, Spanish, Bohemian, or other language as a branch of study in the high school grades as outlined in the state course of study. Sec. 2. Any teacher, principal, superintendent, trustee, or other school official having responsibility in the conduct of the work of the school, and failing to comply with this provision of the law shall be deemed guilty of a misdemeanor and upon conviction thereof in proper court shall be subject to fine of not less than Twenty-five Dollars ($25.00) and not more than One Hundred Dollars ($100.00), cancellation of certificate, or removal from office as the case may be, or both *1000 fine and cancellation of certificate or fine and removal from office. Each day shall be regarded as a separate offense, and it shall be the duty of the trustee, city or county superintendent, or ex-officio superintendent to inspect the schools regularly with regard to the enforcement of this Act and file charges promptly in the court in all cases of violation. [14] It reads: Full Report Having been appointed by your Committee of Education to make a full report on House Bill No. 128 would offer the following, viz: It being made to appear that in many communities of this State, the fixed policy of Texas is being defied and the plain statutes violated in that the free schools of these communities are being actually taught in foreign languages and are actually using the text-books of foreign countries to the ignoring of the text books of this State. As the law now reads no penalty is affixed for this willful violation of its terms and this bill seeks to remedy this evil, and in addition would confine the teaching of foreign languages to the High Schools and those above. We assume that the plain statute referred to is art. 2782 of the 1905 Education Code, to like general effect as H.B. 128 and providing as follows: Art. 2782. Instructions must be given in English. — It shall be the duty of every teacher in the public free schools of this state to use the English language; provided, that this provision shall not prevent the teaching of any other language as a branch of study, but, when any other language is so taught, the use of said language shall be limited to the recitations and exercises devoted to the teaching of said language as such branch of study. Tex.Rev.Civ.Stat. art. 2782 (1911). [15] On March 4, 1918, Mr. Poage offered the following resolution: RELATIVE TO TEACHING GERMAN IN PUBLIC SCHOOLS Whereas, We are now engaged in a war of greater magnitude than all tradition or history tells us against a foe who has adopted many new and terrible methods of warfare, not heretofore used by civilized nations; and Whereas, As our allies we have all of the great Anglo-Saxon races who speak and write the English language — our own mother tongue — the language that was spoken and written by the founders of this incomparable republic. Whereas, We are striving to save and economize in so many ways to enable us to carry to a successful conclusion this war. We are indulging the costly and useless luxury of spending thousands of dollars teaching the language of our German foe— the language that is being used in our country in seditious propaganda to undermine the patriotic efforts of our government to secure world-wide democracy. In many of our high schools the hard-earned money of our taxpayers is being used in the employment of high salaried professors engaged in teaching German. In our University $15,000 is being expended annually for the teaching of German. This unAmerican and unpatriotic practice has already been discontinued in some of the other States. It is high time that kultur of the kind that Germany teaches be cut out; therefore be it Resolved, That all of our public schools in the State of Texas and the University of Texas be requested to discontinue the teaching of German, and that the money now being used for that purpose be hereafter utilized in some manner that will be beneficial in winning the war. Journal of the House of Representatives, 4th Called Sess., 35th Legislature, at 86-87 (1918). [16] In the House Journal, immediately preceding the account of the consideration of Rep. Poage’s resolution — in which it was overwhelmingly defeated — is printed a “Communication from Hon. Rudolph Tschoepe,” of Seguin, in which he says, among other things, that “now to find that I was classed as an alien enemy to the Stars and Stripes — the flag I love so well — has broken my heart.” Mr. Tschope, a former representative, had resigned his position when technically classed as an alien. Id. at 181. [17] Acts 1927, 40th Leg., ch. 188, at 267, providing in part as follows: The fact that under the present law it is unlawful to teach Spanish in the elementary grades in the public free schools of this State and that in counties having cities of over five thousand population bordering on the boundary line between the United States and the Republic of Mexico, a knowledge of the Spanish language is of inestimable value, to the citizens and inhabitants of such counties and cities, and the fact that in order to obtain a speaking knowledge and mastery of any foreign language, it is imperative that instruction in such language be begun at the earliest possible period..... [18] Acts 1933, 43rd Leg., ch. 125, at 325. [19] Before amendment, the statute read, “Except for authorized courses in a foreign language, all courses at all grades in the public schools shall be conducted in the English language.” Tex.Rev.Civ.Stat. art. 2654-ld. After the 1969 amendment, it provided: § 21.109 Language of Instruction (a) English shall be the basic language of instruction in all schools. The governing board of any school district and any private or parochial school may determine when, in which grades, and under what circumstances instruction may be given bilingually. (b) It is the policy of this state to insure the mastery of English by all pupils in the schools; provided that bilingual instruction may be offered or permitted in those situations when such instruction is educationally advantageous to the pupils. Such bilingual instruction may not be offered or permitted above the sixth grade without the express approval of the Texas Education Agency, which approval shall be granted on a three-year basis subject to reapproval at the end of that time. Section 21.109 was further amended in 1973 and at the time of trial read as follows: (a) English shall be the basic language of instruction in all schools. (b) It is the policy of this state to insure the mastery of English by all pupils in the schools; provided that bilingual instruction may be offered or permitted in those situations when such instruction is necessary to insure their reasonable efficiency in the English language so as not to be educationally disadvantaged. [20] A Baptist church is now on the former site of one, a water tower on that of the other. [21] GPISD employed its first Mexican-American teachér in 1957. The former superintendent testified, without contradiction, that he had been recruiting them as far back as he could remember and that they were hard to find. By the 1971-72 school year, GPISD figures matched even present statewide availability figures for such teachers, and they have risen steadily ever since to almost 21% at the time of trial. [22] We are mindful, in so observing, that the exclusive use of English was mandated. The record indicates, however, that in appropriate circumstances the mandate was winked at. [23] Even the existence of entirely one-race schools within a district, and in one that once practiced segregation by law at that, does not necessarily establish segregative intent. Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 26, 91 S.Ct. 1267, 1281, 28 L.Ed.2d 554 (1971). [24] Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. at 16, 91 S.Ct. at 1276. [25] We affirm the judgment of the court below insofar only as it appointed expert witnesses and required GPISD to pay a portion of their fees.