concurring.
One emerging from the rash of opinions with their accompanying clashing of views may well find himself suffering a mental blindness. The Court holds that the appellants have alleged a cause of action. However, it refuses to award relief here — although the facts are undisputed — and fails to give the District Court any guidance whatever. One dissenting opinion, bursting with words that go through so much and conclude with so little, contemns the majority action as “a massive repudiation of the experience of our whole past.” Another describes the complaint as merely asserting conclusory allegations that Tennessee’s apportionment is “incorrect,” “arbitrary,” “obsolete,” and “unconstitutional.” I believe it can be shown that this case is distinguishable from earlier cases dealing with the distribution of political power by a State, that a patent violation of the Equal-Protection Clause of the United States Constitution has been shown, and that an appropriate remedy may be formulated.
I.
I take the law of the case from MacDougall v. Green, 335 U. S. 281 (1948), which involved an attack under the Equal Protection Clause upon an Illinois election statute. The Court decided that case on its merits without hindrance from the “political question” doctrine. Although the statute under attack was upheld, it is clear *252that the Court based its decision upon the determination that the statute represented a rational state policy. It stated:
“It would be strange indeed, and doctrinaire, for this Court, applying such broad constitutional concepts as due process and equal protection of the laws, to deny a State the power to assure a proper diffusion of political initiative as between its thinly populated counties and those having concentrated masses, in view of the fact that the latter have practical opportunities for exerting their political weight at the polls not available to the former.” Id., at 284. (Emphasis supplied.)
The other cases upon which my Brethren dwell are all distinguishable or inapposite. The widely heralded case of Colegrove v. Green, 328 U. S. 549 (1946), was one not only in which the Court was bobtailed but in which there was no majority opinion. Indeed, even the “political question” point in Mr. Justice Frankfurter's opinion was no more than an alternative ground.1 Moreover, the appellants did not present an equal protection argument.2 While it has served as a Mother Hubbard to most of the subsequent cases, I feel it was in that respect ill-cast and for all of these reasons put it to one side.3 Like*253wise, I do not consider the Guaranty Clause cases based on Art. I, § 4, of the Constitution, because it is not invoked here and it involves different criteria, as the Court’s opinion indicates. Cases resting on various other considerations not present here, such as Radford v. Gary, 352 U. S. 991 (1957) (lack of equity); Kidd v. McCanless, 352 U. S. 920 (1956) (adequate state grounds supporting the state judgment); Anderson v. Jordan, 343 U. S. 912 (1952) (adequate state grounds); Remmey v. Smith, 342 U. S. 916 (1952) (failure to exhaust state procedures), are of course not controlling. Finally, the Georgia county-unit-system cases, such as South v. Peters, 339 U. S. 276 (1950), reflect the viewpoint of MacDougall, i. e., to refrain from intervening where there is some rational policy behind the State’s system.4
II.
The controlling facts cannot be disputed. It appears from the record that 37% of the voters of Tennessee elect 20 of the 33 Senators while 40% of the voters elect 63 of the 99 members of the House. But this might not on its face be an “invidious discrimination,” Williamson v. Lee Optical of Oklahoma, 348 U. S. 483, 489 (1955), for a “statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it.” McGowan v. Maryland, 366 U. S. 420, 426 (1961).
It is true that the apportionment policy incorporated in Tennessee’s Constitution, i. e., state-wide numerical equality of representation with certain minor qualifications,5 is a rational one. On a county-by-county compari*254son a districting plan based thereon naturally will have disparities in representation due to the qualifications. But this to my mind does not raise constitutional problems, for the overall policy is reasonable. However, the root of the trouble is not in Tennessee’s Constitution, for admittedly its policy has not been followed. The discrimination lies in the action of Tennessee’s Assembly in allocating legislative seats to counties or districts created by it. \ Try as one may, Tennessee’s apportionment just cannot be made to fit the pattern cut by its Constitution. This was the finding of the District Court. The policy of the Constitution referred to by the dissenters, therefore, is of no relevance here. We must examine what the Assembly has done.6 The frequency and magnitude of the inequalities in the present districting admit of no policy whatever. An examination of Table I accompanying this opinion, post, p. 262, conclusively reveals that the apportionment picture in Tennessee is a topsy-turvical of gigantic proportions. This is not to say that some of the disparity cannot be explained, but when the entire table is examined — comparing the voting strength of counties of like population as well as contrasting that of the smaller with the larger counties — it leaves but one conclusion, namely that Tennessee’s apportionment is a crazy quilt without rational basis. At the risk of being accused of picking out a few of the horribles I shall allude to a series of examples that are taken from Table I.
*"~As is admitted, there is a wide disparity of voting strength between the large and small counties. Some *255samples are: Moore County has a total representation of two7 with a population (2,340) of only one-eleventh of Rutherford County (25,316) with the same representation; Decatur County (5,563) has the same representation as Carter (23,303) though the latter has four times the population; likewise, Loudon County (13,264), Houston (3,084), and Anderson County (33,990) have the same representation, i. e., 1.25 each. But it is said that in this illustration all of the under-represented counties contain municipalities of over 10,000 population and they therefore should be included under the “urban” classification, rationalizing this disparity as an attempt to effect a rural-urban political balance. But in so doing one is caught up in the backlash of his own bull whip, for many counties have municipalities with a population exceeding 10,000, yet the same invidious discrimination is present. For example:
County Population Representation
Carter . 23,303 1.10
Maury. 24,556 2.25
Washington CO CO CO O ccT co
Madison .. Cji O rti <N co
*256Likewise, counties with no municipality of over 10,000 suffer a similar discrimination:
Population Representation County
Grundy ... 'cn rf*. o o co CK
Chester ... OS "CO CO J — 1 to o o
Cumberland CO 05 xo o os CO
Crockett .. CO LCO bO b O
13,264 Loudon ... M to Ox
13,577 Fayette ... W cn O
This could not be an effort to attain political balance between rural and urban populations^' Since discrimination is present among counties of like population, the plan is neither consistent nor rational. It discriminates horizontally creating gross disparities between rural areas themselves as well as between urban areas themselves,8 still maintaining the wide vertical disparity already pointed out between rural and urban.
It is also insisted that the representation formula used above (see n. 7) is “patently deficient” because “it eliminates from consideration the relative voting power of the counties that are joined together in a single election district.” This is a strange claim coming from those who rely on the proposition that “the voice of every voter” need not have “approximate equality.” Indeed, representative government, as they say, is not necessarily one of “bare numbers.” The use of floterial districts in our political system is not ordinarily based on the theory that the floterial representative is splintered among the counties of his district per relative population. His function is to represent the whole district. However, I shall meet the charge on its own ground and by use of its “adjusted *257‘total representation’ ” formula show that the present apportionment is loco. For example, compare some “urban” areas of like population, using the Harlan formula:
Representation County Population
2.65 Washington . 36,967
4.87 Madison . 37,245
1.48 Carter . 23,303
2.05 Greene . 23,649
3.81 Maury. 24,556
Coffee . 13,406 Cm CO C3
Hamblen . 14,090 £-» O t“H
And now, using the same formula, compare some so-called “rural” areas of like population:
County Population Representation
Moore . 2,340 1.23
Pickett . 2,565 .22
Stewart . co CO lO O CO r — ¡
Cheatham co CO to N
6,391 Chester . CD CO r-í
6,540 Grundy . 05 CD
8,731 Smith ... ^ O
8,787 Unicoi .. O ^ O
And for counties with similar representation but with gross differences in population, take:
County Population Representation
Sullivan. 55,712 4.07
Maury. 24,556 3.81
Blount. 30,353 2.12
Coffee . 13,406 2.32
These cannot be “distorted effects,” for here the same formula proposed by the dissenters is used and the result is even “a crazier” quilt.
*258The truth is that — although this case has been here for two years and has had over six hours’ argument (three times the ordinary case) and has been most carefully considered over and over again by us in Conference and individually — no one, not even the State nor the dissenters, has come up with any rational basis for Tennessee’s apportionment statute.
No one — except the dissenters advocating the Harlan “adjusted ‘total representation’ ” formula — contends that mathematical equality among voters is required by the Equal Protection Clause. But certainly there must be some rational design to a State’s districting. The discrimination here does not fit any pattern — -as I have said, it is but a crazy quilt. My Brother Harlan contends that other proposed apportionment plans contain disparities. Instead of chasing those rabbits he should first pause long enough to meet appellants’ proof of discrimination by showing that in fact the present plan follows a rational policy. Not being able to do this, he merely counters with such generalities as “classic legislative judgment,” no “significant discrepancy,” and “de minimis departures.” I submit that even a casual glance at the present apportionment picture shows these conclusions to be entirely fanciful. If present representation has a policy at all, it is to maintain the status quo of invidious discrimination at any cost. Like the District Court, I conclude that appellants have met the burden of showing “Tennessee is guilty of a clear violation of the state constitution and of the [federal] rights of the plaintiffs. . . .”
III.
Although I find the Tennessee apportionment statute offends the Equal Protection Clause, I would not consider intervention by this Court into so delicate a field if there were any other relief available to the people of Tennessee. But the majority of the people of Tennessee have no *259“practical opportunities for exerting their political weight at the polls” to correct the existing “invidious discrimination.” Tennessee has no initiative and referendum. I have searched diligently for other “practical opportunities” present under the law. I find none other than through the federal courts. The majority of the voters have been caught up in a legislative strait jacket. Tennessee has an “informed, civically militant electorate” and “an aroused popular conscience,” but it does not sear “the conscience of the people’s representatives.” This is because the legislative policy has riveted the present seats in the Assembly to their respective constituencies, and by the votes of their incumbents a reapportionment of any kind is prevented. The people have been rebuffed at the hands of the Assembly; they have tried the constitutional convention route, but since the call must originate in the Assembly it, too, has been fruitless. They have tried Tennessee courts with the same result,9 and Governors have fought the tide only to flounder. It is said that there is recourse in Congress and perhaps that may be, but from a practical standpoint this is without substance. To date Congress has never undertaken such a task in any State. We therefore must conclude that the people of Tennessee are stymied and without judicial intervention will be saddled with the present discrimination in the affairs of their state government.
IV.
Finally, we must consider if there are any appropriate modes of effective judicial relief. The federal courts are of course not forums for political debate, nor should they *260resolve themselves into state constitutional conventions or legislative assemblies. Nor should their jurisdiction be exercised in the hope that such a declaration as is made today may have the direct effect of bringing on legislative action and relieving the courts of the problem of fashioning relief. To my mind this would be nothing less than blackjacking the Assembly into reapportioning the State. If judicial competence were lacking to fashion an effective decree, I would dismiss this appeal. However, like the Solicitor General of the United States, I see no such difficulty in the position of this case. One plan might be to start with the existing assembly districts, consolidate some of them, and award the seats thus released to those counties suffering the most egregious discrimination. Other possibilities are present and might be more effective, But the plan here suggested would at least release the strangle hold now on the Assembly and permit it to redistrict itself.
In this regard the appellants have proposed a plan based on the rationale of state-wide equal representation. Not believing that numerical equality of representation throughout a State is constitutionally required, I would not apply such a standard albeit a permissive one. Nevertheless, the dissenters attack it by the application of the Harlan “adjusted ‘total representation’ ” formula. The result is that some isolated inequalities are shown, but this in itself does not make the proposed plan irrational or place it in the “crazy quilt” category. Such inequalities, as the dissenters point out in attempting to support the present apportionment as rational, are explainable. Moreover, there is no requirement that any plan have mathematical exactness in its application. Only where, as here, the total picture reveals incom-mensurables of both magnitude and frequency can it be said that there is present an invidious discrimination.
*261In view of the detailed study that the Court has given this problem, it is unfortunate that a decision is not reached on the merits. The majority appears to hold, at least sub silentio, that an invidious discrimination is present, but it remands to the three-judge court for it to make what is certain to be that formal determination. It is true that Tennessee has not filed a formal answer. However, it has filed - voluminous papers and made extended arguments supporting its position. At no time has it been able to contradict the appellants’ factual claims; it has offered no rational explanation for the present apportionment; indeed, it has indicated that there are none known to it. As I have emphasized, the case proceeded to the point before the three-judge court that it was able to find an invidious discrimination factually present, and the State has not contested that holding here. In view of all this background I doubt if anything more can be offered or will be gained by the State on remand, other than time. Nevertheless, not being able to muster a court to dispose of the case on the merits, I concur in the opinion of the majority and acquiesce in the decision to remand. However, in fairness I do think that Tennessee is entitled to have my idea of what it faces on the record before us and the trial court some light as to how it might proceed.
As John Rutledge (later Chief Justice) said 175 years ago in the course of the Constitutional Convention, a chief function of the Court is to secure the national rights.10 Its decision today supports the proposition for which our forebears fought and many died, namely, that to be fully conformable to the principle of right, the form of government must be representative.11 That is the keystone upon which our government was founded *262and lacking which no republic can survive. It is well for this Court to practice self-restraint and discipline in constitutional adjudication, but never in its history have those principles received sanction where the national rights of so many have been so clearly infringed for so long a time. National respect for the courts is more enhanced through the forthright enforcement of those rights rather than by rendering them nugatory through the interposition of subterfuges. In my view the ultimate decision today is in the greatest tradition of this Court.
TABLE I.
County 1950 voting population Present total representation using J. Clark's formula Present total representation using J. Harlan's formula Proposed total representation (appellants' plan), using J. Harlan's formula
Van Burén. 2,039 .63 .23 ,ii
Moore . 2,340 2.00 1.23 .18
Pickett. 2,565 .70 .22 .24
Sequatchie . 2,904 .63 .33 .19
Meigs. 3,039 .93 .48 .17
Houston. 3,084 1.25 .46 .24
Trousdale . 3,351 1.33 .43 .12
Lewis. 3,413 1.25 .39 .25
Perry. 3,711 1.50 .71 .40
Bledsoe . 4,198 .63 .49 .24
Clay . 4,528 .70 .40 .42
Union 4,600 .76 .37 .45
Hancock . 4,710 .93 .62 .49
Stewart . 5,238 1.75 1.60 .41
Cheatham. 5,263 1.33 .72 .20
Cannon . 5,341 2.00 1.43 .52
Decatur . 5,563 1.10 .79 .52
Lake . 6,252 2.00 1.44 .41
Chester . 6,391 2.00 1.36 .19
Grundy . 6,540 .95 .69 .43
Humphreys. 6,588 1.25 1.39 .72
Johnson. 6,649 1.10 .42 .43
*263County 1950 voting population Present total representation using J. Clark’s formula Present total representation using J. Harlan’s formula Proposed total representation (appellants’ plan), using J. Harlan’s formula
Jackson. 6,719 1.50 1.43 .63
DeKalb . 6,984 2.00 1.56 .68
Benton . 7,023 1.10 1.01 .66
Fentress . 7,057 .70 .62 .64
Grainger . 7,125 .93 .94 .65
Wayne. 7,176 1.25 .69 .76
Polk. 7,330 1.25 .68 .73
Hickman. 7,598 2.00 1.85 .80
Macon. 7,974 1.33 1.01 .61
Morgan. 8,308 .93 .59 .75
Scott . 8,417 .76 .68 .62
Smith. 8,731 2.50 2.04 .67
Unicoi . 8,787 .93 .40 .63
Rhea . 8,937 .93 1.42 .21
White . 9,244 1.43 1.69 .90
Overton ..'. 9,474 1.70 1.83 .89
Hardin. 9,577 1.60 1.61 .93
Cumberland . 9,593 .63 1.10 .87
Crockett . 9,676 2.00 1.66 .63
Henderson . 10,199 1.50 .78 .96
Marion . 10,998 1.75 1.73 .72
Marshall . 11,288 2.50 2.28 .84
Dickson. 11,294 1.75 2.29 1.23
Jefferson . 11,359 1.10 .87 1.03
MeNairy. 11,601 1.60 1.74 1.13
Cocke . 12,572 1.60 1.46 .89
Sevier. 12,793 1.60 1.47 .69
Claiborne . 12,799 1.43 1.61 1.34
Monroe . 12,884 1.75 1.68 1.30
Loudon . 13,264 1.25 .28 .52
Warren . 13,337 1.75 1.89 1.68
Coffee . 13,406 2.00 2.32 1.68
Hardeman . 13,565 1.60 1.86 1.11
Fayette . 13,577 2.50 2.48 1.11
Haywood . 13,934 2.50 2.52 1.69
Williamson . 14,064 2.33 2.96 1.71
*264 County 1950 voting population Present total representation using J. Clark’s formula Present total representation using J. Harlan’s formula Proposed total representation (appellants’ plan), using J. Harlan’s formula
Hamblen. 14,090 1.10 1.07 1.67
Franklin . 14,297 1.75 1.95 1.73
Lauderdale . 14,413 2.50 2.45 1.73
Bedford. 14,732 2.00 1.45 1.74
Lincoln . 15,092 2.50 2.72 1.77
Henry . 15,465 2.83 2.76 1.73
Lawrence . 15,847 2.00 2.22 1.81
Giles . 15,935 2.25 2.54 1.81
Tipton. 15,944 '3.00 1.68 1.13
Robertson. 16,456 2.83 2.62 1.85
Wilson. 16,459 3.00 3.03 1.21
Carroll. 16,472 2.83 2.88 1.82
Hawkins . 16,900 3.00 1.93 1.82
Putnam. 17,071 1.70 2.50 1.86
Campbell . 17,477 .76 1.40 1.94
Roane . 17,639 1.75 1.26 1.30
Weakley . 18,007 2.33 2.63 1.85
Bradley. 18,273 1.25 1.67 1.92
McMinn . 18,347 1.75 1.97 1.92
Obion. 18,434 2.00 2.30 1.94
Dyer . 20,062 2.00 2.36 2.32
Sumner . 20,143 2.33 3.56 2.54
Carter . 23,303 1.10 1.48 2.55
Greene. 23,649 1.93 2.05 2.68
Maury. 24,556 2.25 3.81 2.85
Rutherford. 25,316 2.00 3.02 2.39
Montgomery. 26,284 3.00 3.73 3.06
Gibson. 29,832 5.00 5.00 2.86
Blount. 30,353 1.60 2.12 2.19
Anderson. 33,990 1.25 1.30 3.62
Washington . 36,967 1.93 2.65 3.45
Madison . 37,245 3.50 4.87 3.69
Sullivan. 55,712 3.00 4.07 5.57
Hamilton. 131,971 6.00 6.00 15.09
Knox . 140,559 7.25 8.96 15.21
Davidson . 211,930 12.50 12.93 21.57
Shelby. 312,345 15.50 16.85 31.59
The opinion stated at 551 that the Court “could also dispose of this case on the authority of Wood v. Broom [287 U. S. 1 (1932)].” Wood v. Broom involved only the interpretation of a congressional reapportionment Act.
Similarly, the Equal Protection Clause was not invoked in Tedesco v. Board of Supervisors, 339 U. S. 940 (1950).
I do not read the later case of Colegrove v. Barrett, 330 U. S. 804 (1947), as having rejected the equal protection argument adopted here. That was merely a dismissal of an appeal where the equal protection point was mentioned along with attacks under three other constitutional provisions, two congressional Acts, and three state constitutional provisions.
Georgia based its election system on a consistent combination of political units and population, giving six unit votes to the eight most populous counties, four unit votes to the 30 counties next in population, and two unit votes to each of the remaining counties.
See Part I of the Appendix to Mr. Justice Harlan's dissent, post, p. 341.
It is suggested that the districting is not unconstitutional since it was established by a statute that was constitutional when passed some 60 years ago. But many Assembly Sessions since that time have deliberately refused to change the original act, and in any event “[a] statute [constitutionally] valid when enacted may become invalid by change in the conditions to which it is applied.” Nashville, C. & St. L. R. Co. v. Walters, 294 U. S. 405, 415 (1935).
“Total representation” indicates the combined representation in the State Senate (33 members) and the State House of Representatives (99 members) in the Assembly of Tennessee. Assuming a county has one representative, it is credited in this calculation with 1/99. Likewise, if the same county has one-third of a senate seat, it is credited with another 1/99, and thus such a county, in our calculation, would have a “total representation” of two; if a county has one representative and one-sixth of a senate seat, it is credited with 1.5/99, or 1.50. It is this last figure that I use here in an effort to make the comparisons clear. The 1950 rather than the 1960 census of voting population is used to avoid the charge that use of 1960 tabulations might not have allowed sufficient time for the State to act. However, the 1960 picture is even more irrational than the 1950 one.
Of course this was not the case in the Georgia county unit system, South v. Peters, supra, or the Illinois initiative plan, MacDougall v. Green, supra, where recognized political units having independent significance were given minimum political weight.
It is interesting to note that state judges often rest their decisions on the ground that this Court has precluded adjudication of the federal claim. See, e. g., Scholle v. Secretary of State, 360 Mich. 1, 104 N. W. 2d 63 (1960).
I Farrand, The Records of the Federal Convention of 1787, 124.
Kant, Perpetual Peace.
See Wood v. Broom, 287 U. S. 1; Colegrove v. Green, 328 U. S. 549, rehearing denied, 329 U. S. 825, motion for reargument before the full bench denied, 329 U. S. 828; Cook v. Fortson, 329 U. S. 675, rehearing denied, 329 U. S. 829; Turman v. Duckworth, 329 U. S. 675, rehearing denied, 329 U. S. 829; Colegrove v. Barrett, 330 U. S. 804; MacDougall v. Green, 335 U. S. 281; South v. Peters, 339 U. S. 276; Tedesco v. Board of Supervisors, 339 U. S. 940; Remmey v. Smith, 342 U. S. 916; Cox v. Peters, 342 U. S. 936, rehearing denied, 343 U. S. 921; Anderson v. Jordan, 343 U. S. 912; Kidd v. McCanless, 352 U. S. 920; Radford v. Gary, 352 U. S. 991; Hartsfield v. Sloan, 357 U. S. 916; Matthews v. Handley, 361 U. S. 127; Perry v. Folsom, 144 F. Supp. 874 (D. C. N. D. Ala.); Magraw v. Donovan, 163 F. Supp. 184 (D. C. D. Minn.); cf. Dyer v. Kazuhisa Abe, 138 F. Supp. 220 (D. C. D. Hawaii). And see Keogh v. Neely, 50 F. 2d 685 (C. A. 7th Cir.).