Opinion
Civ. A. No. 2724.
November 15, 1965.
Walter Chandler, Memphis, Tenn., and Hobart F. Atkins, Knoxville, Tenn., for plaintiffs.
Harris Gilbert, Nashville, Tenn., for intervening plaintiff Ben West.
E.K. Meacham and Eugene N. Collins, Chattanooga, Tenn., for intervening plaintiff City of Chattanooga.
C.R. McClain, Director of Law, Knoxville, Tenn., for intervening plaintiff City of Knoxville.
Cecil Branstetter, Nashville, Tenn., for intervening plaintiff Tennessee State Labor Council, AFL-CIO.
Neill S. Brown, Director of Law, and Gilbert S. Merritt, Jr., Asst. Director of Law, Nashville, Tenn., for Metropolitan Government of Nashville and Davidson County, Tennessee.
George F. McCanless, Atty. Gen., and Milton P. Rice, Asst. Atty. Gen., State of Tennessee, Nashville, Tenn., for defendants.
Edwin F. Hunt, Nashville, Tenn., for intervening defendant Tennessee Farm Bureau Federation.
Before PAUL C. WEICK, Circuit Judge, and MARION S. BOYD and WILLIAM E. MILLER, District Judges.
Over six years ago Charles Baker and others filed the initial complaint in this action challenging the constitutionality of Tennessee's Legislative Apportionment Acts. Nearly two years ago this court stated, "malapportionment in Tennessee is an evil of long standing." D.C., 222 F. Supp. 684, 693 (1963). Plaintiffs allege that the evil still exists, and thus bring the issues once again before the court. Before rendering what, it is hoped will be a final disposition of this action, it may be well to summarize its protracted history.
Shortly after the original complaint was filed, the defendants moved to dismiss the action without convening a three-judge court, for the reason that it did not present a substantial, or justiciable federal question. Those were the days of the "political thicket," and the defendants' position was supported by substantial authority. Nevertheless, the court felt that the issues merited consideration by a three-judge court, and the motion was denied. 175 F. Supp. 649, (M.D.Tenn., 1959).
On December 21, 1959, the three-judge court rendered a per curiam opinion dismissing the action ( 179 F. Supp. 824). The court could find "no way in the present case to escape the compelling authority" of numerous Supreme Court decisions to the effect that "whether from a lack of jurisdiction or from the inappropriateness of the subject matter for judicial consideration," the federal courts would not intervene in cases of this type.
Since the court felt powerless to grant a remedy, it could have said no more. But it went on to state that Tennessee's apportionment was a "clear violation" of both the state and federal constitutions. Furthermore, the court agreed with the plaintiffs that "the evil is a serious one which should be corrected without further delay." (emphasis supplied). Perhaps in this language the plaintiffs found a kernel of hope. At any rate, they appealed to the Supreme Court, and the stage was set for constitutional history.
Not every citizen has heard of Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (Mar. 26, 1962), but it is perhaps not fanciful to suggest that the impact of the Supreme Court's decision will be felt for many years to come. It has certainly been felt by the courts. Since that decision, there have been literally hundreds of published opinions dealing with apportionment. And published opinions do not tell the whole story. In this case alone, the single-spaced docket entries require more than 13 pages, and the file is measured not in pages but in feet.
But we digress from our history. The Supreme Court held that these "political" questions were justiciable, and reversed and remanded. On May 7, 1962, following the remand, this court conducted a pre-trial hearing at which the Attorney General of Tennessee informed the court that the Governor was calling a special session of the General Assembly to consider legislative reapportionment. The Attorney General therefore moved the court to stay further proceedings until the General Assembly had acted. The court reserved a ruling on this motion and set June 11, 1962 as the date for a further hearing.
Subsequent to the pre-trial conference, and prior to the scheduled hearing, the General Assembly in extraordinary session reapportioned both the House and the Senate through separate acts passed on June 6, 1962, and approved by the Governor the next day. The interested parties agreed to an amendment of the pleadings so as to bring into issue the constitutionality of the new apportionment statutes. This issue was presented to the court at the June 11 hearing, on plaintiffs' amended motion for summary judgment.
In a per curiam opinion filed June 22, 1962 ( 206 F. Supp. 341) the three-judge court reserved final judgment on all issues until the 1963 General Assembly, constituted and elected under the 1962 statutes, had an opportunity at its regular 1963 session to act on the matter of legislative apportionment. Although it did not enter a judgment, the court expressed its "views" with respect to both the House and the Senate. The 1962 statute reapportioning the House was based on the controversial two-thirds ratio provision of the Tennessee Constitution, and there were "certain discrepancies and inequities." Nevertheless, the court indicated that the House plan was not per se irrational or unconstitutional. The Senate plan was a different matter. The court stated that the only pattern of apportionment in the state Senate was "one of invidious discrimination." The court further indicated that any plan enacted by the 1963 session would have to base apportionment in at least one of the houses "on numbers of qualified voters without regard to any other factor."
At its regular 1963 session, the General Assembly again reapportioned both the House and the Senate. The plaintiffs duly amended their complaints in order to challenge the constitutionality of the new acts, and to request the court to approve reapportionment plans submitted to the court by the plaintiffs.
In an opinion filed October 10, 1963 ( 222 F. Supp. 684), the three-judge court ruled that the reapportionment of the House, having eliminated the "discrepancies and inequities" pointed out by the court in its June 22, 1962 Opinion, complied with the minimum requirements of the equal protection clause. The court rejected plaintiffs' contention that the Fourteenth Amendment required population-based apportionment for both houses of the General Assembly. The reapportionment of the Senate, however, was found to discriminate against the urban counties by allocating senators to them only for every full ratio, although senators were allocated to many rural districts which had less than a full ratio.
With a slight modification, the court approved the plaintiffs' apportionment plan for the Senate, which allocated three more senators to the urban areas than did the 1963 Act. The court did not enter a final order adopting this plan, but allowed the defendants almost four months to file objections to the plaintiffs' plan, as modified, or to submit a plan of their own.
In an Order filed April 17, 1964, the three-judge court noted (1) that the intervenor defendant had filed an alternative senate plan; (2) that the plaintiffs had filed a supplemental brief renewing their contention that both houses must be apportioned on a population basis; and (3) that the plaintiffs' contention would probably be resolved by the imminent decisions in several apportionment cases then pending before the Supreme Court. Consequently, the court ordered a stay of all further proceedings, with the court retaining full jurisdiction.
In an unpublished Opinion filed June 27, 1964, the three-judge court noted that the recent Supreme Court decisions required population-based apportionment for both houses of a state legislature. Accordingly, the court reexamined the entire record and concluded that: (1) the plan for apportioning the Senate, previously submitted by the plaintiffs and modified by the court in its 1963 opinion, was constitutional; and (2) that the plan for apportioning the House previously submitted by the plaintiffs and rejected by the court as unnecessary, was also constitutional. The court directed an Order to be submitted to that effect. Objections to the plaintiffs' House plan could be filed, together with alternative plans, not later than August 3, 1964. If the House and Senate plans, as subsequently modified by the court, were not approved by the 1965 General Assembly, the plans were to "become effective upon adjournment of the 1965 General Assembly, but in no event later than June 1, 1965." [These plans are referred to in Appendix "A" as the Court Approved Plan.]
The defendants then filed objections to the entry of the order requested by the court in its June 27 Opinion. On August 28, 1964, the court directed that the order be filed approving the plans set forth in the appendix to the June 27 opinion. The court, however, directed that the order be without prejudice to the power of the General Assembly to enact apportionment legislation of its own design, so long as the legislation complied with the recent Supreme Court pronouncements.
On April 2, 1965, the three-judge court entered an Order amending the Order of August 28, 1964 to permit the 1965 General Assembly to consider apportionment legislation in an extraordinary session.
On July 15, 1965, the court entered an Order which noted that the 1965 General Assembly, in extraordinary session, had once again enacted reapportionment legislation, and that some of the parties had questioned the validity of the new legislation. The court therefore ordered the action to be re-opened for consideration, and set September 8, 1965 as the hearing date before the three-judge court.
Ratio of the largest to the smallest district
On August 23, 1965, a pre-trial conference was held pursuant to the direction of the court. In a Pre-Trial Order filed August 24, 1965, the court summarized the positions taken at the pre-trial conference and ordered various parties to prepare maps, charts and other illustrative materials. All parties were directed to file briefs not later than September 2, 1965. On September 8, 1965 a hearing on the merits was duly held before the three-judge court to determine the validity of the 1965 legislation.
Such is the history of this case. The court will now proceed to examine the 1965 Act.
The intention of the 1965 Act, as expressed in its preamble, is to design senate and house districts "as nearly of equal population in the respective houses as is practicable." The basis of apportionment is the total number of persons 21 years of age or older based on the 1960 U.S.C.ensus (this number will hereafter be referred to as the total of qualified voters). The Act intends to form senate districts with an average qualified voter population of 63,420, and house districts with an average qualified voter population of 21,140,
deviating from those figures only as necessary to preserve county lines and to avoid combining small rural and medium-sized counties with large urban counties, discriminating against no county or segment of the population, and limiting deviations from average insofar as possible to a maximum of fifteen percent (15%) plus or minus.
The Act divides all counties entitled to two or more senators or representatives into single-member districts. This is done, as the Act recites, in order to eliminate long and cumbersome ballots, to provide identifiable constituencies, to assure voters of a specific senator or representative, and to minimize the dilution or cancellation of the voting strength of various ethnic, political, economic, or social elements of the population within such counties.
In order to maintain the political integrity of counties, the Act establishes floterial districts for both the Senate and the House which, for the most part, include counties not entitled singly to elect one or more senators or representatives. It is the stated purpose of the Act to equalize in one house any inequalities in the representation of certain areas found to be necessary in the other house.
Section 2 of the Act provides for 33 senators and 99 representatives, and defines the counties, precincts, and wards which make up each of the 33 senate districts.
Section 3 defines the counties, precincts, and wards which make up each of the 99 house districts. Fourteen counties elect one direct representative. Each of these counties is, in effect, a district. Ten of these counties also participate in the election of a floterial representative. All counties which are entitled to elect more than one representative or senator are divided into districts corresponding to the number of representatives or senators to be elected. These "sub"-divisions are set out in sections 4 through 9 of the Act.
Section 10 establishes 38 floterial districts which are each entitled to elect one representative. Since in six instances, the same counties which are used to form one floterial district are also used to form an additional floterial district, it appears that there are, in effect, six floterial districts which elect two representatives.
Section 11 provides that in counties entitled to elect more than one representative, the candidate for election shall not be required to reside in the district from which he seeks to be elected, so long as he is a resident of the county. Since one stated purpose of sub-districting counties was to provide for closer contact between the representative and his constituency, a provision which allows the representative to reside in a portion of the county which could be quite remote from his constituency is, at best, incongruous. Furthermore, no similar exception is provided for senators. Apparently, in those counties which are sub-districted into senate districts, the candidate must reside in the district from which he is to be elected. This difference in treatment of senators and representatives is not explained in the Act, but it has not been challenged by the parties and will not be further considered in this opinion.
Section 12 provides that until the next election, the present senators and representatives shall continue to represent their prior constituencies.
Section 13 declares the provisions of the Act to be severable to the extent that the invalidity of any provision does not affect other provisions "which can be given effect without the invalid provision."
STATISTICAL ANALYSIS
Mathematical analysis alone cannot dictate whether the 1965 Act should stand or fall, but in each of the six cases decided by the Supreme Court on June 22, 1964, and in subsequent cases decided by the district courts (see Appendix "D"), two statistics have been of great importance: first, the ratio between the largest and the smallest districts, and second, the minimum percentage of the population which could be represented by a majority of either house.
Minimum percentage represented by a majority
In Tennessee under the 1965 Act, a ratio of 1.39-1 was found to exist when the average of the five smallest House districts was compared with the average of the five largest House districts.
Reference to Appendix "B" indicates the following lowest ratios and highest percentages which the Supreme Court and the district courts have held to be unconstitutional:
HOUSE SENATE
Supreme Court: Ratio 4.36-1 Ratio 2.60-1 Percentage 40.5% Percentage 47%
District Court: Ratio 1.60-1 Ratio 1.60-1 Percentage 42.34% Percentage 48.2% These figures must be analyzed very carefully. For example, although the Supreme Court has struck down an apportionment plan which involved the rather high minimum senate percentage of 47%, there was in the same plan a 5-1 ratio between the largest and the smallest senate districts. Similarly, while the lowest senate ratio which the Supreme Court has yet struck down was only 2.60-1, there is, nevertheless, clearcut language elsewhere which suggests that even a 2-1 ratio is unconstitutional. See Reynolds v. Sims, 377 U.S. 533, 562-563, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964); Davis v. Cameron, 238 F. Supp. 462 (D.C.Iowa, Feb. 11, 1965) (McManus, D.J., dissenting).
1965 SENATE
As we have seen, there are 33 senators. A majority of the Senate represents 1,011,021 qualified voters, which, compared with the total of 2,092,891 qualified voters, results in a minimum percentage of 48.38%. The ratio between the largest and the smallest senate districts, 78,922 and 56,996, is 1.38-to-1. These figures, taken together, compare favorably with present constitutional limits established by case law. The ratio is well below the minimum that has been struck down by either a district court or the Supreme Court — 1.38-to-1 in the instant case compared with 1.60-to-1 and 2.60-to-1. Similarly, the minimum percentage represented by a senate majority is within currently judicially approved limits — 48.38% in the instant case compared with a district court's disapproval of 48.2% and the Supreme Court's disapproval of 47%. (See Appendix "B"). This brief analysis suggests that the Senate apportionment, in mathematical terms, meets the requirements of the Supreme Court's recent "one person, one vote" interpretations of the Fourteenth Amendment. E.g., Reynolds v. Sims, supra.
1965 HOUSE
The minimum percentage of qualified voters represented by a house majority is 46.38%. This percentage is well within house determinations, 42.34% being the highest percentage yet struck down by a district court, and 40.5% being the highest percentage yet struck down by the Supreme Court. The Supreme Court has expressly reserved judgment on the constitutionality of the Colorado House, which had a ratio of 1.7-to-1 and a minimum percentage of 45.1%, but stated that these figures were "arguably" constitutional. Lucas v. Colorado General Assembly, 377 U.S. 713, 730, 734, 84 S.Ct. 1459, 12 L.Ed.2d 632 (1964). The Tennessee House has a lower ratio (1.51-to-1), and higher percentage (46.38%). This ratio was computed by the original defendants based on figures reflected in their Exhibit B. The 1.51-1 figure is, in a sense, a distortion arising from the General Assembly's attempt to compensate for the over and under-representation of two counties in the General Assembly. For instance, by virtue of population Shelby County was entitled to 17.01 representatives and 5.67 senators. Under the 1965 Act, Shelby County is allotted 6 senators and 16 representatives in order that its representation, when considered as a whole, may be equalized. The compensation, however, does create abnormally large house districts. Similarly, on a population basis, Knox County is entitled to 2.40 senators and 7.19 representatives. Under the Act it receives 2 senators and 8 representatives. This equalization of Knox County's representation in the General Assembly is rational and fair, but it tends to create abnormally low house districts in that county. Therefore, since the most divergent population districts resulted from the attempt to compensate in the House for the over-representation of Shelby County and the under-representation of Knox County in the Senate, it would seem that a more meaningful ratio would be obtained by comparing the average of the five smallest house districts, three of which are in Knox County, with the average of the five largest house districts, all of which are in Shelby County. The respective averages, 17,862 and 24,900, result in a ratio of 1.39-1, a ratio which we find to be constitutional.
See also Davis v. Cameron, 238 F. Supp. 462 (D.C.Iowa, Feb. 11, 1965) which suggested without holding, that a House with a 2.23-to-1 ratio in which 44-48% of the population could elect a majority, was "arguably close." One judge dissented on this point. In League of Nebraska Municipalities v. Marsh, 242 F. Supp. 357 (D.C.Neb., May 12, 1965), however, a district court struck down a ratio which was lower than the House ratio in Lucas. The district court distinguished Lucas on the ground that a 1.7-to-1 ratio was "arguably" constitutional only if no better plan could be devised. Since plans had been submitted to the court which reduced a 1.60-to-1 ratio to 1.26-to-1, the district court struck down the 1.60-to-1 ratio. It should be noted that plaintiffs in the present case have previously submitted plans, which the court has tentatively approved, which would reduce the House ratio from 1.51-to-1, to 1.16-to-1, and which would reduce the Senate ratio from 1.38-to-1, to 1.18-to-1. Thus, if the court were to adopt the test used in Nebraska, without regard to other considerations, it could be argued that the 1965 Act should not be allowed to stand.
Consequently, in terms of a strict mathematical analysis, we find that the 1965 Senate and House apportionments are constitutional.
BASE
We turn now to non-mathematical considerations. The apportionment base for the 1965 Act is qualified voter population. Since several court decisions indicate that the initial basis should always be total population, the court has also analyzed the 1965 Act as if it were based on total population, with particular attention being directed to the five major urban counties. It was found that the qualified voter population per county bore a uniform and reasonable relation to the total population. See Appendix "C." Whichever base is used, 50% of the total base should be represented by 50% of each House. In Tennessee, 1,752,954 persons out of a total population of 3,567,089 are represented by a majority of the Senate. This means that a majority of the Senate represents 49.1% of the total population. Senators from the same districts represent 48.38% of the qualified voters. A similar comparison for the House shows that 51.50% of the House represents 48.5% of the total population, and that these same representatives represent 46.38% of the qualified voter population. It was further found that the five major urban counties (Davidson, Shelby, Hamilton, Sullivan and Knox) elect fifteen senators, or 45.45% of the Senate. The combined total population of these counties is 45.67% of the total population (1,629,329 out of 3,567,089). Thus, the five urban counties elect .22% fewer senators than they should elect if apportionment were based on total population. The same counties make up 46.06% of the total qualified voters (963,770 out of 2,092,891). Therefore, on the present qualified voter basis these counties elect .60% fewer senators then they should elect. These very minor discrepancies indicate that there is no invidious discrimination in the use of qualified voters rather than total population as the apportionment base. Moreover, it should be noted that if apportionment were based on total population, these five counties would receive fifteen senators and forty-five representatives — the same number they receive under the 1965 Act on a qualified voter basis.
E.g., Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964): "Population is, of necessity, the starting point for consideration and the controlling criterion for judgment in legislative apportionment controversies." ( 377 U.S. p. 567, 84 S.Ct. p. 1384). "Both houses of a state legislature must be apportioned on a population basis." (p. 577, 84 S.Ct. p. 1389). However, it should be noted that even in Reynolds, there is language which suggests that apportionment may be based on something other than total population. For example: "We realize that it is a practical impossibility to arrange legislative districts so that each one has an identical number of residents, or citizens, or voters. Mathematical exactness or precision is hardly a workable constitutional requirement." (p. 577, 84 S.Ct. p. 1390, emphasis supplied). Several district courts have approved apportionments based on something other than total population. For example, in Holt v. Richardson, 238 F. Supp. 468 (D.C.Hawaii, Feb. 17, 1965) the court approved a house apportionment based on registered voters. Of course, registered voters would be a more extreme deviation from total population than would qualified voters. The district court did find, however, that the number of registered voters represented a "reliable and fairly uniform relationship to the total population." In Stout v. Hendricks, 228 F. Supp. 568 (D.C.Ind., 1964), the district court expressed some concern, but approved an apportionment based on the total number of males 21 or older. The court did not attack the base as such, but declared the apportionment to be unconstitutional because of inequalities resulting from non-adherence to the base. In Buckley v. Hoff, 243 F. Supp. 873 (D.C.Vermont, June 28, 1965), a district court approved a plan by which the Senate was apportioned on the basis of total residents, and the House was apportioned on the basis of registered voters. The court noted that this plan did not involve changing the historical basis for apportionment in Vermont, did not perpetuate existing discrimination, and was justified in Vermont, although it might well be unconstitutional in other states. In WMCA, Inc., v. Lomenzo, 238 F. Supp. 916 (D.C.N.Y., Feb. 1, 1965) a district court stated that apportionment based on citizen population (total population minus aliens) or on resident population, would satisfy the requirements of Reynolds. The court, however, struck down an apportionment plan based on voter population, for the reason that the state had, for seventy years, used citizen population, and the change was intended to be a discrimination against the urban districts. Since voter population was determined by the 1962 vote for Governor, it would be a much narrower base than qualified voters. Furthermore, it appears that Tennessee's Constitution specifies qualified voters as the apportionment base, and that this has been the base for many years.
MULTI-DISTRICT COUNTIES
In the present case, the original plaintiffs contend that "sub"-districting all counties entitled to elect more than one senator violates the State Constitution. They have also suggested that sub-districting violates the Federal Constitution. We turn now to the federal question.
In Fortson v. Dorsey, 379 U.S. 433, 85 S.Ct. 498, 13 L.Ed.2d 401 (Jan. 18, 1965), the plaintiffs contended that Georgia's failure to sub-district all counties entitled to elect more than one senator was unconstitutional. The district court agreed, but the Supreme Court disagreed and reversed the decision below (D.C., 228 F. Supp. 259). Also, in Davis v. Cameron, 238 F. Supp. 462 (D.C.Iowa, Feb. 11, 1965), a district court, with some reluctance, approved the use of multi-member districts. One judge dissented, in part, because he felt that, in the absence of discrimination, whether to sub-district or to hold county-wide elections was a political matter to be determined solely by the legislature. It appears, then, that county-wide voting may be constitutional. But the Fortson decision does not say that county-wide voting will always be constitutional, and it certainly does not say that sub-districted voting will be unconstitutional. Since the district court had granted the plaintiffs' motion on summary judgment, the Supreme Court treated the district court's ruling as a decision that the statute providing for county-wide voting was unconstitutional on its face. The Supreme Court was careful in Fortson to point out that although the statute was not, on its face, unconstitutional, it might be so as applied:
It might well be that, designedly or otherwise, a multi-member constituency apportionment scheme, under the circumstances of a particular case, would operate to minimize or cancel out the voting strength of racial or political elements of the voting population. When this is demonstrated it will be time enough to consider whether the system still passes constitutional muster. ( 379 U.S. p. 439, 85 S.Ct. p. 501, emphasis supplied).
The Court relied on Reynolds for the proposition that equal protection does not "necessarily" require single-member districts. The court noted that Fulton County Georgia had a population nearly seven times larger than a single-member constituency, and that the county elected seven senators. Thus, there was "clearly no mathematical disparity." (p. 437).
Fortson suggests two things: First, at least with respect to the Federal Constitution, the states will be left considerable leeway in their choice of single-member or county-wide voting districts; and second, the states should consider the possibility of discrimination against racial or political elements inherent in county-wide voting.
Furthermore, in Lucas v. Colorado General Assembly, 377 U.S. 713, 84 S.Ct. 1459, 12 L.Ed.2d 632 (1964), Chief Justice Warren, speaking for the Court, stated that county-wide voting in multi-member counties is "extremely objectionable," (p. 726-727, n. 13, 84 S.Ct. 1459), and set forth several reasons for sub-districting such counties (p. 731, 84 S.Ct. 1459). It should be noted that these reasons are precisely the reasons set forth in the preamble to the 1965 Act. Apparently, the sub-districting provisions of the Act are an attempt to conform to the Lucas suggestion. Were the Act simply an attempt to conform to the Supreme Court's expressed preference for sub-districting, without more, we would be justified, absent a showing of discrimination, in sustaining the validity of the Act. And where, as here, the state has made an express legislative finding that sub-districting will eliminate long and cumbersome ballots, provide identifiable constituencies, assure voters of a specific senator or representative, and minimize the dilution or cancellation of the voting strength of various ethnic, political, economic, or social elements of the population within such counties, it would be an improper exercise of our judicial function to hold an Act embodying such an express finding to be violative of the principle of "one person, one vote" enunciated by the Supreme Court. The court, therefore, holds that sub-districting counties does not violate the Federal Constitution.
As mentioned above, the plaintiffs primarily attack sub-districting as a violation of the State Constitution. The alleged Constitutional prohibition occurs in Article II, Section 6, which refers only to the apportionment of the Senate, and which reads, in part, as follows:
There is no language in the Tennessee Constitution which can possibly be construed as prohibiting sub-districting in the apportionment of the House; indeed, the plaintiffs do not seriously challenge house sub-districting.
When a district is composed of two or more counties, they shall be adjoining; and no county shall be divided in forming a district.
There are several possible interpretations of this sentence. One interpretation, urged by the intervening defendant, is based on the fact that the original Constitution contained a comma rather than a semi-colon before the last clause, and that the transition from a comma to a semicolon may be nothing more than a scrivener's mistake. This interpretation would uphold sub-districting even under the State Constitution.
The Tennessee Supreme Court has not yet construed this portion of the State Constitution. Absent a determination by a state court, a federal court may construe and apply a state constitutional provision. Nevertheless, we decline to determine whether sub-districting is permitted or prohibited by the state constitution, although the reason for abstention may be somewhat more compelling when the issue is pending in a state court. See WMCA, Inc. v. Lomenzo, 238 F. Supp. 916, 921-922 (S.D.N.Y., 1965). The fact that such a suit is not now before the Tennessee courts, however, does not mean that this court, in the exercise of its discretion, may not permit the state courts an opportunity to determine the proper construction of Article II, Section 6. Abstention is particularly justified where, as here, there are difficult questions of state constitutional interpretation involved, and where the resulting interpretation would significantly affect the political structure of the state — a structure which, absent discrimination, the state has the undoubted right to determine. Our ruling either for or against the plaintiffs' interpretation would not be binding on the state court. And, if we considered, and ruled in favor of plaintiffs' interpretation of the state constitution, the validity of the Act could still be sustained by applying the severability clause. Our refusal to consider this difficult state question could hardly be considered an abdication of judicial responsibility.
We use the term abstention in its broadest sense. Technically, abstention presupposes that the court has jurisdiction, which, for various reasons, it declines to exercise. It should be noted that if the alleged violation of the state constitution were the sole basis of the plaintiffs' case, this court would not even have jurisdiction. Since it appears, after hearing, that the plaintiffs offer no serious challenge to the federal constitutionality of the 1965 Act, the court has considered the possibility of dismissing the original plaintiffs' motion for summary judgment. challenging the 1965 Act, for want of jurisdiction for failure to present a substantial federal question. This course has been rejected for the reason that the federal questions, although not strongly supported, are nonetheless present, and the issue of valid apportionment in Tennessee which has long been before this court should be settled with finality — at least insofar as the federal constitution is concerned. Still, since dismissal on this ground would be within the court's discretion, our abstention as to the unsettled issue of state constitutional interpretation should not be subject to challenge.
The clause, if applied, would invalidate the senate sub-districting provisions, with the result that the senators from multi-senator counties would be elected at large.
SUB-DISTRICT BASES
Representatives and senators are, in every case, allocated to counties on the basis of the 1960 Federal Census enumeration of qualified voters per county. The Census, however, does not contain adequate information to enable the counties which are entitled to more than one senator or representative to be sub-districted on the basis of actual qualified voter populations. The task of sub-districting multi-member counties was left up to the county delegations. Basically, two methods were used. In Davidson County, for example, the county was sub-districted on the basis of total population. In Knox County, the county was sub-districted on the basis of a "calculated" qualified voter figure. This calculated figure was arrived at by determining two ratios: first, the ratio of registered voters to qualified voters in Knoxville; and second, the ratio of registered voters to qualified voters in the remainder of Knox County. Two ratios were used in order more accurately to reflect the fact that the ratio of registered voters to qualified voters was somewhat higher in Knoxville than in the remainder of the county. These ratios, or "multipliers," were then applied to the registered voter figures for the various precincts in order to determine the calculated qualified voter population per precinct. On the basis of these calculated figures, the county was then sub-districted. The original plaintiffs challenge the use of a total population base, or a "multiplier" base as a violation of the Fourteenth Amendment. The plaintiffs apparently contend that the difference between using a qualified voter base for the county-wide allocation, and using total population or a calculated qualified voter base for sub-districting is per se unequal, and, therefore, unconstitutional. Plaintiffs, however, have not alleged a single, specific instance of discrimination or inequality, even though requested by the court to do so. It may be that specific inequalities resulting from the use of different bases could be proved, but we cannot agree that the use of different bases is per se unconstitutional.
OMITTED VOTERS IN DAVIDSON COUNTY
Another problem arises from the fact that the Act failed to include in any Davidson County representative district the 5th Precinct of the 32nd Councilmanic District, composed of approximately 800 qualified voters. The omission apparently occurred by a combination of two events. Prior to the date of the Act (May 27, 1965), the Metropolitan County Council divided the 2nd Precinct of the 32nd Councilmanic District into two precincts — the 2nd Precinct and the 5th Precinct. Thereafter, the Act fixed and enumerated precincts in Davidson County as they were delineated on May 10, 1965, a delineation which, of course, should have taken into account the newly created 5th Precinct. Although the General Assembly included the 1st, 2nd, and 3rd Precincts in the 8th Representative District and the 4th Precinct in the 10th Representative District, it made no provision for the 5th Precinct.
The omission was obviously the result of legislative inadvertence and should not result in undermining the whole apportionment scheme. The problem is the appropriate remedy. Several alternatives have been considered, including supplying words by statutory construction, application of the elision clause, and declaring the whole Act unconstitutional coupled with enforcement of a court plan identical to the Act itself, except for the addition of the omitted Davidson Precinct to one of the eleven districts in that county.
The first named solution appears to us to be compatible with settled principles of statutory construction as reflected in current Tennessee case law. Metropolitan Government of Nashville and Davidson County v. Poe, Tenn., 383 S.W.2d 265 (1964); Hilliard v. Park, 212 Tenn. 588, 370 S.W.2d 829 (1964); Hudgins v. Nashville Bridge Co., 172 Tenn. 580, 113 S.W.2d 738 (1937); Southern Ry. Co. v. Rowland, 152 Tenn. 243, 276 S.W. 638 (1925).
We think the legislative intent was that the omitted precinct was to be a part of the 8th Representative District. While legislative history on this point is somewhat meager, it appears that the local Davidson County delegation in the House, to whom responsibility for devising county district lines had been committed, intended to include what is now the 5th Precinct in the 8th Representative District. That this may fairly be assumed to have been the legislative purpose also appears from two further facts: first, the omitted precinct is geographically contiguous to the 8th Representative District; and second, its inclusion in the 8th Representative District would, from a population standpoint, more nearly achieve a "one person, one vote" apportionment than would its inclusion within the 10th Representative District, the only other representative district to which it is contiguous.
For these reasons we find that the General Assembly intended to incorporate the 5th Precinct of the 32nd Councilmanic District as a part of the 8th Representative District of Davidson County, and the 1965 Apportionment Act is so read and construed. On this basis the inadvertent omission gives rise to no constitutional problem.
DIRECT-MEMBER FLOTERIAL DISTRICTS
As presently structured, it appears that there may be some discrimination against counties within the seven direct-member floterial districts, that is, districts which elect a floterial representative even though one or more counties within the district also elect a direct representative. However, although there is possible discrimination against counties, there is no discrimination between districts, and it is with the districts that we are primarily concerned. Although this discrimination could be eliminated through the use of at-large elections, still, in a practical sense, the same counties which control the election of the direct and floterial representatives under the present Act would still control the election of the increased number of floterial representatives in at-large elections. Furthermore, an at-large election would create two problems which may be more troublesome than the problems presented by the Act. First an at-large election would mean that ten counties within these districts, whose populations entitle them to a direct representative, would be, unlike all the other full-ratio counties of the state, forced to participate with other counties in the election of "their" representative. Second, it would be possible in six of these seven districts, under an at-large election, for one county to control the election of all the floterial representatives. Finally, it should be noted that the use of direct-member floterial districts is not questioned, and that in open court plaintiffs expressly disclaimed any challenge to the Act upon this ground.
Under these circumstances, although the use of direct-member floterial districts possibly leads to some inequality, the court is not prepared to find that such use is unconstitutional when considered in the light of the whole plan envisaged by the 1965 Act and its stated purpose to preserve the integrity of county lines.
The judgment to be entered pursuant to this opinion will deny the original plaintiffs' motion for summary judgment, will sustain the original defendants' motion for summary judgment, and will declare the 1965 Apportionment Act to be a federally constitutional apportionment of Tennessee's General Assembly. This being a final disposition of this protracted proceeding, it will be placed on the retired docket.
APPENDIX "A" COMPARISON OF THE REPRESENTATION ACCORDED THE FIVE URBAN COUNTIES UNDER COURT APPROVED PLAN AND 1965 ACT
Under County Entitled: Got: Result: Entitled: Got: Result: Result:
HOUSE SENATE COMBINED Ct. Plan Knox 7.19 7 .19 UR 2.92 3 .08 OR .09 UR Act 7.19 8 .81 OR 2.40 2 .40 UR .41 OR Ct. Plan Shelby 17.01 17 .01 UR 5.67 6 .33 OR .32 OR Act 17.01 16 1.01 UR 5.67 6 .33 OR .68 UR Ct. Plan Davidson 11.49 11 .49 UR 3.83 4 .17 OR .32 UR Act 11.49 11 .49 UR 3.83 4 .17 OR .32 UR Ct. Plan Sullivan 3.18 3 .18 UR 1.06 1 .06 UR .24 UR Act 3.18 3 .18 UR 1.06 1 .06 UR .24 UR Ct. Plan Hamilton 6.76 7 .24 OR 2.25 2 .25 UR .01 UR Act 6.76 7 .24 OR 2.25 2 .25 UR .01 UR UR stands for under-represented; OR stands for over-represented. The "Entitled" figure for Knox County Senators is slightly different under the Court Approved Plan and the 1965 Act. This is because the Court Plan combined Knox and Anderson Counties in the formation of senate districts. The total of the Combined Results shows that under the Court Approved Plan the five urban counties are .34 under-represented; under the 1965 Act, the same five counties are .84 under-represented. Of course, the above analysis weighed senators and representatives as if they were equal. Since the denial of one-third of a senator is more important than the denial of one-third of a representative some compensating factor should be applied in computing the Combined Result figure. However, since the Supreme Court has apparently not discussed this troublesome issue, no attempt has been made to determine, much less apply, the correct compensating factor. fn__ fn__ fn__ fn_ fn_ fn___ fn_ fn_ fn_ fn_ fn_ fn_ STATE LEGISLATIVE ACTS HELD TO BE UNCONSTITUTIONAL STATE HOUSE SENATE COURT Ratio Percentage Ratio Percentage ALABAMA 16-1 25.7% 41-1 25.1% Supreme Court ARKANSAS 7.5-1 — — — 2.22-1 44.0% District Court COLORADO 1.7-1 45.1% 3.6-1 33.2% Supreme Court CONNECTICUT 424-1 11.9% 8-1 31.9% District Court (later aff'd) DELAWARE 12-1 28% 15-1 21% Supreme Court GEORGIA 2.00-1 42.34% 1.81-1 48.2% District Court INDIANA 5.8-1 35.3% 3.9-1 37.4% District Court IOWA 2.23-1 44-48% 3.2-1 38.9% District Court MARYLAND 6.00-1 35.6% 32-1 14.1% Supreme Court MINNESOTA 7-1 35% 4.1-1 39.1% District Court MISSOURI 12.7-1 — — — 1.66-1 — — — District Court NEBRASKA 1.60-1 — — — 1.60-1 — — — District Court NEW YORK 12.7-1 37.5% 2.6-1 38.1% Supreme Court OKLAHOMA 5-1 31% 4.73-1 — — — District Court (later aff'd) PENNSYLVANIA 3.7-1 — — — 1.96-1 — — — District Court TENNESSEE1.51-1 46.38% 1.38-1 48.38% 1965 Act UTAH 18.2-1 37.8% 5.4-1 25.3% District Court VERMONT — — — 12% 5-1 47% District Court (later aff'd) VIRGINIA 4.36-1 40.5% 2.65-1 41.1% Supreme Court WASHINGTON 4.65-1 38% 7.25-1 35.6% District Court (later aff'd)APPENDIX "C" 1965 Act — Senate
Approximate Total Qualified Percent District Population Voters Deviation -------- ---------- --------- ------------ 1 117,175 67,121 59.3 2 114,139 69,478 58.8 3 111,103 64,835 58.4 4 102,413 59,042 57.7 5 104,325 58,270 55.9 (6 (7 250,523 151,999 60.7 8 105,533 107,260 57.6 9 107,462 117,932 56.9 (10 (11 237,905 142,979 60.1 12 110,370 59,569 53.9 13 98,452 57,317 58.2 14 115,696 65,614 56.6 15 93,055 56,996 61.2 16 60,914 66,353 59.6 17 61,191 69,067 59.9 18 111,338 62,595fn_ 61.0 19 115,286 57,557fn_ 61.0 20 102,615 59,570fn_ 61.0 21 94,356 62,417fn_ 61.0 22 97,656 60,094 57.2 23 102,323 61,919 59.7 24 105,053 67,423 63.3 25 103,646 63,824 60.5 26 106,507 60,969 56.8 27 105,354 65,452 55.5 (28 (29 (30 (31 (32 (33 627,019 359,532 57.4
1965 Act — House
Anderson 60,032 33,554 55.9 Blount 57,525 32,849 57.0 Bradley 38,324 21,881 57.1 Carter 41,578 23,907 55.1 Gibson 44,699 27,791 62.2 Greene 42,163 25,248 59.9 Hamblen 33,092 19,465 58.8 Madison 60,655 36,033 59.4 Maury 41,699 25,033 60.0 McMinn 33,662 19,696 59.6 Montgomery 55,645 30,419 54.7 Roane 39,133 21,957 56.1 Rutherford 52,368 30,347 57.9 Sumner 36,217 21,776 60.1 Washington #1 2 64,832 39,287 60.6 Sullivan #1 2 3 114,139 67,121 58.8 Hamilton #1 2 3 4 5 6 7 237,905 142,979 60.1 Knox #1 2 3 4 5 6 7 8 250,523 151,999 60.7 Davidson #1 36,742 22,413fn_ 61.0 2 35,752 21,809fn_ 61.0 3 38,689 23,600fn_ 61.0 4 37,850 23,089fn_ 61.0 5 35,053 21,382fn_ 61.0 6 32,994 20,126fn_ 61.0 7 34,963 21,327fn_ 61.0 8 37,908 23,124fn_ 61.0 9 35,201 21,473fn_ 61.0 10 38,292 23,358fn_ 61.0 11 33,506 20,439fn_ 61.0 Shelby #1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 627,019 359,532 57.3 Approximate Floterial Total Qualified Percent District Population Voters Deviation ------------------------------------------------------------
1 109,588 64,183 58.6 2 38,225 21,997 57.5 (3 (4 69,134 39,807 57.6 5 40,071 22,627 56.5 6 81,282 46,903 57.7 7 35,476 19,614 55.3 8 99,165 55,511 55.9 9 43,349 23,209 53.5 10 40,158 22,026 54.8 11 42,253 23,171 54.8 12 38,433 20,668 53.8 13 38,679 23,189 59.9 (14 (15 65,643 37,119 56.5 16 35,976 21,750 60.5 17 38,469 22,648 58.9 18 71,679 42,242 58.9 (19 (20 67,292 41,286 61.3 21 32,582 20,039 61.5 22 36,763 21,986 59.8 23 66,966 39,568 59.1 (24 (25 62,367 36,544 58.6 26 36,970 22,135 59.9 27 71,950 40,257 55.9 28 30,126 19,193 63.7 29 34,138 21,731 63.7 30 39,281 23,912 60.9 31 35,482 21,178 59.7 32 46,094 24,377 52.8 33 37,987 20,311 53.5 34 105,354 63,824 60.6 (35 (36 60,756 38,076 62.6 (37 (38 45,141 79,945 56.5
APPENDIX "D" Compilation by state of apportionment cases considered by the Court -------------------------------------------
ALABAMA: Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964).ARKANSAS: Yancey v. Faubus, 238 F. Supp. 290 (E.D.Ark., Jan. 28, 1965).
COLORADO: Lucas v. Colorado General Assembly, 377 U.S. 713, 84 S.Ct. 1459, 12 L.Ed.2d 632 (1964); on remand, 232 F. Supp. 797 (D.C.Colo., 1964).
CONNECTICUT: Butterworth v. Dempsey, 229 F. Supp. 754 (D.C. Conn., 1964), aff'd sub nom. Pinney v. Butterworth, 378 U.S. 564, 84 S.Ct. 1918, 12 L.Ed.2d 1037 (1964); on remand, 237 F. Supp. 302 (D.C. Conn., Jan. 22, 1965).
DELAWARE: Roman v. Sincock, 377 U.S. 695, 84 S.Ct. 1449, 12 L.Ed.2d 620 (1964); on remand, 233 F. Supp. 615 (D.C.Del., 1964).
GEORGIA: Dorsey v. Fortson, 228 F. Supp. 259 (N.D.Ga., 1964), rev'd, 379 U.S. 433, 85 S.Ct. 498, 13 L.Ed.2d 401 (Jan. 18, 1965). See also Fortson v. Toombs, 379 U.S. 621, 85 S.Ct. 598, 13 L.Ed.2d 527 (Jan. 18, 1965), and 241 F. Supp. 65 (N.D.Ga., April 1, 1965).
HAWAII: Holt v. Richardson, 238 F. Supp. 468 (D.C.Hawaii, Feb. 17, 1965); 240 F. Supp. 724 (D.C.Hawaii, April 23, 1965), on appeal, Burns v. Richardson, 86 S.Ct. 74.
ILLINOIS: Germano v. Kerner, 220 F. Supp. 230 (N.D.Ill., 1963), rev'd, 378 U.S. 560, 84 S.Ct. 1908, 12 L.Ed.2d 1034 (1964). See also, Scott v. Germano, 381 U.S. 407, 85 S.Ct. 407, 14 L.Ed.2d 477 (June 1, 1965).
INDIANA: Stout v. Hendricks, 228 F. Supp. 568 (S.D.Ind., Nov. 8, 1963, as corrected April 20, 1964).
IOWA: Davis v. Cameron, 238 F. Supp. 462 (S.D.Iowa, Feb. 11, 1965).
MARYLAND: Maryland Committee v. Tawes, 377 U.S. 656, 84 S.Ct. 1429, 12 L.Ed.2d 595 (1964).
MINNESOTA: Honsey v. Donovan, 236 F. Supp. 8 (D.C.Minn., 1964).
MISSOURI: Jonas v. Hearnes, 236 F. Supp. 699 (W.D.Mo., 1964).
NEBRASKA: League of Nebraska Municipalities v. Marsh, 242 F. Supp. 357 (D.C.Neb., May 12, 1965). See also, 232 F. Supp. 411 (D.C.Neb., 1964), on appeal, docketed July 30, 1965, Docket No. 411.
NEW HAMPSHIRE: Levitt v. Stark, 233 F. Supp. 566 (D.C.N.H., 1964).
NEW YORK: WMCA, Inc., v. Lomenzo, 377 U.S. 633, 84 S.Ct. 1418, 12 L.Ed.2d 568 (1964). See also, 238 F. Supp. 916 (S.D.N.Y., Feb. 1, 1965), motion to accelerate appeal or to stay district court order denied, sub nom. Travia v. Lomenzo, 381 U.S. 431, 85 S.Ct. 1582, 14 L.Ed.2d 480 (June 1, 1965).
NORTH DAKOTA: Paulson v. Meier, 232 F. Supp. 183 (D.C.N.D., 1964).
OKLAHOMA: Moss v. Burkhart, 220 F. Supp. 149 (W.D.Okla., 1963), aff'd sub nom. Williams v. Moss, 378 U.S. 558, 84 S.Ct. 1907, 12 L.Ed.2d 1026 (1964). See also, Reynolds v. State Election Board, 233 F. Supp. 323 (W.D.Okla., Aug. 7 and 19, 1964).
PENNSYLVANIA: Drew v. Scranton, 229 F. Supp. 310 (M.D.Pa., 1964).
UTAH: Petuskey v. Clyde, 234 F. Supp. 960 (D.C.Utah, 1964).
VERMONT: Buckley v. Hoff, 234 F. Supp. 191 (D.C.Vt., 1964), aff'd sub nom. Parsons v. Buckley, 379 U.S. 359, 85 S.Ct. 503, 13 L.Ed.2d 352 (Jan. 12, 1965). See also, Buckley v. Hoff, 243 F. Supp. 873 (D.C.Vt., June 28, 1965).
VIRGINIA: Davis v. Mann, 377 U.S. 678, 84 S.Ct. 1453, 12 L.Ed.2d 609 (1964); on remand, 238 F. Supp. 458 (E.D.Va., 1964).
WASHINGTON: Thigpen v. Meyers, 211 F. Supp. 826 (W.D.Wash., 1962), aff'd, 378 U.S. 554, 84 S.Ct. 1905, 12 L.Ed.2d 1024 (1964); on remand, 231 F. Supp. 938 (W.D.Wash., 1964).
I concur in the majority opinion in this case upholding the 1965 Apportionment Act of the State of Tennessee except that I deem unconstitutional that portion of the Act which divides all counties entitled to two or more senators or representatives into single member districts. Instead, I would require that the senators and representatives from such populous counties be elected on an "at large" basis, thus creating multi-member districts which are concurrent with existing county lines. Such a plan is similar to the plan heretofore approved in this case (referred to in the majority opinion as "The Court Approved Plan"). Moreover, the creation of multi-member districts quite recently has been validated expressly under the Fourteenth Amendment of the United States Constitution. Fortson v. Dorsey, 379 U.S. 433, 85 S.Ct. 498, 13 L.Ed.2d 401 (1965); Davis v. Cameron, 238 F. Supp. 462 (D. Iowa 1965).
1965 Public Chapter No. 3, Extraordinary Session (84th General Assembly), Senate Bill No. 10, signed by the Governor on May 28, 1965.
(I note in this regard that said portion of the Act is severable in that its invalidity does not affect the other provisions. See Section 13 of the Act.)
In my judgment, the creation of multi-district counties by the 1965 Apportionment Act only for a few of the more populous counties of the State, thus destroying their political integrity while maintaining the political integrity of all the other counties of the State, is unwarranted and unreasonable and invidiously discriminates against the citizens of the more populous counties. It does not comport, therefore, with the requirements of the equal protection clause of the Fourteenth Amendment. As held by the United States Supreme Court in Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964), the preservation of traditional and historic political subdivisions is a legitimate end to be sought in the proper apportionment of senators and representatives in a State Legislature. Not only do such political subdivisions have historic significance to the people of this State, but more important they are charged with grave and direct responsibilities in the carrying out of State policies and programs. Moreover:
"Indiscriminate districting, without any regard for political subdivision * * * may be little more than an open invitation to partisan gerrymandering." ( 377 U.S. at 578, 579, 84 S.Ct. at 1390)
Such, in my opinion, has been the effect of the carving up of the more populous counties of the State of Tennessee by the 1965 Apportionment Act.
Therefore, respectfully I dissent to this limited extent.