Plaintiffs appeal from an adverse summary judgment against their challenge to the constitutionality of the districting plan for elections to the County Commission of Dallas County, Alabama. They contend that Dallas County's at-large election system, which requires that candidates be residents of certain districts that are not equally populated, violates the Fourteenth Amendment. Relying on Dusch v. Davis, 387 U.S. 112, 87 S. Ct. 1554, 18 L. Ed. 2d 656 (1964), the District Court reasoned that the at-large feature of the plan preserved it against constitutional attack. In our view the plan denies 'fair and effective representation' in violation of the Fourteenth Amendment, Reynolds v. Sims, 377 U.S. 533, 565, 84 S. Ct. 1362, 1383, 12 L. Ed. 2d 506 (1964), and so we reverse and remand.

The Statement of Agreed Facts submitted to the trial court shows that the County Commission is the governing body of Dallas County. Its responsibilities include, among others, building roads and bridges, managing and disposing of county properties, providing sewerage improvements, enforcing state health laws, settling legal claims against the county, and raising and investing county revenues. 12 Ala.Code 11, 12 (1958, 1973 Cum.Supp.).

For the purpose of electing county commissioners Dallas County is divided into four districts (City, South, West, and Fork). Laws of Alabama No. 328 (Feb. 8, 1901) (as amended).1 The City District, containing the City of Selma, has a population of 27,379-- approximately half of the total for the county. The West, South, and Fork Districts, which are rural areas, have populations of 6,209, 14,203, and 7,505 respectively. Four commissioners and a probate judge, who can vote only to break ties, are elected from these districts. Although all the voters of Dallas County vote for all four commissioners, only one of the candidates residing in each district may be elected.2 The probate judge may reside in any of the districts.

II. Plaintiff's Prima Facie Case of Invidious Discrimination

This case does not present the one man-- one vote issue in its conventional form. It is undisputed that the votes cast in Dallas County elections are of equal weight. Each voter can vote for four candidates, and the election is county-wide. Rather this case is an instance of dilution of equally weighted votes-- of 'minimiz(ing) or cancel(ling) out the voting strength' of a group that has an identifiable set of common interests. Fortson v. Dorsey, 379 U.S. 433, 439, 85 S. Ct. 498, 501, 13 L. Ed. 2d 401 (1965). The one man-- one vote principle is violated when some votes carry more weight than others. Dilution, in contrast, minimizes the impact of one group's votes, even though they are equal in weight to nondiluted votes.3 Recent decisions have recognized that 'access to the political process and not population is the barometer of dilution of minority voting strength.' Zimmer v. McKeithen, 5 Cir., 1973, 485 F.2d 1297, 1303 & n. 14. See White v. Register, 412 U.S. 755, 765, 93 S. Ct. 2332, 2339, 37 L. Ed. 2d 314 (1973).

Dilution, or restriction of 'access to the political process,' is usually accomplished by drawing district lines either to disperse the votes of one faction so that they cannot influence the outcome of elections, or to concentrate those votes in as few districts as possible, thus wasting their strength. In this case none of the familiar devices of dilution, such as gerrymandering, is present. The dilution here is alleged to be accomplished by a method usually encountered in one man-- one vote cases: burdening the voters of a particular political unit with an explicit, numerical disadvantage.

B. Testing the Dallas County Plan for Dilution

Because of the explicit, numerical disparity created by the residency requirement, the injury this plan inflicts on Selma voters is apparent on its face. The citizens of Selma are forbidden to elect resident commissioners in proportion to their numbers.4 They must select candidates who reside not in their subdistrict but who reside in the rural subdistricts to represent Selma's interests. It is to be expected that commissioners who are elected from rural subdistricts will give greater priority to needs of the subdistricts in which they reside than to the interests of the City of Selma. Unlike the residents of Selma, the other voters in Dallas County can choose their proportional share of the representative body from among rural candidates who, by reason of their residence in rural areas, can be expected to share their interests.

The Supreme Court and this Court have repeatedly stated that 'lack of provision for at-large candidates running from particular geographical subdistricts' is a signal of unconstitutional dilution of one group's voting strength. Zimmer v. McKeithen, 5 Cir., 1973, 485 F.2d 1297, 1305. 'If districts are not appropriately subdistricted to assure distribution of legislators that are resident over the entire district,' an important element of fair and effective presentation is sacrificed. Burns v. Richardson, 384 U.S. 73, 88, 86 S. Ct. 1286, 1294, 16 L. Ed. 2d 376 (1966). See White v. Regester, 412 U.S. 755, 766 n. 10, 93 S. Ct. 2332, 2340, 37 L. Ed. 2d 314 (1973).

The clear implication of this conclusion is that having representatives who live among constituents is vital to maintain effective representation. Therefore, a districting scheme that inhibits the right of the residents of a subdistrict from electing representatives living in their own area is undeniably an injury. See Moore v. Leflore County Bd. of Election Com'rs, N.D.Miss., 1973, 361 F. Supp. 609, 613, aff'd, 5 Cir., 1974, 502 F.2d 621.

Dallas County has a subdistrict residency requirement, but it falls upon the four subdistricts unequally. The number of commissioners permitted to reside in Selma is less, in proportion to its population, than the number permitted to reside in the other three subdistricts. Selma-- the County's urban area, with half the total population-- is prohibited from having more than one resident commissioner, whereas the three remaining subdistricts-- all rural, with the other half of the population-- have three resident commissioners. Thus Selma, the urban community, is outvoted three to one by the rural areas, even though its population is virtually equal to their populations.

Dilution is established if a districting scheme works an invidious effect on an identifiable group. See, e.g., White v. Regester, supra, 412 U.S. at 765, 93 S. Ct. at 2339. In this case identifying the disadvantaged group is simple, because it is described by the boundaries of Dallas County's City District. The discrimination, however, is not based on place of residence alone. The group of voters adversely affected by the plan is virtually coextensive with the group of city dwellers in Dallas County. The voters of Selma confront not only the common concerns that develop among residents of the same area, but also a perspective on local issues that the rural population, with its district social and economic interests, is unlikely to share. Defendants' expression of concern at the prospect of a city-dominated county government5 clearly shows the divergence of interests. Thus the discriminatory effect of the Dallas County plan is political as well as geographical.

Finding an adverse effect on an identifiable group of voters is not sufficient in itself to establish dilution. The effect must be invidiously discriminatory. See Reynolds v. Sims, supra, 377 U.S. at 561, 84 S. Ct. at 1381. Multimember districts are not unconstitutional per se, but they are not constitutional per se either. See White v. Regester, supra, 412 U.S. 765, 93 S. Ct. at 2339. The circumstances of the present case differ significantly from most of the decisions involving multimember districts. The Supreme Court and this Court have upheld multimember districting schemes with residency requirements somewhat similar to the one challenged here. See Fortson v. Dorsey, 379 U.S. 433, 85 S. Ct. 498, 13 L. Ed. 2d 401 (1965); Goldblatt v. City of Dallas, 5 Cir., 1969, 414 F.2d 774.6 In these cases, however, the subdistricts were equally populated and thus did not cause the discrimination that appears in the present case.7

The Supreme Court has reviewed a multimember plan with unequally populated residency subdistricts only once. In Dusch v. Davis, 387 U.S. 112, 87 S. Ct. 1554, 18 L. Ed. 2d 656 (1967), the statute under review provided for a county government of eleven representatives elected at large by the voters of the seven unequally populated boroughs of Virginia Beach, Virginia. That plan resembles the Dallas County arrangement in that residency requirements provided that seven of the representatives must live in different boroughs. The other four representatives, however, were elected without regard to residence. The Court held the plan constitutional. This Court has held, however, that the Dusch-type plan is not immune from constitutional attack. See Lipscomb v. Josson, 5 Cir., 1972, 459 F.2d 335. Like any other multimember plan, the Virginia Beach plan cannot be used to perpetrate a 'denial of effective participation in the political process.' Lipscomb v. Jonsson, supra, at 339.8

We conclude that the circumstances of the present case differ significantly from Dusch. In that case the presence of the four positions unrestricted as to residency assured that the heavily populated boroughs would not be forced to depend on representatives from the sparsely populated ones. The effect of the plan was to give all voters an equal opportunity to choose their share of the members of the representative body from their own area. Thus as an instrument of fair and effective representation the plan could not be said to create a 'distinction on the basis of race, creed, or economic status or location.' Dusch v. Davis, supra, 387 U.S. at 115, 87 S. Ct. at 1555.

A panel of this Court articulated this reading of the Dusch case in Keller v. Gilliam, 5 Cir., 1972, 454 F.2d 55. Keller presented facts indistinguishable from those of the present case. Judge Tuttle found important differences between Keller and Dusch:

What was clear in the Virginia Beach (Dusch) case is that was no combination by which representatives residing in the districts comprising a minority of the population could combine to work their will over the majority residing in the heavier populated precincts, except with the concurrence of some of the four councilmen who could reside in any of the boroughs and who were elected at large. Quite the contrary is obviously true here.

454 F.2d at 56. In Keller, as in the present case, preserving majority rule was not possible, and the plan was struck down. We believe Keller is sound and decline defendant's invitation to overrule it.9 See Moore v. Leflore County Bd. of Election Com'rs, N.D.Miss., 1973, 361 F. Supp. 609, aff'd 5 Cir., 1974, 502 F.2d 621 (Leflore III); id., N.D.Miss., 1973, 361 F. Supp. 603 (Leflore II); id., N.D.Miss., 1973, 351 F. Supp. 848 (Leflore I).

In Davis v. Thomas County, 5 Cir., 1967, 380 F.2d 93, decided one month after Dusch v. Davis, supra, a panel of this Court upheld an at-large election-- subdistrict residency plan in which the districts were unequally populated. In light of our belief in the correctness of Keller, the per curiam opinion in Thomas County cannot be followed.

The result directly supported by Keller is also in harmony with the Supreme Court's two most recent decisions on the dilution issue. In Gaffney v. Cummings, 412 U.S. 735, 93 S. Ct. 2321, 37 L. Ed. 2d 298 (1973), the Court upheld a legislative reapportionment plan admittedly drawn to reflect the statewide political strengths of the major parties. The use of political criteria was permissible because no invidious discrimination occurred. Although a group of voters in a particular district might have exerted more influence if the lines had been drawn differently, no identifiable political faction was treated differently from any other.

In White v. Regester, 412 U.S. 755, 93 S. Ct. 2332, 37 L. Ed. 2d 314 (1973), the Court struck down multimember legislative districting plans because minorities in the area historically had been officially excluded from the electoral process. These obstacles, the Court reasoned, prevented black and Mexican-American groups from functioning as effective political forces in the context of a multimember district.

The present case is more like Regester than Gaffney. Concededly the discrimination here may be less severe than in Regester, insofar as city dwellers are a more heterogeneous faction than blacks or Mexican-Americans. On the other hand, the discrimination here is current, explicit, and official, unlike the informal historical disadvantages that supported the decision in Regester. 412 U.S. at 766-769, 93 S.Ct. at 2340.

It is unclear whether dilution of a group's voting power is unconstitutional only if deliberate. See Comment, Political Gerrymandering: A Statutory Compactness Standard as an Antidote for Judicial Impotence, 41 U.Chi.L.Rev. 398, 408-409 (1974). In any event, the intention underlying the Dallas County plan is unambiguous. Although intent often can be inferred from the effects of a districting scheme, see Wright v. Rockefeller, 376 U.S. 52, 56-57, 84 S. Ct. 603, 605, 11 L. Ed. 2d 512 (1964), in this case there are explicit declarations to consider.

Defendants assert that the districting plan is designed to assure 'a governing body which is knowledgeable of subjects relating to agriculture' and to prevent the city from establishing a 'tyranny of the majority' that would 'encroach upon legitimate minority interests.' By its sponsors' own admission, the plan is designed to 'minimize or cancel out' the voting strength of one group of voters. The Supreme Court has repeatedly declared that voters may not be excluded from the electoral process out of concern for the way they might vote. See Evans v. Cornman, 398 U.S. 419, at 423, 90 S. Ct. 1752, at 1755, 26 L. Ed. 2d 370 (1970); Cipriano v. City of Houma, 395 U.S. 701, 705-706, 89 S. Ct. 1897, 1900, 23 L. Ed. 2d 647 (1969); Carrington v. Rash, 380 U.S. 89, 93-94, 85 S. Ct. 775, 778-779, 13 L. Ed. 2d 675 (1965). This case presents a less extreme form of the same strategy. Instead of excluding a group of voters outright, the county has weakened their ability to participate by restricting unequally their opportunity to elect candidates from their own area.

Plaintiffs have established a strong prima facie case of official discrimination intended to dilute their voting strength. Unless some justification can be discovered that rebuts this prima facie case, the discrimination is invidious and violates the Fourteenth Amendment. See, e.g., Kilgarlin v. Hill, 386 U.S. 120, 122, 87 S. Ct. 820, 822, 17 L. Ed. 2d 771 (1967).

The standard of review by which justifications of the plan are to be judged is not entirely clear. See Casper, Apportionment and the Right to Vote: Standards of Judicial Scrutiny, 1973 Sup.Ct. Rev. 1, 29-32. Compare Kramer v. Union Free School District, 395 U.S. 621, 89 S. Ct. 1886, 23 L. Ed. 2d 583 (1969), with McDonald v. Board of Election Com'rs of Chicago,394 U.S. 802, 89 S. Ct. 1404, 22 L. Ed. 2d 739 (1969). The most recent articulation of the standards of review under the Equal Protection Clause appears in Dunn v. Blumstein, 405 U.S. 330, 335, 92 S. Ct. 995, 999, 31 L. Ed. 2d 274 (1972):

To decide whether a law violates the Equal Protection Clause, we look, in essence, to three things: the character of the classification in question; the individual interests affected by the classification; and the governmental interests asserted in support of the classification.

The Court concluded that statutes burdening the right to vote 'must meet close constitutional scrutiny' and be 'necessary to promote a compelling state interest.' Id. at 336-337, 92 S.Ct. at 1000. See Kramer v. Union Free School District, supra; Bullock v. Carter, 405 U.S. 134, 143, 92 S. Ct. 849, 855-856, 31 L. Ed. 2d 92.

This standard has been applied irrespective of any political or racial implications in the one man-one vote cases concerning congressional districts. Disparities in the weight accorded the votes of residents of different places has been sufficient to trigger strict scrutiny, even though location is not a suspect classification. See, e.g., Ellis v. Mayor and City Council of Baltimore, 4 Cir., 1965, 352 F.2d 123, 129. Here the presence of political discrimination, which the Supreme Court has suggested as a suspect classification,10 underscores the propriety of a strict standard of review.

In the one man-one vote cases, greater deference has been shown toward apportionment plans for state legislative districts than for congressional districts.11 Mahan v. Howell, 410 U.S. 315, 93 S. Ct. 979, 983-985, 35 L. Ed. 2d 320 (1973). No such distinction has appeared in adjudication of the dilution issue. There is, to be sure, a strong suggestion of judicial deference to legislative judgment attacked on grounds of dilution through political discrimination. See Gaffney v. Cummings, supra, 412 U.S. at 752-753, 93 S. Ct. at 2331-2332. This deference, however, is motivated by wholly different policies from those that underlie the deference shown in Mahan v. Howell, supra. Apportionment plans for state legislative districts that deviate from the numerical ideal of equally weighted votes are tolerated out of respect for important values of federalism. The federal judiciary is reluctant to apply the principles of Equal Protection rigidly where the vitality of local political institutions and subdivisions is at stake. See Mahan v. Howell, supra, 410 U.S. at 325-330, 93 S. Ct. at 985-987. Thus a more moderate standard of review is appropriate.

In the dilution cases, however, regard for federalism is not the reason for judicial restraint, but rather respect for the institutional limitations on the courts' ability to gauge the ramifications of districting patters. Intentional dilution of a group's voting strength is a serious affront to the Constitution. Establishing this injury in court, however, is often very difficult. In most dilution cases the convenient numerical guidelines available in conventional one man-one vote cases are absent. Instead the courts must evaluate evidence of the political alignments of allegedly disadvantaged factions and infer the intent of the legislature from actions that may have several plausible motives. It is not surprising, therefore, that so-called 'partisan gerrymandering,' Casper, Apportionment and the Right to Vote: Standards of Judicial Scrutiny, 1973 Sup.Ct.Rev. 1, 24, is more difficult to verify than racial gerrymandering, because dilution on the basis of race can be demonstrated by reliable demographic data.12 Compare Gaffney v. Cummings, 412 U.S. 735, 93 S. Ct. 2321, 37 L. Ed. 2d 298 (1973) with Gomillion v. Lightfoot, 364 U.S. 339, 81 S. Ct. 125, 5 L. Ed. 2d 110 (1960). The difficulty of proof, however, is the principal differences.

The kinds of justifications that may rebut a prima facie showing of dilution are not as clearly defined as the justifications the courts entertain in defense of deviations from numerical equality. See Reynolds v. Sims,supra, 377 U.S. at 622-623, 84 SCt. at 1413 (Harlan, J., dissenting). Adjudication of the latter issue long ago established that preservation of rural-urban balance in a legislative body is a categorically unacceptable reason for departing from the one man-one vote standard. See, e.g., Davis v. Mann, 377 U.S. 678, 692, 84 S. Ct. 1441, 1448, 12 L. Ed. 2d 609 (1964). This purpose-- the protection of 'legitimate minority interests'-- is the only one offered to justify the discrimination caused by the Dallas County plan.

Although defendants urge that their concern is to ensure 'knowledgeable' representatives, their fear of a 'tyranny of the majority' exercised by the voters of Selma demonstrates that influence rather than mere expertise, is their objective. Moreover, defendants do not explain why the residents of Selma cannot be trusted to elect representatives who are qualified to protect their interests, if agriculture is indeed so vital to the entire county.

In Dusch v. Davis, supra, the Supreme Court suggested that if all votes are equally weighted, residency districts of unequal population that preserved agrarian influence might withstand constitutional scrutiny. The Court surmised that the plan under review 'seem(ed) to reflect a detente between urban and rural communities.' Id., 387 U.S. at 117, 87 S. Ct. at 1556. It is not clear that the Court intended to permit discriminatory plans to be justified by invoking the protection of agricultural interests as a purpose. The Virginia Beach plan, as we noted earlier, worked no real discrimination. Therefore the question whether protection of agricultural interests would justify a discriminatory plan was not squarely presented.

Even if Dusch does approve the use of unequally populated subdistricts under the circumstances of that case, in Dallas County there is no 'detente between urban and rural communities' as found in Dusch, but instead a domination by the latter. However sufficient the preservation of agricultural influence may have been as a justification for the Virginia Beach plan, which did not compromise the democratic principle of fair and effective representation, that purpose does not justify the burdens on electoral participation created by the Dallas County Plan. Under the views we have expressed, the districting plan is constitutionally impermissible.

We remand this case, therefore, for the fashioning of appropriate relief by the District Court under the principles herein announced.

Reversed and remanded.

COLEMAN, Circuit Judge (dissenting) with whom GEWIN, DYER, SIMPSON, CLARK and RONEY, Circuit Judges, join.

I regret that I must respectfully dissent.

A majority of the Court today holds, as I understand its opinion, that the statutory scheme for the election of county commissioners in Dallas County, Alabama, is prima facie unconstitutional, although the opinion admits, 'It is undisputed that the votes cast in Dallas County elections are of equal weight. Each voter can vote for four candidates, and the election is countywide.'

The record is totally bereft of any discrimination on the basis of race, creed, or other personal status, either individually or as a group.

Every commissioner, regardless of the place of residence, is answerable to every voter in the entire county, including the voters residing in Selma. While the record fails to reflect the voing strength of Selma as contrasted to the remainder of the county, it is pointed out that Selma contains approximately half the population of the county. It goes against human experience to hold as a matter of law, I think, that any commissioner residing in the rural area and answerable to the county as a whole would likely consider that his chances of retention in elective office would be enhanced by either neglecting or ignoring the best interests of Selma voters.

Therefore, I cannot agree with the language of the opinion which states (Page 883) that, 'It is to be expected that commissioners who are elected from rural subdistricts will give greater priority to needs of the subdistricts in which they reside than to the interests of the City of Selma'.

This case is before us on appeal from summary judgment and there is no evidence to support that statement.

I think the voters of Selma are getting better than equal treatment.

An examination of Title 12 (Counties) of the Code of Alabama indicates that the commissioners, along with the judge of probate, who is elected at-large as a general county official, have jurisdiction over county roads, bridges, and stock law districts, whereas they have no jurisdiction over the streets of Selma. The commissioners direct and control the property of the county, such as the courthouse. Within statutory limits they levy taxes for county purposes. They appropriate the money required for the maintenance of general county functions. They administer the liquidation of county bonded debt. They are required to publish semi-annually an itemized report of the receipts and expenditures of the county.

The governing authorities of Selma control similar functions within the corporate limits of that city. Citizens residing outside the corporate limits of Selma have no vote for officials or on other public issues in the city of Selma.

Yet, the citizens of Selma have a vote as to each and every commissioner of the entire county, including areas outside the city. Except for the levying of county taxes and expending county funds, the commissioners appear to perform very few functions of particular interest to the residents of Selma. Nevertheless, each and every commissioner on election day must submit to being weighed in the Selma scales, but the governing authorities of Selma do not have to go before the people of the county.

This is not a White v. Regester case. There the District Court found that ethnic minorities who were also a minority of the total population 'had less opportunity than did other residents in the district to participate in the political processes and to elect legislators of their choice.' The Supreme Court affirmed, upon consideration of the totality of the circumstances. There is nothing to indicate in the instant case that Selma voters have any less opportunity than do county residents to participate in the political processes or to elect commissioners of their own choice. Such rights do not include the right to be elected to office or to have a neighbor elected.

I offer the further thought that the 'factual' conclusions of the majority are wholly hypothetical. I think the Supreme Court decisions require proof of 'a real-life impact', Whitcomb v. Chavis, 403 U.S. 124, 91 S. Ct. 1858, 29 L. Ed. 2d 363 (1971). In Chavis, the Supreme Court said, 'The real-life impact of multi-member districts on individual voting power has not been sufficiently demonstrated, at least on this record, to warrant departure from prior cases', Id. at 146, 91 S.Ct. at 1870.

I would therefore adhere to the teachings of the Supreme Court in Dusch v. Davis, 387 U.S. 112, 87 S. Ct. 1554, 18 L. Ed. 2d 656 (1967), which, in the absence of distinction on the basis of race, creed, economic status, or location, held:

'The Seven-Four Plan makes no distinction on the basis of race creed, or economic status or location. Each of the 11 councilmen is elected by a vote of all the electors in the city. The fact that each of the seven councilmen must be a resident of the borough from which he is elected, is not fatal. In upholding a residence requirement for the election of state senators from a multi-district county we said in Fortson v. Dorsey, 379 U.S. 433, 438, 85 S. Ct. 498, 501, 13 L. Ed. 2d 401, 404: 'It is not accurate to treat a senator from a multi-district county as the representative of only that district within the county wherein he resides. The statute uses districts in multi-district counties merely as the basis of residence for candidates, not for voting or representation. Each district's senator must be a resident of that district, but since his tenure depends upon the county-wide electorate he must be vigilant to serve the interests of all the people in the county, and not merely those of people in his home district; thus, in fact he is the county's and not merely the district's senator.' 'By analogy the present consolidation plan uses boroughs in the city 'merely as the basis of residence for candidates, not for voting or representation.' He is nonetheless the city's, not the borough's, councilman. In Fortson there was substantial equality of population in the senatorial districts, while here the population of the boroughs varies widely. If a borough's resident on the council represented in fact only the borough, residence being only a front, different conclusions might follow. But on the assumption that Reynolds v. Sims controls, the constitutional test under the Equal Protection Clause is whether there is an 'invidious' discrimination.' (387 U.S. at page 115, 87 S.Ct. at page 1556, 18 L.Ed.2d at page 659)

With deference, nothing in this case justifies an effort to undermine the teachings of Dusch. I therefore respectfully dissent.

Separate Opinion of COLEMAN, Circuit Judge.

Since I wrote the dissenting opinion in which Judges Gewin, Dyer, Simpson, Clark, and Roney joined, the majority opinion has been amended (as the majority had a right to do) by including in Footnote 9 a reference to Lytle v. Commissioners of Election of Union County, 4 Cir., 1974, 509 F.2d 1049 (No. 74-1619, June 17, 1974). The author of the majority opinion has graciously delayed its release to give me an opportunity to file this separate opinion, which I have not circulated to the other Judges for possible concurrences as I am reluctant to cause any further delay in the release of the en banc opinion.

Lytle v. Commissioners of Election, supra, was consolidated and decided with McCain v. Lybrand in the Fourth Circuit. Both cases, of course, had common issues. Lytle was concerned with the election of the Board of Control in Union County, South Carolina. Lybrand involved similar elections in Edgefield County. In both counties the members of the Board of Control were elected at large but each member had to reside in separate fixed districts of the county. The District Court had invalidated the system in both counties. The Fourth Circuit (Judges Craven, Russell, and Field) affirmed as to Union but reversed as to Edgefield. The unsuccessful appellees in Edgefield sought certiorari. On November 25, 1974, the Supreme Court denied review, McCain v. Lybrand, U.S. , 95 S. Ct. 515, 42 L. Ed. 2d 308 (No. 74-278, 43 USLW 3303). I have not had an opportunity to see the petition as filed in the Supreme Court but the point as recited in USLW reads as follows:

'County officials elected at large may validly be required to reside in separate fixed districts of county, even though districts have wide population variances; attempt through residency requirement to strike balance between interests of rural and urban areas in county, provided voting rights for both areas are protected by at-large elections, meets 'rational' test; county plan represents proper balancing of rural-urban interests and is without such population variances among districts as to require deletion of residency requirements on constitutional grounds.'

In Union County, the urban area composed 58% Of the population but only one member of the eight man board was required to reside in that area. It was pointed out that Union District (the urban area) had 17,127 people, Goshen Hill, had 561. Each had a residential member of the Board. There were no floating members as in Dusch.

This was held invalid, but the Court was careful to point out that 'to strike down completely all residential requirements would mean that the urban area of Union could completely dominate and control the Board, leaving the rural areas of the county without any representative to voice their peculiar and unique territorial interests'. In decreeing transitional relief the District Court was directed to avoid such a result, saying that while awaiting a new Act of the Legislature.

'The remedy adopted should be one that recognizes and gives proper effect to the population variances among the districts as well as the economic differences among them. It was the virtue of the Dusch plan that it did exactly that.'

No figures are cited in the opinion as to Edgefield County, but it is said that 'There is not the wide disparity of population among the districts as was the case in Union County. It is not obvious that a minority, either in numbers or interritorial or economic interest, can dominate the Board'.

The Edgefield case is the one on which the Supreme Court denied certiorari.

I think the Fourth Circuit opinion in Lytle and Lybrand contains a valuable in Dallas County. Assuming that in Dallas County, Assuming that this case is, as ordered, to go back to the District Court to 'fashion relief', I suggest that it could very advantageously look carefully to the teachings of Lytle and Lybrand.

Since the Dallas County system reveals no such glaring disparities as appeared in Union County, and not now knowing the figures as to Edgefield, I adhere to my original dissent.

BELL, Circuit Judge (concurring in part and dissenting in part):

Although I agree with the majority that this cause must be remanded, I would remand for an evidentiary hearing on the question of invidious discrimination vel non which, as I understand the law, is the since qua non for finding a violation of the Equal Protection Clause in apportionment cases. See e.g., Dusch v. Davis, 1967, 387 U.S. 112, 116, 87 S. Ct. 1554, 18 L. Ed. 2d 656; Reynolds v. Sims, 1964, 377 U.S. 533, 561, 84 S. Ct. 1362, 12 L. Ed. 2d 506. 'The Equal Protection Clause does not, of course, require that the State never distinguish between citizens, but only that the distinctions that are made not be arbitrary or invidious.' Avery v. Midland County, 1968, 390 U.S. 474, at 484, 88 S. Ct. 1114, at 1120, 20 L. Ed. 2d 45.

This appeal is from a summary judgment upholding the districting plan for electing county commissioners in Dallas County. The implicit holding of the district court was that as a matter of law, there was no invidious discrimination in the plan. The majority of this court has found invidious discrimination as a matter of law. Because I find a fact issue for resolution in the district court, I respectfully dissent from the final determination in this court.

The invidious discrimination, if any, is against urban voters as a group-- the residents of the City of Selma. There is no claim of racial discrimination. It is simply an urban versus rural resident question somewhat like that faced by the Supreme Court in Dusch v. Davis, supra. The districting plan there was upheld on facts which fell short of showing invidious discrimination. The ratio decidendi of the decision is that, given an at-large election, coupled with a district residential requirement for the candidates where the population of the districts vary widely, the constitutional test under the Equal protection Clause nevertheless is whether there is invidious discrimination as a result of the inequality in population of the districts.

A substantial disparity from the one man-one vote standard makes out invidious discrimination in the ordinary apportionment plan but we are here faced with an at-large plan where all voters of the county have a voice in the election of the candidates. The dilution, if any, in the vote of some residents depends on finding a group of voters who are treated differently from other groups. Having established the existence of such a group, here the voters of Selma, the next question is whether the difference, or the dilution in the terminology of the majority, reaches proportions of invidiousness.

As noted, Avery v. Midland County, supra, equates an invidious distinction between voters with arbitrariness, and appellants alleged that the Dallas County districting plan was arbitrary.

The record as now constituted shows a population disparity in the districting plan in question on an urban versus rural basis. Just under one-half of the residents in the county reside in the district comprising the City of Selma and they are allocated one commissioner for residence. There are three districts outside the city and each is allocated one commissioner for residence. Thus approximately one-half of the population residing outside Selma has three resident commissioners as against one for the other one-half residing in Selma. All four commissioners are elected at-large. But Dusch v. Davis, supra, teaches that disparity in the population of districts may be justified in an at-large election system.

It may be that taxes are not levied on citizens of Selma for expenditure in the rural areas to such a degree as would rise to the level of an invidious form of discrimination. It may be that the system can be justified in other ways. Much will depend on the structure and operation of the county government from the standpoint of taxes levied, expenditures made, and services rendered.

The majority opinion assumes invidious discrimination in the following statement:

'. . . It is to be expected that commissioners who are elected from rural subdistricts will give greater priority to needs of the subdistricts in which they reside than to the interests of the City of Selma. Unlike the residents of Selma, the other voters in Dallas County can choose their proportional share of the representative body from among rural candidates who, by reason of their residence on rural areas, can be expected to share their interests.' P. 883.

This is disputed by the dissenting opinion; indeed, a strong argument is made that there is no discrimination at all against the citizens of Selma. I have examined the record in vain for a factual basis for either position.

My understanding of the judicial process in that constitutional rulings should not be anticipated. Ashwander v. Tennessee Valley Authority, 1936, 297 U.S. 288, 346-347, 56 S. Ct. 466, 80 L. Ed. 688, J. Brandeis, concurring. Indeed, Art. III of the Constitution prevents our deciding questions in the abstract. We are not foreclosed from requiring a factual determination because of the district court's view that only a question of law was presented. That view may well have been proper in ruling for the defendants on the authority of Davis v. Thomas County, 5 Cir., 1967, 380 F.2d 93, now to be overruled by the majority. A decision by the district court for plaintiffs might well have required an evidentiary hearing but it is enough to say that the case was not considered by the district court in that posture.

The sum of the majority opinion is to rule for plaintiffs on the one man-one vote standard despite Dusch v. Davis, supra, which holds that an at-large election with a district residential requirement superimposed thereon may accord with that standard even with substantial disparity in the population of the districts. The determination turns on the totality of the facts and not on a mechanical or arithmetical computation to determine a variation from the one man-one vote measurement.

I am not willing to decide the case for plaintiffs or defendants absent a resolution of the disputed facts which bear on the question of invidiousness. The pretrial order recites at least one such dispute under the heading 'DISPUTED FACTS', as follows:

'In their complaint, the plaintiffs allege: 'That three (3) out of every four dollars collected by Dallas County in the last fiscal year were levied on and collected from the citizens of Selma, Alabama. Notwithstanding, the services offered by the County to the citizens of Selma are comparatively insubstantial and grossly discriminatory.' This fact is disputed by the defendants.'

I would hold that the disparity appearing from the districting plan itself does make out a prima facie case of invidious discrimination. Aught else appearing, the Selma voters are entitled to have two of the four commissioners reside in Selma. The burden should be on defendants to overcome the prima facie case, if they can, by offering sufficient justification for the disparity. Within this framework, I suggest that the proper disposition of this cause is to vacate and remand with direction that it go forward to an evidentiary hearing.