Full opinion text
JOHN W. OLIVER, District Judge, joined by WILLIAM R. COLLINSON, District Judge. I. This case presents for judicial review the third effort made by a Missouri General Assembly since the 1960 decennial census to divide this State into ten congressional districts of as nearly equal population as is practicable, in conformity with the dual mandates of Article 3, Section 45 of the 1945 Missouri Constitution, V.A.M.S. and of Art. I, § 2 of the Constitution of the United States. Application of principles established by Art. I, § 2 of the Constitution, as construed by the Supreme Court of the United States, requires that we hold that this third effort does not pass constitutional muster. The initial decision of this Court, rendered January 4, 1965, determined that the 1961 Missouri Congressional Redistricting Act was unconstitutional but deferred granting any judicial relief “until the Legislature of the State of Missouri has once more had an opportunity to deal with the problem;” this Court refused to presume that “the Legislature of the State of Missouri will refuse to take all necessary action to comply with its duty under the Federal, as well as its own State, Constitution.” Preisler v. Secretary of State of Missouri, 238 F.Supp. 187, 191 (W.D.Mo.1965) (Preisler I). Following Preisler I the Seventy-third General Assembly of Missouri enacted the 1965 Congressional Redistricting Act, Mo.Stat.Ann. Title 9, §§ 128.202-128.305 (1965). That Act was subjected to judicial scrutiny and held to be constitutionally void on the ground that it also failed to comply with the command of Art. I, § 2 of the Constitution. Preisler v. Secretary of State of Missouri, 257 F.Supp. 953 (W.D.Mo.1966) (Preisler II). Our decree in the second case, however, for reasons fully stated in light of Swann v. Adams II, 383 U.S. 210, 86 S. Ct. 767, 15 L.Ed.2d 707 (1966), permitted the 1966 Congressional elections to be conducted under the constitutionally void 1965 Act. We retained jurisdiction for the purpose of reviewing any new Congressional redistricting plan enacted by a future General Assembly and signed by the Governor. 257 F.Supp. at 982. On January 9, 1967 the judgment of this Court was affirmed by the Supreme Court of the United States. Kirkpatrick v. Preisler, 385 U.S. 450, 87 S.Ct. 613, 17 L.Ed.2d 511 (1967). In 1967 the Seventy-fourth General Assembly of Missouri again tackled the problem. Its effort culminated in legislation which became effective on October 13, 1967, hereinafter referred to as the 1967 Act, Mo.Stat.Ann. Title 9, §§ 128.-202-128.305 (1967). This case pends on the Attorney General of Missouri’s motion for approval of the 1967 Act and for dismissal of this case. Plaintiffs contend that the 1967 Act is unconstitutional because the populations of the districts were not “as equal as is practicable,” since the variances between district populations are substantial. The 1967 Act is not challenged on any ground other than failure to comply with Art. I, § 2 of the federal Constitution. The parties who intervened in Preisler II and urged approval of the 1965 Act have appeared and participated in the same role in regard to the 1967 Act. Defendants attached to their pending motion a map of the State of Missouri which illustrated the boundaries of the districts and purported to show the actual population of each district. Defendants’ motion alleged that “all of the population figures referred to [on that exhibit] are based on the United States Census for 1960.” Plaintiffs’ responsive pleading suggested that the population data presented by defendants in their motion did not accurately reflect the 1960 census population data. Procedures were therefore adopted at a prehearing conference under which accurate figures were promptly obtained from the Bureau of Census. The following table illustrates the differences between the population data presented by Appendix B attached to defendants’ motion and that established by the accurate 1960 census figures: Comparison of Population Figures Represented As Accurate in Defendants’ Appendix B With Actual 1960 Census Figures in Evidence District Population Variance Represented Represented Def. App. B Def. App. B Actual Census Population Actual Census Variation 1 436,417 + 4,436 439,746 + 7,765 2 442,302 +10,321 436,448 + 4,467 3 431,507 — 474 436,099 + 4,118 4 423,815 — 8,166 419,721 —12,260 5 430,412 — 1,569 431,178 - 803 6 425,238 - 6,743 422,238 — 9,743 7 436,769 + 4,788 436,769 + 4,788 8 439,984 + 8,003 445,523 +13,542 9 428,223 - 3,758 428,223 - 3,758 10 423,866 — 8,115 423,868 — 8,113 It is apparent that defendants’ Appendix B accurately reflected the actual population of only two of the ten districts (Districts 7 and 9). The variations in regard to two other districts (Districts 5 and 10) are obviously minor. Before the facts concerning the actual population variances were established, defendants’ sole legal contention that the 1967 Act be approved was based on the notion that “the districts resulting from the 1967 Congressional Redistricting Act comply with the doctrine of de minimis in regard to population variances between the districts.” After defendants learned that the actual population variances established by the 1967 Act were in many instances substantially greater than those set forth in defendants’ pending motion, they advised the Court that defendants would like to adduce evidence to attempt to justify the greater variations that had in fact' been established.’ Defendants and intervenors continue to contend that, in any event, the variances actually established, even though greater than those that the 1967 Missouri Legislature believed it was creating, nevertheless should still be ruled to be de minimis. Defendants’ motion was set for hearing before the full panel of this Court as requested by defendants. The evidence adduced at that hearing established that the 1967 General Assembly of Missouri at no time ever considered accurate 1960 federal census population figures in its consideration of the 1967 Act. That evidence further established that the accurate 1960 census figures for all appropriate 1967 political subdivisions were in fact furnished the experienced chairman of the Missouri Senate Reapportionment Committee (the same Senator served in the same capacity in regard to both the void 1961 and 1965 Acts) for his consideration and use. The inaccurate population figures presented by defendants’ pending motion were in fact the inaccurate population figures used by both the Missouri Senate and House in connection with its legislative action that produced the 1967 Act. Chief Judge Biggs’ statement in Sincock v. Duffy, D. Dela.1963, 215 F.Supp. 169 at 194, is apposite. In considering a proposed apportionment plan for Delaware it was found as a fact that “census enumeration districts were included when they should have been excluded, or excluded when they should have been included.” After noting that “to the extent they were erroneously included, the population for the proposed district is overstated,” Chief Judge Biggs said: The erroneous exclusion or inclusion of enumeration districts in representative districts throws population figures askew. When one vote — one person is the aim, it is impossible to hit the bull’s eye with a defectively sighted weapon. The difficulties of attempting to reapportion the State of Missouri with inaccurate population figures is made vividly apparent by, a comparison of the variances created by the unidentified population figures actually used by the chairman of the Senate Reapportionment Committee when he prepared the original of the bill eventually enacted, Senate Bill No. 182, (see pages 27-28 of the Chairman of the Senate Reapportionment Committee’s deposition), with the variances in fact produced by that original bill under the 1960 federal census figures. The following table makes that comparison: Senate Bill No. 182 As Originally Introduced by the Chairman of the Senate Reapportionment Committee District Figures Used by S.B. 182 Author Apparent Variance Actual Census Population Actual Census Variance 1 423,712 - 8,269 408,950 —23,031 2 444,462 +12,481 460,469 a +28,488 3 448,824 +16,843 462,716 +30,735 4 428,113 - 3,868 428,224 - 3,757 5 430,412 - 1,569 434,459 + 2,478 6 427,710 - 4,271 427,710 - 4,271 7 428,173 - 3,808 423,103 - 8,878 8 426,260 - 5,721 429,928 - 2,053 9 427,313 - 4,668 412,249 a -19,732 10 432,281 + 300 431,966 - 15 4,317,260 4,318,774 + 39b 4,318,813 a These figures represent the actual population of the respective .districts as outlined in the Bill, adding the 1960 figures for the new Townships of Spanish Lake (formerly in Florissant) to the 9th District and of Ferguson to the 2nd District. b The actual census figures for wards in St. Louis City as per Stipulated Exhibit No. 18 are 39 less than the total population. The Bureau of the Census has explained that this discrepancy “results from a number of technical factors in the processing of the data.” (Stipulated Exhibit No. 18). The author of Senate Bill No. 182 twice testified that he was still under the impression the Missouri Legislature could comply with the “as nearly as is practicable” standard of Wesberry v. Sanders, 376 U.S. 1, 84 S.Ct. 526, 11 L. Ed.2d 481 by getting within 2% of the equality demanded by Art. I, § 2 (Deposition, pp. 13-14, 18). Defendants’ evidence, consisting of the testimony of the party leaders in both houses of the 1967 Legisla- ture, and the Chairman of the Senate Reapportionment Committee, was simply that many bills were introduced; that the one finally adopted had a stormy passage and was amended many times; and that finally it barely passed the House. Defendants contend that these facts demonstrate that this was the best bill, in respect to practicably equal population, that could be passed, and, therefore, attempt to argue that the division of population provided is, in fact, as “nearly equal as is practicable.” In other words, defendants would have this Court hold that their definition of “practicable” is legally equivalent to “acceptable to a majority of the legislature.” We reject that argument. We find that defendants’ evidence leads to the inevitable conclusion that a majority of the members of the Legislature were motivated by other considerations than the constitutional objective of “practicable equality.” Our specific factual findings in regard to the 1967 Act and in regard to the rejection of better plans by the 1967 Missouri Legislature are fully stated in Appendix A made a part of this opinion by this reference. The legislative history and changes made by the 1967 Act in prior acts held to be constitutionally void are also there stated in detail, together with the facts concerning how the shifts of particular political subdivisions from one 1967 Act district to an adjoining district would have made both districts more nearly equal in population. It is not necessary to restate the constitutional principles applied in Preisler II. That case was affirmed by the Supreme Court on January 9, 1967 and is the law of this case. The same day Preisler II was affirmed, the Supreme Court in Duddleston v. Grills, 385 U.S. 455, 87 S.Ct. 611, 77 L.Ed.2d 508 (1967), vacated the judgment rendered in Grills v. Branigin, S.D.Ind.1966, 255 F.Supp. 155, and remanded that case to the three-judge Indiana court for further consideration in light of Swann v. Adams, 385 U.S. 440, 87 S.Ct. 569, 17 L.Ed.2d 501 (1967), Wesberry v. Sanders, 376 U.S. 1, 84 S.Ct. 526, 11 L.Ed.2d 481 (1964), and Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Swann v. Adams, 385 U.S. 440, 87 S.Ct. 569, 17 L.Ed.2d 501 (1967), also decided January 9, 1967, although a state reapportionment ease, nevertheless established burden of proof principles applicable to all reapportionment eases. Mr. Justice Harlan stated in his dissent, joined by Mr. Justice Stewart, that the burden of proof rule there established “stand [s] on its head the usual rule governing * * * the validity of legislative enactments, state as well as federal, which is, of course, that they come to us with a strong presumption of regularity and constitutionality” (supra, at 447, 87 S.Ct. at 574). Mr. Justice Harlan’s strongly held conviction that “the burden of showing unconstitutionality should be left here, as in other cases, on the attacking party” was, of course, explicitly rejected by the majority of the Supreme Court. In Kilgarlin v. Hill, 386 U.S. 120, 87 S.Ct. 820, 17 L.Ed.2d 771 decided February 20, 1967, Mr. Justice Harlan and Mr. Justice Stewart dissenting, the Supreme Court reversed a Texas three-judge court for the reason that: Under that case [Swann v. Adams III], it is quite clear that unless satisfactorily justified by the court or by the evidence of record, population variances of the size and significance evident here are sufficient to invalidate an apportionment plan. Without such justification, appellants’ analysis of H.B. 195 made out a sufficient case under the Fourteenth Amendment (supra, at 122, 87 S.Ct. at 820). The basic contentions made by defendants in this case were accepted by' the Texas three-judge court only to be rejected by the Supreme Court. The impact of Swann v. Adams III and Kilgarlin v. Hill was immediate and consistent. All three-judge federal courts and state courts deciding congressional apportionment cases, with the benefit of either or both of those cases, have been able, with one exception, unanimously to determine every such case that has been decided this year. See Dinis v. Volpe, D.Mass.1967, 264 F.Supp. 425; Wells v. Rockefeller, S.D.N.Y., 273 F.Supp. 984 (decided May 10, 1967, and not yet reported) ; Baker v. Ellington, N.D. Tenn. 1967, 273 F.Supp. 174; Civ.No. 3945; Gong v. Kirk, S.D.Fla.1967, No. 64-143 Civ.E.C., 278 F.Supp. 133; (decided August 2, 1967); Silver v. Reagan, Cal., 62 Cal.Rptr. 424, 432 P.2d 26, 1967; and Exon v. Tiemann, D.Nebr. decided November 22, 1967, 279 F.Supp. 603. The only split decision was that in Lucas v. Rhodes, N.D.Ohio 1967 (decided May 10, 1967). That case was reversed by the Supreme Court on December 4, 1967, 389 U.S. 212, 88 S.Ct. 416, 19 L.Ed.2d 423. The notion expressed in the majority opinion in the reversed Ohio case to the effect that “an element of ‘practicality’ is implicit in the concept of ‘practicability’ as that term is to be applied in a districting case” was implicitly rejected by the Supreme Court when it reversed that case in its one sentence per curiam opinion which cited only Wesberry v. Sanders. The principle is firmly settled that the proponents of a redistricting plan must sustain the burden of justifying any deviation from practicable equality of population. Swann v. Adams, supra, 385 U.S. at 445, 87 S.Ct. 569, 17 L.Ed.2d 501; Reynolds v. Sims, supra, 377 U.S. at 577-580, 87 S.Ct. 1362; Maryland Citizens Committee for Fair Cong. Redistr. v. Tawes, 253 F.Supp. 731, 733 (D.Md.1966), aff’d sub nom. Alton v. Tawes, 384 U.S. 315, 86 S.Ct. 1590, 16 L.Ed.2d 586 (1966). This Court must also follow and apply Kilgarlin v. Hill’s express requirement that before a particular plan may be approved as constitutional the District Court must be able, from the evidence adduced by the defendant, to articulate legally acceptable reasons for the specific variations in population between districts that have been shown to exceed the “as nearly as is practicable” standard of Art. I, § 2. Both Swann v. Adams III and Kilgarlin v. Hill teach that courts can not refuse to adjudicate bona fide controversies concerning alleged abridgments of the constitutional right to an equal vote. The infrequent judicial use of the phrase “de minimis” in apportionment decisions cannot properly be converted into a generalized rationalization of variances that are shown, on the precise facts presented in a particular case, to have been both avoidable and greater than those variances that would have been produced by a more faithful application of Art. I, § 2’s standard of “as nearly equal as is practicable.” Defendants’ basic de minimis argument is not tenable. The phrase “de minimis” was first used in an apportionment case in Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962). Mr. Justice Harlan there suggested that the variances in the 1901 Tennessee state apportionment were only “de minimis” from an apportionment formula established by the Tennessee Constitution. 369 U.S. at 335-336, 82 S.Ct. 773-774. Mr. Justice Clark countered by stating that Mr. Justice Harlan’s efforts to justify those variances on “such generalities as ‘classic legislative judgment,’ ‘no significant discrepancy,’ and ‘de minimis’ ‘departures’ [are shown by] even a casual glance at the present apportionment * * * to be entirely fanciful” (369 U.S. at 258, 82 S.Ct. at 732). Only three lower federal courts have made reference to “de minimis.” Cal-kins v. Hare, E.D.Mich.1964, 228 F.Supp. 824, used the phrase in explaining and emphasizing why the “mathematical precision” caveat in Wesberry v. Sanders could not be used as “an escape hatch for the reluctant.” (Id. at 829). This Court, in Preisler II, made similar explanatory use of the “de minimis doctrine.” We stated that “use of such words as ‘feasible’ and ‘practicable’ in a careful statement of the fundamental constitutional principle constitutes but a recognition that the familiar doctrine of de minimis is applicable and is designed to make clear that a State legislature would not be expected to create entirely new political subdivision lines in order to have absolute and precise mathematical equality in its congressional districts” (257 F.Supp. at 973). The third lower federal court case to mention “de minimis”, however, attempted to give that phrase a definitive and fixed conceptual meaning. That case was Kilgarlin v. Hill, S.D.Tex., 252 F. Supp. 404. The majority opinion in that case attempted to create a burden of proof formula under which (1) “de minimis” variations were to be placed at one pole and considered to be constitutionally permissible, (2) “per se” variations were to be placed at the opposite pole and held to be constitutionally void, and (3) all cases in the middle area of the mathematical scale would be held to be constitutionally permissible unless the plaintiff was able to “negate the existence of any state of facts which would sustain the constitutionality of the legislation” (252 F.Supp. at 414). That case was reversed by the Supreme Court. In reversing the Supreme Court simply held that the burden of proof rule established by Swann v. Adams III, which had not been decided when the Texas three-judge court attempted to establish a different burden of proof rule, was controlling. The Supreme Court did not deem it necessary even to discuss the elaborate lower court formula that included “de minimis” as a conceptual part of its involved legal theory of percentage and ratio justification. In fact, the only other mention of the de minimis phrase by the Supreme Court, other than' that noted in the concurring and dissenting opinions in Baker v. Carr, was in Swann v. Adams III, supra, 385 U.S. at 444, 87 S.Ct. 569. Mr. Justice White’s use of “de minimis” was. in precisely the same pattern as the explanatory use made of that phrase by Calkins v. Hare, and Preisler II. He simply stated that “mathematical exactness or precision is not required” because “de minimis deviations are unavoidable” (385 U.S. at 444, 87 S.Ct. 569). The lower federal court reversed by Swann v. Adams III had not even mentioned the words “de minimis” in its opinion. It is therefore obvious that no question concerning the phrase was presented to or placed in focus for Supreme Court determination by the appeal from that lower court decision. We believe that the Supreme Court’s reversal in Kilgarlin v. Hill is at least an implicit rejection of any notion that the judicial use accorded the phrase “de minimis” can be converted into either an exception to the burden of proof rule so recently established by the Supreme Court in apportionment cases or as an independent defense that may, without proof, be said to justify variances greater than those that would have been created by better plans in fact presented to but rejected by a legislature or those created by similar plans that may be presented to a court in the trial of an apportionment case. We know of no court, state or federal, with the sole exception of the reversed Texas three-judge court, that has ever expressed a contrary idea concerning the very place that the “de minimis” phrase occupies in the constitutional law applicable to apportionments. “De minimis,” we suggest, belongs where Mr. Justice Clark put it in his concurring opinion in Baker v. Carr; among discarded generalities such as “classic legislative judgment” and “no significant discrepancy.” So long as counsel representing State legislatures attempt to get courts to convert the infrequent judicial use of the phrase “de minimis” into some sort of numerical formulae under which ratios or percentage deviations fixed in one case become the rule in another for purposes of an attempted measurement of “how much variance can we expect to get away with,” the whole thrust of the Supreme Court’s apportionment cases is misconceived. Indeed, the whole thrust of many earlier Supreme Court cases, decided before the apportionment cases were decided, is ignored. Gray v. Sanders, 372 U.S. 368 at 379-380, 83 S.Ct. 801, at 808, 9 L.Ed.2d 821, held long before either Reynolds v. Sims and Wesberry v. Sanders were on the books, that “the concept of ‘we the People’ under the Constitution visualizes * * * equality” and that “the idea that every voter is equal to every other voter in his State * * * underlies many of our decisions.” All the Supreme Court’s apportionment cases from Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962) to the present time teach that a particular apportionment case must be held to present a bona fide justiciable controversy and that a congressional apportionment plan cannot be approved by a lower court as constitutional unless the variances above those which result mathematically from the application of the “as nearly as is practicable” standard did in fact result from and are justified by the application of a legally acceptable principle of rational State policy. The burden of proof to so establish rests upon the defendant. When it becomes apparent, therefore, from the facts presented in a particular case, that the apportioning body could have districted the particular state in a substantially more nearly equal manner, without doing violence to any articulated legally acceptable state policy, consideration of de minimis is at an end in that particular case. Kilgarlin v. Hill based its reversal on the dual ground that the lower court had not applied the proper burden of proof rule and upon its failure to “articulate any satisfactory grounds for rejecting at least two other plans presented to the court, which respected county lines but which produced substantially smaller deviations” (386 U.S. supra, at 124, 87 S. Ct. at 823). The undisputed and unexplained facts of this case show (a) that the 74th Missouri General Assembly did not in its 1967 Act in fact provide districts which “as nearly as is practicable one man’s vote for a congressional election is * * worth as much as another’s;” (b) that the 1967 Missouri Legislature in fact rejected plans that came much closer to so doing; and (c) that any number of shifts of political subdivisions including counties, to say nothing of townships, wards and precincts, would have produced a congressional districting plan that would have been in closer compliance with the constitutional command of Art. I, § 2. In Preisler II we took detailed notice of more practicable plans that had been proposed to and rejected by the 1965 Missouri General Assembly. The 1967 74th General Assembly enacted a bill that malapportioned this State in a manner more extreme than the plan proposed to the 1965 Legislature as Senate Committee Substitute for Senate Bill No. 320, to which this Court devoted detailed attention in Preisler II. Table IV on page 964 of 257 F.Supp. reflects a comparison of the changes proposed by that rejected 1965 plan with those finally enacted by the 1965 Legislature. The 1967 Legislature’s action in enacting its 1967 Act was taken in the face of the express caveat stated in Preisler II that our discussion of the better plan rejected by the 1965 Legislature was not to be construed by a future Missouri General Assembly as an approval of the rejected plan but only to show that the 1965 Act did not in fact comply with the constitutional standard. The fact that the 1965 Legislature could have adopted a more practicable plan with districts more nearly equal in population had it not rejected Senate Committee Substitute for Senate Bill No. 320 was demonstrated by our discussion in Preisler II. The fact that the 1967 Legislature could have done the same thing demonstrates that its 1967 Act was not a plan that represented the most faithful application of the “as nearly equal as is practicable” standard. The rejected 1965 plan obviously would have produced variances that departed less from the ideal than those created by the 1967 Act. Our discussion of the rejected 1965 plan is not to be taken by a future Missouri Legislature as any intimation that such rejected plan, if adopted, would pass constitutional muster. Determination of that question would turn upon the sufficiency of the evidence adduced by defendants to justify and explain the quite obvious and substantial variances contained in that plan. The 1967 Missouri Legislature rejected a better plan than that presented to the 1965' session of the Legislature. House Bill No. 870, proposed to and rejected by the 1967 Missouri General Assembly, provided districts that varied substantially less from the ideal than any bill ever before proposed to any Missouri Legislature. It is also apparent that particular districts containing populations much closer to the ideal than those eventually created by the 1967 Act were proposed by earlier versions of Senate Bill No. 182, as that bill progressed toward final passage, and by still other bills, all of which were rejected by the 1967 Missouri Legislature. Our detailed findings concerning House Bill No. 870 and the other better plans rejected by the 1967 Legislature that demonstrate how Missouri could have been divided into ten districts of more practicably equal population than those provided in the 1967 Act are made in Appendix B which is incorporated in this opinion by this reference. We find that the variances produced by those better plans are in fact substantially less than those created by the 1967 Act. Our discussion of those better plans is solely for the purpose of demonstrating that the 1967 Act was not in fact the most practicable plan that could have been enacted. We intimate no judgment on the question of whether House Bill No. 870 or any of the other plans discussed would or would not pass constitutional muster for the reasons that have been sufficiently stated. The prime difficulty with all three of the past reapportionment efforts of the respective Missouri Legislatures in the years 1961, 1965, and 1967 has been that the task has not been approached with an understanding that, since Wesberry v. Sanders, no State legislature in the United States can unnecessarily abridge one’s right to an equal vote in a congressional election any more than it may constitutionally deny one his right to trial by jury, his right of free speech, his right to freedom of religion, his right of peaceful assemblage, or any one of his many other rights guaranteed by the Constitution of the United States. Indeed, the discharge of this delicate and quite fundamental constitutional duty has been treated by three successive Missouri legislatures as a legislative step-child; apparently qualfication for appointment to the committees which have unsuccessfully attempted to do the actual work seems to have been limited to those who insist upon believing that something less than full performance of the dual mandates of the Missouri and federal Constitutions is all that is required. It was explicitly held in Wesberry v. Sanders that “Our Constitution leaves no room for classification of people in a way that unnecessarily abridges” their right to an equal vote in a congressional election (376 U.S. at 17, 84 S.Ct. at 535. Legislation implementing the command of Art. I, § 2 must divide the population into districts in which “as nearly as is practicable one man’s vote in a congressional election is to be worth as much as another’s” (376 U.S, at 7-8, 84 S.Ct. at 530). Those who have been responsible for the drafting of Missouri’s three congressional redistricting plans have adamantly refused to recognize that these constitutional principles are as binding on all State legislatures of the United States as upon all State and federal courts. The facts established in the three cases decided by this Court demonstrate that in all three legislative efforts the leadership of both political parties in the Senate and the House were given nothing better to work with than a makeshift bill produced by what has been candidly recognized to be no more than that deemed to be an expedient political compromise. And even that was not furnished until the closing hours of the session when nothing could have been done to either discover or correct even obvious errors. This Court repeats for the third time that the Constitution does not leave room for intentional 85%, 95% or even 98% compliance with its mandate; that Art. I, § 2 commands equality of population in congressional districts “as nearly as is practicable;” and that any enacted plan that fails to comply with that constitutional standard, unless otherwise justified by substantial evidence in accordance with applicable law, is constitutionally void. This Court obviously has no alternative other than to deny defendants’ pending motion for the reasons that (1) we must find that the 1967 Act does not in fact divide the population of Missouri into districts as nearly equal as is practicable; and (2) because the obviously avoidable population variances in the 1967 Act can not be satisfactorily justified on any legally acceptable ground supported by any evidence in this record. It is therefore totally impossible for this Court, on the basis of any substantial evidence, to relate any “declared justification to any specific inequalities among the districts” or to “articulate any satisfactory grounds for rejecting * * * other plans presented [both] to the court [and to the 74th General Assembly of Missouri] which produced substantially smaller deviations” than those created by the 1967 Act. Kilgarlin v. Hill, supra, 386 U.S. at page 124, 87 S.Ct. 820, 823. Defendants have simply failed to carry the burden of proof imposed on them by Swann v. Adams III, a case decided before Senate Bill No. 182 was introduced in the 1967 Missouri Legislature. We so find and determine. Defendants’ pending motion should be and will therefore be denied. Because of the public interest involved in this, as in all State and congressional apportionment cases, we have attached as Appendix C our discussion of all the contentions presented by either the defendants or by the intervenors that have not been explicitly answered in the body of this opinion. II. There remains the question of remedy. Significant developments have occurred since Preisler II was decided that could in the near future require that our order of August 5, 1966 be appropriately modified. That order, consistent with the applicable congressional statute then in effect required that Missouri’s congressional delegation be elected at large. In Preisler II we directed specific attention to the fact that a “failure of the Missouri General Assembly to properly redistrict after the 1930 census forced the election of Missouri’s entire congressional delegation to be held at large in 1932” (257 F.Supp. at 955). We also noted that Missouri had been fully advised “over thirty years ago that a Missouri Genera] Assembly’s failure to re apportion congressional districts in accordance with law required that ‘unless and until new districts are created, all Representatives allocated to the state must be elected by the state at large’ ” (257 F.Supp. at 981), quoting from Smiley v. Holm, 285 U.S. 355 at 374-375, 52 S.Ct. 397, 76 L.Ed. 795 (1932). Carroll v. Becker, Secretary of State, 285 U.S. 380, 52 S.Ct. 402, 76 L.Ed. 807, which compelled Missouri’s 1932 congressional elections at large, was decided on the authority of Smiley v. Holm and made application of the congressional statute in effect at that time and still in effect at the time Preisler II was decided. Our order of August 5, 1966 was made, as stated in Preisler II, pursuant to the Congressional “command of Section 2a(c) of Title 2, United States Code, enacted pursuant to Art. I, § 4 of the Constitution of the United States” (257 F.Supp. at 981). That section of the United States Code was enacted as a part of the Act of June 18, 1929 (46 Stat. 26). As we also noted in Preisler II, Wood v. Bloom, 287 U.S. 1, 53 S.Ct. 1, 77 L.Ed. 131 (1932), held that Congress had failed to reenact any legislation requiring the election of Representatives from districts when it refused to include the provisions of the Act of August 8, 1911 (37 Stat. 14) in its Act of June 18, 1929. Indeed, the latter Act, inconsistent with legislation that had required Representatives to be elected from districts since the Act of 1842 (5 Stat. 491), expressly provided that in the event a State failed to properly redistrict after a decrease in the number of Representatives apportioned to that State, the Representatives of such State “shall be elected from the State at large” (Section 2a(c) (5), of Title 2, United States Code). Section 2a(e) did not require an election at large in the event a State failed to reapportion in cases in which the State either gained or stayed the same in the number of Representatives apportioned to it after a decennial census. It was for these reason that federal three-judge courts in Maryland and Tennessee, for examples, were not prevented by congressional mandate from redistricting those particular states by court order. The Supreme Court’s reversal of the Indiana three-judge court in Grills, and the affirmance of Preisler II last January, of course, focused Congressional attention on the fact that the Act of June 18, 1929 was indeed a Congressional command to the federal three-judge court in Indiana and to this Court to order elections-at-large in the event the Legislatures of Indiana and Missouri should fail to pass a constitutional redistricting act in time for the 1968 elections. The order of this Court, of course, expressly so provides. Recent Congressional developments reveal that long standing efforts to reenact the 1842 requirement for district elections in the present Congress seemed doomed to the same failure suffered by similar efforts in every Congress over at least the past fifteen years. But, most recently, the 1842 requirement was successfully attached as a rider to a private immigration bill for the relief of one Dr. Ricardo Vallejo Samala, 81 Stat. 581 (H.R. 2275, 90th Cong., 1st Sess.) The rider, in the precise language of the Act of June, 1842, provides that “there shall be established by law a number of districts equal to the number of Representatives to which such State is so entitled, and Representatives shall be elected only from districts so established, no district to elect more than one Representative * * * ”. That private bill, with its attached rider, passed the House on November 28, 1967, passed the Senate on November 30, 1967, and became law on December 14, 1967 when it was signed by the President. See Congressional Record: November 28, 1967, H15901 and November 30, 1967, S17458. When that legislation became law, this Court was relieved of the prior existing Congressional command to order that the 1968 and succeeding congressional elections in Missouri be held at large. This Court, therefore, will be free, in the event no constitutional plan is enacted at the coming Special Session of the Missouri Legislature, to direct appropriate proceedings that will enable it to make an appropriate redistrieting order. In the event it should become necessary for this Court to exercise that power we shall follow procedures similar to those that were followed by other three-judge federal courts that have drawn congressional districts for particular states. Judge Matthes, in his concurring opinion in Priesler II, stated that: We have emphasized that the responsibility for enacting a constitutionally permissible redistricting plan rests on the Legislature of Missouri. Although the task may be fraught with difficulties, certainly it is not an impossible one. We are convinced that a conscientious effort, motivated by a desire to satisfy the requirements of the federal and State constitutions, will result in the formulation of congressional districts which will comport with the demands of the Constitution of the United States and the Constitution of Missouri (supra, 257 F.Supp. at 982). That statement continues to reflect the view of this Court. We would be hopeful that all appropriate steps will be taken by all responsible persons to the end that a constitutionally valid redistricting plan be presented to this Court in order that it not be required to take further action in this case. Should it become apparent that further action is necessary, we shall, as we must, and on our own motion, direct further appropriate proceedings in this case. For the reasons stated, we find and determine that the 1967 Missouri Congressional Reapportionment Act presented by defendants’ pending motion does not pass constitutional muster. It is therefore Ordered that defendants’ pending motion should be and is hereby denied. It is further Ordered that the judgment and decree of this Court entered August 5, 1966 should be and are hereby ordered to remain in full force and effect, subject, however, to future appropriate modification. It is further Ordered that any filings for the House of Representatives of the United States Congress that may have already been made under and pursuant to the 1967 Missouri Congressional Reapportionment Act should be and are hereby declared to be null and void, that no further filings be accepted under that Act, and that any and all future action that may be taken in connection with any filings that may have been made should be and will be considered a violation of the order of this Court. It is further Ordered that pursuant to Rule 56(c) of the Rules of Civil Procedure, this opinion and the appendices attached thereto shall serve as our findings of fact and conclusions of law. It is further Ordered that jurisdiction of this cause be retained for the purposes that have been stated. APPENDIX A FACTUAL FINDINGS IN REGARD TO 1967 ACT AND REJECTION OF MORE PRACTICABLE DISTRICTS, INCLUDING PRACTICABILITY OF SHIFTS OF POLITICAL SUBDIVISIONS We follow the pattern of Preisler II in making our specific findings of fact in regard to the 1967 Act. Appendix A attached to Preisler II at page 983 of 257 F.Supp. illustrated the changes made by the 1965 Legislature in the 1961 Act held void in Preisler I. The undisputed facts in Preisler III establish that the changes made by the 1967 Legislature in the 1965 Act held void in Preisler II follow precisely the same pattern followed by the 1965 Legislature in its modification of the 1961 Act. Change “A” on Preisler II's Appendix A showed that Mercer County (pop. 5,750) was shifted by the 1965 Act from District 9 to District 6. The 1967 Act added to that same shift the counties of Grundy (pop. 12,220); Putnam (pop. 6,999) and Sullivan (pop. 8,783) with a total population of 28,002. The 1967 shift of additional rural population from District 9 to District 6 in the 1967 Act commanded still further invasion of the urban population in the St. Louis metropolitan area in exactly the same pattern of invasion established by the 1965 Act. While the overvaluation of District 6 was reduced to -9,743 by the 1967 Act, it is perfectly obvious that had Schuyler County (pop. 5,052) been added to the 1967 shift from the District 9, the variance of District 6 would have been reduced to a variance of -4,691. The inclusion of Schuyler County in the 1967 shift was in fact proposed to the 74th General Assembly in Senate Bill No. 182 as that bill was originally introduced; remained there in the Senate Substitute for Senate Bill No. 182, and was removed from the perfected Senate Committee Substitute for Senate Bill No. 182 only when the shift of Adair and Macon Counties was proposed. Had the plan proposed in the Senate Committee Substitute for Senate Bill No. 182 been adopted District 6 would have had a variance of only +268. Change “B” illustrated in Preisler II's Appendix A stated that the 1965 Act shifted several Kansas City wards from District 4 to District 5. The 1967 Act’s shuffle of Kansas City wards, quite by accident, brought the variance of District 5 to within -803 of an ideal district of 431,981. Such a variance, if it could be separately considered, would apparently be within constitutional range. But congressional apportionment plans can not be viewed piecemeal; they must be viewed as a whole. The 1967 shifting of Kansas City ward population of 25,398 to District 5 required that additional population be added to District 4 which was already 29,168 deficient before the shift. Accordingly, more populous Saline County (pop. 25,148) was shifted from District 8 and St. Clair (pop. 8,421) and Benton (pop. 8,737) Counties were shifted from the deficient District 7; — a total addition of only 42,306. The 1967 shift resulted in a reduction of the minus variance in the 1967 Act’s District 4 to -12,260. But, once again, it is perfectly obvious that had Howard County (pop. 10,859) been included in the shifts made by the 1967 Act the variation of District 4 would have been further reduced to -1,401. The inclusion of Howard County in District 4 was specifically proposed to the 1967 General Assembly along with Saline and other counties in Senate Bill No. 182 as that bill was originally introduced. Had that plan not been rejected and had proper, unquestionably practicable, allocation been made concerning the wards of Kansas City, both Districts 4 and 5 would have had variances of only -639 and -640, respectively. Change “C” on Preisler II's Appendix A illustrated the 1965 Act’s shuffle, of “several St. Louis wards and St. Louis County townships * * * among the First, Second, Third, and Ninth Districts.” The same sort of shuffle was made in the 1967 Act; only more so. Under the 1967 Act St. Louis County was fragmented into' five, rather than four separate congressional districts. The reason why further invasion and fragmentation of the population of the St. Louis metropolitan area was necessary is obvious. In order to improve the 1965 Act’s treatment of District 4, the 1967 Act shifted population formerly in District 8. Similar depletion of District 8’s 1965 allocation of population was rendered necessary in order to bring the rural District 10’s 1967 allocation of population closer to the ideal. The architect of the 1967 Act frankly stated that the attempt of the 1967 Legislature to use the population allocated to District 8 under the 1965 Act to bolster the deficient population of adjoining districts under the 1967 Act was “just like squeezing a sponge” so far as District 8 was concerned (Dep. p. 13). Although he did not know it at the time, the author of Senate Bill No. 182 actually had more water in District 8 than he thought he had. Because he used population figures from a still unidentified source, the Chairman of the Senate Committee was under the erroneous impression that District 8 was being allocated only 8,003 excess population under the 1967 Act. The 1960 census figures establish that an excess of 13,542 population was actually placed in District 8 by the 1967 Act. It is therefore apparent that it was neither practicable nor even necessary for the 1967 Act to have extended the boundaries of District 8 to the city limits of St. Louis. But even if the invasion of the St. Louis metropolitan population could be justified, which is impossible on the facts presented, it is obvious that a shift of Howard County (pop. 10,859) from overpopulated District 8 to either of the adjoining and underpopulated Districts 4 or 6 would have most substantially reduced the variations of those transferor and transferee districts. In further specific regard to the 1967 Act’s shuffle of St. Louis wards and St. Louis County townships we find that the 1967 Missouri General Assembly, like the 1961 and 1965 Missouri General Assemblies, attempted to disregard and ignore basic population data that will not go away. The 1960 census figures make it obvious that the City of St. Louis with its 1960 population of 750,026 needed to have added only 113,936 additional population to entitle that total population to two ideal congressional districts. It cannot be validly argued that it was not practicable for the 1967 Legislature to have created two districts with extremely small variations without violating the integrity of political subdivisions as large as townships, to say nothing of violating ward lines as practiced in the 1967 Act. For example, and only for example, the addition of the populations of the townships of St. Ferdinand (pop. 66,420), Hadley, (pop. 24,720), and Lincoln (pop. 22,380) in St. Louis County to the population of the City of St. Louis would have provided sufficient population for two congressional districts, each of which would have varied only -208 from the ideal. The drawing of the line to divide that 864,378 population into two congressional districts as nearly equal as practicable was in fact an extremely easy task because the 74th General Assembly established no policy that prevented its use of existing precinct lines, at least in the metropolitan geographical areas of Missouri for use in dividing the two districts. Urban precincts, generally regulated as they are by the number and capacity of voting machines assigned them, necessarily have relatively small populations. The ability to shift precincts from one side of a district dividing line to the other makes it entirely practicable to get exceedingly close to equal populations for adjoining districts as was demonstrated in the Jackson and Clay County apportionments to which attention was directed in Preisler II (257 F.Supp. at 978, footnote 16). The Kansas City Star for December 5, 1967 reports that the new boundary lines for the Kansas City School Board divided the population within that school district “into six subdistricts each with about 71,500 population.” Local government public officials in this judicial district have consistently demonstrated that the one man-one vote constitutional principle can be applied without difficulty given the desire so to do. The particular St. Louis townships mentioned have been used solely for illustrative purposes. That use, however, supports our factual finding that it was in fact entirely practicable for the 1967 Legislature to have created two compact and contiguous congressional districts in the St. Louis metropolitan area in accordance with the command of the Missouri Constitution that would have in fact avoided all except exceedingly small variations for two of the congressional districts to which that population was by law entitled. We make the same finding in regard to how a third congressional district could have been formed from the 1960 population situated in the remaining portion of St. Louis County. Such a third congressional district could not have included all of the remaining townships in the only charter county in the State that would have been left unassigned in the example just stated in regard to how two practicable St. Louis metropolitan congressional districts could have been created; the most northerly townships of St. Louis County would necessarily have to have been assigned to still a fourth congressional district, some of which would have included a part of “out-state” Missouri. The practicable third district that could have been made to adjoin the area described for the first two practicable districts, with full observance given an assumed established policy against cut-ling township lines outside of metrópolitan cities, would have included the foilowing townships of St. Louis County: TABLE SHOWING HOW A THIRD PRACTICABLE DISTRICT COULD HAVE BEEN CREATED USING ST. LOUIS COUNTY 1960 POPULATION St. Louis County Township of: 1960 Population Lemay 40,095 Concord 36,273 Gravois 52,205 Jefferson 35,649 Clayton 51,092 Creve Coeur 48,681 Midland 50,417 Washington 19,304 Normandy 37,673 Bon Homme 53,779 Meramac 9,528 Total Population 434,696 Ideal District 431,981 Variance +2,715 Whether the plus variance of 2,715 indicated in the above table would, on the facts, be “as nearly equal as is practicable” would, of course, be entirely dependent upon the proof adduced in any case in which such variance might be challenged. As Chief Justice Weintraub noted in Jones v. Falcey, 48 N-J- 25, 222 A.2d 101 at 108, one of the valid teachings of Meeks v. Avery, D.C., 251 F.Supp. at 251, is that where “a pattern is disclosed in which the integrity of political subdivisions is ignored, then the districts could be formed with population substantially equal in each of them; and when * * * lines [are] completely disregarded in * * * many instances the court [is] justified in holding that a pattern had developed in which the integrity of political subdivisions was ignored.” That sound principle merely says that a State legislature must practice what it preaches. A State legislature should therefore anticipate that a purported State policy of observing particular political subdivisions lines can not be relied upon to justify variances above those which would have been created by proper application of the “as nearly as is practicable” principle unless the legislature’s enacted plan does in fact observe and apply such an articulated State policy. The evidence in this case establishes that the township lines of St. Louis County were in fact broken by the 1967 Act; precinct lines were in fact used to define the boundaries of particular congressional districts created by the 1967 Act. It is therefore not necessary to apply the principle of Jones v. Falcey, supra, to the factual situation presented in Preisler III because the avoidable variances created by the 1967 Act’s Districts 1, 2, and 3 of +7,765, +4,467, and +4,118, respectively, are far above the variances that would have been created had the 1967 Legislature in fact established and followed the less stringent policy of recognizing townships and ward lines. Change “D” in Appendix A to Preisler II illustrated how the 1965 Legislature changed the allocation of population made in its 1961 Act to District 7. We there noted that the 1965 Missouri Legislature had shifted only Barton County (pop. 11,113) from District 7 to District 4. As noted above, the 1967 shuffle of Kansas City’s wards made it necessary to shift St. Clair and Benton Counties from District 7 to District 4. Defendants’ hearing Exhibit A, the map used by the majority leader of the House, shows that those particular counties, along with other counties involved in other shifts made by the 1967 Act apparently were the subject of some sort of special discussion because those counties were shaded with a lead pencil. The problem created by the shift of St. Clair and Benton Counties required that the 1967 Legislature shift additional population into District 7. Once again the population allocated to District 8 under the 1965 Act had to be “squeezed like a sponge.” Camden (pop. 9,116) and Laclede (pop. 18,991) Counties (both of which were pencil shaded on the House majority leader’s map) were shifted from old District 8 with a resultant total net gain of 10,949 population for District 7. That shift produced a +4,788 variance under the 1967 Act. But this Court must face the undisputable fact that the addition of Hickory County (pop. 4,516) to the 1967 shift from old District 7 to new District 4, would have reduced new District 7’s variance to a +278 and would have, at the same time, reduced the variance of new District 4 from a -12,260 to a -7,754. Hickory County was in fact proposed for inclusion in District 7 in several versions of Senate Bill No. 182 as that bill progressed through the 1967 Missouri Legislature. Indeed, the Senate Substitute for Senate Bill No. 182 made provision for a District 7 that included Hickory County in which the variation from the ideal district was -1,292, a substantially smaller, variation than that eventually created by the 1967 Act. Change “E” noted in Preisler II’s Appendix A reflected the 1965 Act’s shift of Wayne County from the 1961 Act’s District 8 to that Act’s District 10. Following the same pattern established in 1965, the 1967 Legislature shifted the additional and quite sparsely populated counties of Shannon (pop. 7,087), Carter (pop. 3,973), Reynolds (pop. 5,161), Iron (pop. 8,041), and Madison (pop. 9,366) to form the 1967 Act’s District 10. The 1967 Act’s additional shift reduced the greatest minus variation of the 1965 Act only to a minus variation of 8,113. The undisputable facts establish that had Dent County (pop. 10,445) been added to the 1967 shift, the variance of new District 10 would have been reduced to +2,332. That shift would have also had the effect of reducing the variance in adjoining new District 8 to +3,105. We further find that House Committee Substitute for Senate Committee Substitute for Senate Bill No. 182 was proposed to and rejected by the 1967 Missouri General Assembly. That plan proposed the inclusion of Dent County in the 1967 shift (together with every other one of the other five counties that were in fact shifted by the 1967 Act). Acceptance of that proposed and rejected plan would have produced a variance of +1,064 for new District 10. The 1967 Missouri Legislature was required to put the squeeze on old District 8 to provide new District 10 the population assigned it by the 1967 Act. The 1967 Act’s shift of additional rural counties from old District 8 to new District 10 was compensated for by siphoning still more population from the St. Louis metropolitan area. The 1967 Legislature invaded St. Louis County population by extending the boundaries of still a fifth congressional district to include part of that County. The 1967 Act added Concord and Meramac townships and particular precincts in Bon Homme in St. Louis County to new District 8. The addition of that portion of St. Louis County by the 1967 Act caused new District 8 to stretch from the city limits of St. Louis to within less than 75 miles of the city limits of Kansas City. Thus, in total disregard of the duty to draw compact and contiguous districts placed on it by the Missouri Constitution, the 74th General Assembly added the population of urban St. Louis County to an essentially rural district in order to balance out the population of the five rural counties shifted from old District 8 to new District 10 by the 1967 Act. For the applicable Missouri law on gerrymander, see State ex rel. Barrett v. Hitchcock, 1912, 241 Mo. 433, 146 S.W. 40, and Preisler v. Doherty, 1955, 365 Mo. 460, 284 S.W.2d 427. Compare Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960), for circumstances under which gerrymandering can become a federal question. In Preisler II we found as a fact that “the discrimination against the St. Louis metropolitan area, evidenced by the 1961 Act, was continued under the 1965 Act” (257 F.Supp. at 958, fn. 4). We find that the 1967 Act continues the same discrimination. We found in Preisler II that under both the 1961 Act and the 1965 Act “the City of Kansas City was split between three separate Congressional districts,” and that under both acts the “population of the City of Kansas City * * * continued to be utilized to increase the deficient population of two essentially rural districts” (257 F.Supp. at 960). We make the same finding in regard to the use of Kansas City’s Platte and Clay County population in regard to District 6 created by the 1967 Act. In Preisler II we found that the 1961 Act and the 1965 Act “both have the same pattern of overvaluation and overweight of the votes in the rural geographical areas, as contrasted with the votes in the two metropolitan areas of the State” (257 F. Supp. at 960). We make the same finding in regard to the 1967 Act, particularly as it applies to the St. Louis metropolitan area and the area of Kansas City north of the Missouri River. In Preisler II we found and held that “all that the Seventy-third General Assembly did [in its 1965 Act] in regard to the 1961 Act held constitutionally impermissible in Preisler I was to move three small counties from one district to another and to shift a few wards from one district to another in both metropolitan areas” (257 F.Supp. at 980). We find and hold in Preisler III that all the 74th General Assembly did in regard to the 1965 Act held constitutionally void in Preisler II was to add a few more counties to the shifts of rural counties and to invade more grossly than ever before the population concentrated in the two metropolitan areas of this State. In Preisler II we found that the overvaluation of votes in the particular districts created by the 1965 Act had been established by “the undisputed and unexplained facts” presented in that case. We noted in Preisler II that “counsel did not attempt to suggest any reason why it was not practicable to avoid variances * * *” (257 F.Supp. at 969). We expressly rejected the argument presented by counsel for the intervenors that the 1965 Act represented “the best that could be gotten under the practical political circumstances” (257 F.Supp. at 969). We emphasized again and again that the “substantial population deviations and variances” were “unexplained” (257 F.Supp. at 974). We directed attention to what we called the most significant language in Drum I in regard to what we believed to be the burden of proof in apportionment cases. We italicized the language of Judge Sobeloff in Tawes I from page 976 of 257 F.Supp. that anticipated the Supreme Court’s later holding in Kilgarlin v. Hill. On page 977 of 257 F.Supp. of Preisler II, we italicized the precise language from Tawes III that the Supreme Court quoted as the proper burden of proof rule in Swann v. Adams III. And, finally, we expressly rejected in Preisler II defendants’ arguments that attempted to establish considerations of “political stability” and their arguments concerning “political,” “economic,” “historical,” and “traditional” factors as legitimate grounds upon which disparities from population based districts could be justified (257 F.Supp. at 980). The legal principles that underlie those findings are, of course, the law of this case as affirmed by the Supreme Court. APPENDIX B (In Two Parts) I. Discussion of House Bill No. 870 Our specific findings in regard to House Bill No. 870 proposed to and rejected by the 1967 Missouri Legislature can best be understood by preliminary examination of the following table: House Bill No. 870 as Originally Introduced and Rejected by the 1967 Missouri General Assembly H. B. 870 1967 Act District District No. No. 1960 Census Population 1960 Population Variance 1 (D 431.896 - 851 2 (2) 431.896 1 — 851 3 (3) 431.896 1