This case involves the narrow question of whether an at-large election system in Lufkin, Texas, population 23,049, 72% white and 28% black, works an unconstitutional dilution of votes for the City Commission election. The district court held the at-large plan unconstitutional and ordered an election with one at-large and six single member districts. Because the district court’s factual findings are not adequate to support the finding of unconstitutionality, we vacate and remand.
Plaintiffs are black residents and registered voters in the City of Lufkin, Texas. They brought this suit in 1973 alleging violation of the Fourteenth and Fifteenth Amendments, as well as their civil rights under 42 U.S.C.A. § 1983. The defendants are members of the City Commission, the governing body of Lufkin.
Lufkin, in Angelina County, is a small town situated in eastern Texas near the Louisiana border. According to the 1970 census, Lufkin had a population of 23,049, of whom 6,604 were black, or about 28%. The city is governed by a City Commission of seven members, one of whom acts as Mayor. All seven commissioners are elected at-large, however, four of the commissioners are required by law to reside in designated wards. There are no primaries. The elections are nonpartisan. In order to be elected, a candidate must carry a majority of those residents voting. This plan of government has been in effect, without substantial change, since 1919.
About 85% of the city’s black residents live in the northern part of the city. The City Commission has never had a black commissioner, although a local black community leader, Inez Tims, did become the first black candidate for the City Commission in 1971. Tims received a plurality but not a majority of the votes cast in his election race, and was forced into a runoff, which he lost. His opponent was white.
The district court concluded that the system of municipal elections used by Lufkin operated “to minimize and dilute the voting strength of black citizens, and to prevent them from participating equally with white citizens in the political process relative to selection of the City Commission.”
The consideration of a dilution claim requires a clear understanding of the plaintiffs’ precise argument. Dilution is an elusive concept because, unlike a one person-one vote attack on an election scheme, it cannot be proven by mathematics alone. Conceptually, protesting dilution conflicts with the political theory of majority control. It reflects a challenge to the usual election system where the candidate of the majority of the voters wins an election. But the courts have held that when a majority of voters, and the successful governing authority elected by that majority, simply ignore the governmental needs of a substantial minority of the voters and remains arrogantly unresponsive to the voting strength of that minority, there has been a dilution of those votes that violates the United States Constitution. Amendment Fifteen to the Constitution provides: “The right of citizens of the United States to vote shall not be denied or abridged by the United States or by any State on account of race, color, or previous condition of servitude.” The Courts have interpreted this amendment to require that the protected persons be given some meaningful participation in the political process, not just the right to cast a vote that can be completely ignored in the provision of governmental protection and governmental services merely because an election system is so operated as to make that vote meaningless in the election outcome.
Although generally found in multimember or at-large voting systems where the same voting majority controls the balance of governmental power, the evil is not nec*926essarily cured by single member districts. And with the cure uncertain, the Courts must be careful to upset the legislative plan adopted by the people only when the Constitution clearly dictates that such plan is unlawful.
With this sketchy introduction to the problem, a brief review of the decisions is necessary to determine the present state of the law against which this ease must be assessed.
Historically, the initial constitutional concern with state and local • election schemes was whether the system, by apportionment of the elected representatives among districts with unequal numbers of voters, made one person’s vote unequal to the other as to the final matters upon which the elected representatives, casting equal ballots among themselves, voted. Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). A totally at-large system of voting eliminates any defect in the system viewed as a one person-one vote matter.
If each person’s vote, assuming 100% turnout, has perfect mathematical equality, however, the second constitutional concern is whether the system minimizes or cancels out the voting strength of a segment of the voting population in actual practice. Burns v. Richardson, 384 U.S. 73, 88, 86 S.Ct. 1286, 16 L.Ed.2d 376 (1966); Fortson v. Dorsey, 379 U.S. 433, 439, 85 S.Ct. 498, 13 L.Ed.2d 401 (1965). This infirmity is called “dilution.” This defect in the system gets away from pure mathematics, and concerns the sociological realities of the election process. See Reese v. Dallas County, 505 F.2d 879, 882 (5th Cir. 1974) (en banc), rev’d, 421 U.S. 477, 95 S.Ct. 1706, 44 L.Ed.2d 312 (1975).
The law has become clear that mathematical disparity renders a system unconstitutional. Chapman v. Meier, 420 U.S. 1, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975); Swann v. Adams, 385 U.S. 440, 87 S.Ct. 569, 17 L.Ed.2d 501 (1967). This defect in the system can be demonstrated by figures and the only question is how much variance from an absolute equality will be permitted. Compare Kilgarlin v. Hill, 386 U.S. 120, 87 S.Ct. 820, 17 L.Ed.2d 771 (1967), with Gaffney v. Cummings, 412 U.S. 735, 93 S.Ct. 2321, 37 L.Ed.2d 298 (1973). Needless to say, an all at-large system, where every voter gets to vote for every representative who is making the decisions in government provides mathematical perfection.
But, as illustrated above, mathematical perfection is not the sole criterion to be used in deciding whether an election system is constitutional. If a class of voters can prove that their voice in the government, their access to the political system, is diluted because of facts at work not reflected by the ¡figures, they are entitled to a remedy. White v. Regester, 412 U.S. 755, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973). So far, this Court has relied on Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir. 1973) (en banc), aff’d per curiam on other grounds sub nom., East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976), as setting forth the factors which must be addressed in considering the constitutional infirmity based on dilution.
We should not, however, confuse the standard by which to judge a legislatively-enacted scheme of election for dilution with the standard by which to judge a court-ordered scheme which is to be substituted for a constitutionally infirm system.
In the court-ordered remedy, single member districts are almost mandated by the United States Supreme Court. East Carroll Parish School Board v. Marshall, 424 U.S. 636, 639, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976); Chapman v. Meier, 420 U.S. 1, 18, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975); Mahan v. Howell, 410 U.S. 315, 333, 93 S.Ct. 979, 35 L.Ed.2d 320 (1973); Connor v. Williams, 404 U.S. 549, 551, 92 S.Ct. 656, 30 L.Ed.2d 704 (1972); Connor v. Johnson, 402 U.S. 690, 692, 91 S.Ct. 1760, 29 L.Ed.2d 268 (1971). Even though a legislative preference is shown for a court-imposed at-large plan which would not be unconstitutional under either the “one person-one vote” standard, or the “dilution” standard, the federal court is not permitted to favor such an at-large legislative plan submitted as a cure to a *927constitutionally infirm system merely because it expresses the will of the elected representatives of the people, but must follow a preferred single member district remedy. This strong preference apparently derives from the supervisory power of the Supreme Court over the remedies to be given for a constitutional infirmity. Chapman v. Meier, 420 U.S. 1, 18, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975). By invoking this preference in East Carroll Parish School Board, supra, the Supreme Court has indicated that no attention should be given to the size of the electorate in a proposed court-imposed at-large system. In that case, the at-large district would have had a population of only 12,884. Yet single member districts were required notwithstanding the statement in Connor v. Johnson, 402 U.S. 690, 91 S.Ct. 1760, 29 L.Ed.2d 268 (1971) that “single-member districts are preferable to large multi-member districts”. 402 U.S. at 692, 91 S.Ct. at 1762 [emphasis supplied].
When it comes to the constitutionality of the at-large system, however, the size of the electorate would seem to be important. The number of people involved must logically be considered. For a voter to know the candidates, for the candidates to get the attention of the voter, the commitment of time, money, and other resources to an election process must vary widely with the size of the electorate.
This Court has held, in cases which involve a fairly small electorate, that the plaintiffs must prove the precise way in which the system dilutes their voting strength. E. g., Nevett v. Sides, 533 F.2d 1361 (5th Cir. 1976).
The Court has clearly stated that no one group is constitutionally entitled to elect representatives from that group. Kirksey v. Board of Supervisors of Hinds Co., 528 F.2d 536, 541 (5th Cir. 1976), pet. for reh. en banc granted, 528 F.2d 543 (argued Sept. 9, 1976); Turner v. McKeithen, 490 F.2d 191, 197 n.24 (5th Cir. 1973); Howard v. Adams County Board of Supervisors, 453 F.2d 455, 458 (5th Cir.), cert. denied, 407 U.S. 925, 92 S.Ct. 2461, 32 L.Ed.2d 812 (1972). On the other hand, a majority bloc is not entitled to continually elect representatives who simply ignore the governmental interests of a substantial portion of the population.
Consequently, rather than providing us with a specific definition of dilution, the courts have stated that dilution occurs when the minority voters have no real opportunity to participate in the political process. In addition to the size of the district, the courts have used various other indicators to determine if dilution is present in a particular situation. In White v. Regester, 412 U.S. 755, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973), the indicators were: a history of official racial discrimination visa-vis voting and registering, a rule requiring a majority vote as a prerequisite to nomination in a primary, its so-called “place” rule which limits candidates for legislative offices from multimember districts to a “place” on the ballot which reduces the election to a head-on contest for each position, the fact that only two blacks had been elected since Reconstruction, a finding that there was no good faith concern for political and other needs and aspirations of the minorities, a white-dominated candidate slate in one county, racially discriminatory campaign tactics to destroy minority competition, cultural and language barriers to diminish Mexican-American abilities to participate in the political process, past use of the poll tax and the most restrictive voting registration procedures in the nation. In Whitcomb v. Chavis, 403 U.S. 124, 91 S.Ct. 1858, 29 L.Ed.2d 363 (1971), the indicators were: a history of discrimination in the attempts of minorities to register, general history of discrimination in voting against minorities, and a history of discrimination in minority attempts to participate in the political process. In Turner v. McKeithen, 490 F.2d 191 (5th Cir. 1973), the indicators were: a history of discrimination and the lingering effects thereof, lack of access to the political process, blatant opposition to minority participation in the candidate selection process, lack of responsiveness of the elected officials, and the strength of the state interest in multimember districts or at-large voting.
*928These indicators and structural devices denoting dilution are echoed in Perry v. City of Opelousas, 515 F.2d 639 (5th Cir. 1975); Wallace v. House, 515 F.2d 619 (5th Cir. 1975), vacated on other grounds, 425 U.S. 947, 96 S.Ct. 1721, 48 L.Ed.2d 191 (1976); Bradas v. Rapides Parish Police Jury, 508 F.2d 1109 (5th Cir. 1975); and Zimmer v. McKeithen, 467 F.2d 1381 (5th Cir. 1972).
The Court is trying to find its way in this developing area of the law. To enfranchise the diluted minority, care must be taken not to disenfranchise the majority. Any action with this result merely replaces one evil with another. To replace an at-large system with several single member districts invites variance from the perfect one person-one vote goal, and forever compartmentalizes the electorate, reinforces the bloc voting syndrome, and prevents members of a minority class from ever exercising influence on the political system beyond the bounds of their single member districts. They remain forever a minority in their representative influence. In the final analysis, therefore, the determination of dilution vel non must either rise or fall on the facts of the individual cases. See Dallas County v. Reese, 421 U.S. 477, 480-481, 95 S.Ct. 1706, 44 L.Ed.2d 312 (1975); Turner v. McKeithen, 490 F.2d 191 (5th Cir. 1973).
In making that determination, this Circuit has followed the articulation of factors in Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir. 1973) (en banc), aff’d per curiam on other grounds sub nom., East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976). The Court set out the relevant test as follows:
[W]here a minority can demonstrate [1] lack of access to the process of slating candidates, [2] the unresponsiveness of legislators to their particularized interests, [3] a tenuous state policy underlying the preference for multi-member or at-large districting, or [4] that the existence of past discrimination in general precludes the effective participation in the election system, a strong case [of dilution] is made. Such proof is enhanced by [a] showing of the existence of large districts, [b] majority vote requirements, [c] anti-single shot voting provisions and [d] the lack of provision for at-large candidates running from particular geographic subdistricts. The fact of dilution is established upon proof of the existence of an aggregate of these factors . however, . . .all these factors need not be proved in order to obtain relief. [Bracketed letters and numbers supplied].
Id. at 1305. It is clear that these factors are aimed at attempting to distinguish between the invidious and the innocuous at-large schemes. If black voters in an area with no past discrimination are free to slate candidates, and are having their particular interests well attended to by their representatives then it is unlikely that the adoption or retention of the at-large system dilutes their Fifteenth Amendment right to vote. Contrariwise, if the district is extremely large, and a majority is required to elect, and black citizens are excluded from the slating process, there is far more reason for a federal court to be skeptical about validity of the at-large scheme.
The case before us today is very much like Nevett v. Sides, 533 F.2d 1361 (5th Cir. 1976), where this Court reversed a finding of dilution and remanded for further findings in accord with the Zimmer tests. In Nevett the district court had sought to reduce the multi-faceted and admittedly imprecise set of four primary and four secondary factors to a simple question, to-wit, “does the present system, regardless of purpose, operate to minimize or cancel the voting strength of the blacks in the City . .” On appeal Judge Rives pointed out for us, “we cannot affirm the ultimate conclusion of a dilution without findings of fact to fit proper standards. To hold merely that the plan unintentionally ‘simply does act to inhibit and has inhibited voting strength’ ... is not enough.” 533 F.2d at 1365.
The conclusion of the district court in the instant case is phrased in much the same *929language as that used by the trial judge, and found inadequate in Nevett. We note that eonclusory findings as to each of the Zimmer criteria are no more helpful than an overall eonclusory finding of dilution. The factual predicates for such conclusions must be clearly stated by the trial court.
The first factor which must be addressed is minority access to the slating process. The district court made no findings as to the existence of any organization which was involved in the slating process or the method of slating, if any, which forecloses minority participation. It did discuss the candidacy of Inez Tims, a black, for City Commissioner. The presence of Mr. Tims on the ballot is suggestive of the fact that there is minority access to the nomination process. There is no indication that any other blacks have since sought election. The lack of any findings or discussion of this point may indicate that the district court failed to consider this factor, which it was obliged to weigh in coming to its ultimate conclusion.
Any consideration of access to the election process must necessarily concern itself with the size of the electorate. The time, money, and number of persons needed for a campaign in a small electorate, and the ability of the voters to know the various candidates with a minimum of effort, differs so much from districts of large population and extended geographical area that cases decided under one set of circumstances may be questionable guides to decisions under the other. In a small electorate, the plaintiffs must show facts that overcome what would seem to be apparent — that candidates with only modest support could wage an effective campaign in which the merits of their candidacy could be amply exposed to the voters.
The second factor concerns legislative responsiveness to the particularized needs of the minority group. There are two somewhat distinct facets to this issue. The first concerns the provisions of municipal services to neighborhoods populated by minority group members. On this subject the district court found that “more than any other area within the city, [the black area] is characterized by dilapidated housing, poor enforcement of the city building codes, and despite recent efforts by the city to upgrade the streets in the area, inferior streets.” The court made no findings as to which of these services or conditions are the responsibility of the City Commission however. More importantly, these findings are incomplete. They neither accept nor reject the undisputed documentary evidence offered by the defendants in the form of a comprehensive master plan for the city. That evidence purported to show that the black section of the city had better drainage, fire protection, neighborhood school coverage and neighborhood recreational facilities than the most affluent white section of Lufkin. It also concludes that the black section possessed equal qualities of police protection, library and medical facilities, although health services in the black area were slightly inferior to the white section. Further, five of the city’s ten parks were located in the black residential area, and in the last five years the city had expended 44% of its funds for maintenance of streets, new streets and repairs in the black section. This percentage is greatly disproportionate to the comparative size of the black area to the rest of the city. Clearly, findings on all these issues are required before this Court can fully address and resolve the question of responsiveness.
The other facet of the responsiveness inquiry concerns the distribution of municipal jobs and appointments to various boards and commissions. On this subject the district court found that there are no black policemen or firemen and virtually no black city employees in Lufkin. There are no findings, however, as to whether any blacks applied for these positions, nor as to whether any black applicants were qualified for the positions they sought. Furthermore, there is no finding as to whether or not present employees are long-term employees with a low turnover, creating fewer opportunities for black employment. Finally, there are no findings as to how much control, if any, is exercised by the City Commission in making these hiring decisions. There is some question as to whether hiring *930disparities reflect a lack of governmental responsiveness at all if the all white city employees see to it that the black community receives equal city services. There are after all other remedies for discrimination in hiring by the city. See 42 U.S.C.A. § 2000e-2(a). To the extent that the hiring patterns of the city can reflect its lack of responsiveness to the particularized needs of the black community, however, findings on each of the issues isolated above are essential.
The third factor which Zimmer identifies is the strength of the state policy favoring at-large elections. Here, there are no findings on this precise question. There is an indication that the Lufkin system has been in use for almost 60 years, and that for Texas cities under 5000 population at-large elections are required. See Vernon’s Tex. Rev.Civ.Stat.Ann. art. 977. As was pointed out in McGill v. Gadsden Co. Commission, 535 F.2d 277 (5th Cir. 1976), where a 75-year-old at-large scheme was considered, “[t]his policy could not have had racist underpinnings because other, less subtle state mechanisms had already disenfranchised almost all black voters by the turn of the century.” Since Zimmer teaches that this factor must be considered along with the others, explicit findings on this issue are required.
The fourth major factor which is relevant on the dilution question concerns whether the past effects of discrimination have the effect of precluding effective participation in the election system by blacks today. On this subject, the district court found that there had been past discrimination in Lufkin. It also found that there had been no blacks elected to any city or county office. However there were no findings that, because of the past discrimination, present black participation in elections is less than effective. There were no findings that blacks were afraid to vote, or campaign, and in fact there was evidence that the turnout among black voters was unusually strong. As stated in McGill, supra, 535 F.2d at 281:
We reaffirm the importance of past discrimination to decisions about the dilutive effects of at-large voting schemes. The key question, however, is whether ‘the existence of past discrimination in general precludes the effective participation [of blacks] in the election system’ today, [citation omitted].
Without more explicit and concrete findings the general conclusion that past discrimination was presently precluding a current effective participation is unsupported.
On the four secondary factors articulated in Zimmer the record reflects that there is a geographic subdistrict requirement, which cuts in the city’s favor, and that there is a majority vote requirement, which cuts in the plaintiff’s favor. There are no findings on whether or not the 23,000 person single district in Lufkin was “large” or “small” as compared with other cases which have considered this point in reference to other cities, and there is no finding as to single-shot voting, although the system which assigns each candidate to a place on the ballot and pits him against a single opponent presumably precludes the use of single-shot voting.
Thus the findings of the district court are not adequate to support its decision that the at-large election plan is unconstitutional. To the extent that there are findings on the crucial factors, they are inconsistent and point in opposing directions. A full weighing and balancing of all of them is called for so that a decision based on the aggregate of the factors may be made. As this Court has previously noted, “[t]he weighing of these factors is ordinarily a trial court function which we will not undertake initially unless the record is so clear as to permit of only one resolution.” Paige v. Gray, 538 F.2d 1108, 1111 (5th Cir. 1976).
The point made by Judge Rives in Nevett v. Sides, 533 F.2d 1361 (5th Cir. 1976), should be emphasized here.
Specifically, the trial court’s findings may be read as indicating that elections must be somehow so arranged — at any rate where there is evidence of racial bloc voting — that black voters elect at least *931some candidates of their choice regardless of their percentage turnout. This is not what the constitution requires.
533 F.2d at 1365.
It thus appears to be the law at this time, from the decisions of this Court and the Supreme Court, that at-large systems in small communities are not unconstitutional merely because a minority of voters cannot elect a candidate from among themselves.
We remand this case to the district court for reconsideration in the light of this opinion and the other decided cases. Since there are many relevant factors that may not have been appreciated by the parties at the trial of this case, on remand the district court shall conduct such further evidentiary hearings as it may deem necessary to allow it to make the factual findings required for appropriate disposition of this case in the light of this opinion.
VACATED AND REMANDED.