Appellants, black residents of Woodville, Mississippi, brought this action in 1985 against the City of Woodville alleging that the city's aldermanic election system diluted the voting strength of Woodville's black residents in violation of Section 2 et seq. of the Voting Rights Act of 1965, as amended, 42 U.S.C. Sec. 1973 et seq. (1982). Appellants would have the court superintend the division of Woodville's single electoral district into four single-member districts. The district court originally granted summary judgment to appellees. Monroe v. City of Woodville, Mississippi, 636 F. Supp. 423 (S.D.Miss.1986). We reversed and remanded for a trial on the merits. Monroe v. City of Woodville, Mississippi, 819 F.2d 507 (5th Cir.1987). After trial, the district court concluded that Woodville's at-large election system did not violate Section 2. Having considered appellants' broad attacks on this result, we nevertheless affirm the judgment on somewhat different reasoning than the district court employed.

Woodville, Mississippi, is exactly one mile square and claims 1,512 inhabitants. Black residents in Woodville account for 64.3% of the total population and 60.5% of the voting age population. According to the district court, Woodville is highly socially segregated and lies in one of the most rural, impoverished and economically depressed areas of Mississippi.

Woodville's mayor and four aldermen are elected in at-large contests. When this suit was filed, Mississippi prohibited "bullet," or "single-shot," voting in municipal elections.1 Further, until 1987, Woodville's residents were required to register both with the city and county in order to vote in their respective elections. The town has never included more than one voting precinct.

Although blacks constitute 60.5% of Woodville's voting age population and at least two black candidates have run for aldermanic seats in each city election since 1965, when this case was filed in 1985, only one black candidate had been elected to that post.2 During the same twenty year period, four black candidates have unsuccessfully run for mayor, and three blacks failed in bids for election as the town marshall, another at-large post. None of the seven black residents who have run for a position on Woodville's democratic executive committee since 1965 has been successful.

Congress, in amending Section 2, affirmed its commitment to an "effects" test of voting rights discrimination based on the "totality of the circumstances." See S.Rep. No. 417, 97th Cong., 2d Sess., reprinted in 1982 U.S.Code Cong. and Admin.News 177 at 192-93; Thornburg v. Gingles, 478 U.S. 30, 35, 106 S. Ct. 2752, 2758, 92 L. Ed. 2d 25, 37 (1986); Brewer v. Ham, 876 F.2d 448, 450 (5th Cir.1989). At the same time, however, Section 2 explicitly disallowed resort to a rule of proportional representation by race or ethnicity. Our judicial resources strain to walk the statutory tightrope in this case, because Woodville's black plaintiffs contend that they were denied effective political representation notwithstanding their decisive numerical voting majority. Confronted by a similar paradox, a previous panel of our court asked whether

supposing that the proposed new system still fails to produce a black winner, we will then be asked to continue down the slippery slope, mandating new designs which segregate blacks into greater and greater concentrations until at last a black is elected? Somewhere along this downward course, the goal of an open and pluralistic political process, where groups bargain among themselves, is transformed into one of proportional representation by persons beholden for office to discrete ethnic groups.

Houston v. Haley, 859 F.2d 341, 342-43 (5th Cir.1988), vacated, 869 F.2d 807 (5th Cir.1989).

Nevertheless, when this case appeared before us after summary judgment, we rejected the city's claim that even if a violation of the Voting Rights Act had occurred, Section 2 offers no possible remedy because of the blacks' significant majority at the poll. The case accordingly went to trial, and we review it now on a full record.

Congress instructed courts adjudicating Section 2 claims to conduct a "searching and practical evaluation" of "past and present reality" to determine whether participation in the political process is "equally open" to all persons. See Thornburg, 106 S. Ct. at 2763-64 (quoting from S.Rep. No. 417, supra ). As guidance for this broad inquiry, Congress directed attention back to our Circuit's opinion in Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir.1973) (en banc), aff'd, sub nom. East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S. Ct. 1083, 47 L. Ed. 2d 296 (1976) (per curiam), which identified a number of factors generally relevant to a fact-bound, intensely local appraisal of the challenged electoral system. Congress did not intend these factors to be exclusive. Thornburg, 106 S. Ct. at 2764.

The Supreme Court, in Thornburg v. Gingles, acknowledged the totality of the circumstances approach and introduced an additional, threshold analysis to be used in challenges to at-large election systems.3 As the Court stated, unless the threshold factors are established by the Section 2 plaintiff, "the use of multimember districts generally will not impede the ability of minority voters to elect representatives of their choice." 106 S. Ct. at 2766. Satisfying the threshold test, therefore, does not prove a plaintiff's Section 2 claim; the district court must then proceed to the totality of the circumstances inquiry.4 Cf. Overton v. City of Austin, 871 F.2d 529 (5th Cir.1989) (plaintiff may lose at the threshold step pretermitting need to conduct totality of circumstances test). In this case, the district court made findings both on the threshold factors and on the totality of the circumstances inquiry.

We conclude that the district court was not clearly erroneous in its determination that Woodville's electoral structure does not violate Section 2. In so doing, however, we must clarify the district court's analysis of the Thornburg threshold factors, which is vigorously attacked by appellants. We affirm the district court's findings on the totality of circumstances test for a Section 2 violation.

Appellants contend that the trial court's Thornburg analysis erred in its findings on the lack of political cohesion among the black voters and the legal significance of the white bloc vote. The first Thornburg factor, that the "minority" is sufficiently large and geographically compact to form a majority in a single-member district, concededly exists here. We review the district court's findings on the other two Thornburg factors by the standard of clear error. Overton, 871 F.2d at 533.

At the outset, we note that the district judge discounted the statistical evidence presented by the appellants as severely flawed.5 The weaknesses he observed are particularly damaging to the appellants' case because this information constituted the bulk of their evidence on the issues of black political cohesiveness and white bloc voting. Dr. Love, appellants' statistical expert, faced difficulties in producing useful data for the court. Since Woodville uses only one precinct in its elections, statistical methods such as ecological regression and "overlapping percentages," which require multiple precincts, cannot be used. Therefore Dr. Love examined 1983 and 1987 elections in Wilkinson County, which includes Woodville. We have previously cautioned district courts to undertake "fact-specific assessments" of the relevance and probative value of exogenous election results. Westwego Citizens for Better Govern. v. Westwego, 872 F.2d 1201, 1208 n. 8 (5th Cir.1989). Dr. Love's analysis of the 1987 county elections was further hampered by results that were not statistically significant because of large amounts of crossover voting. Additionally, as the district court noted, appellees' expert identified specific flaws in Dr. Love's analysis: failure to provide a test of statistical significance;6 failure to analyze certain relevant election contests; and mixing of data from primary and general elections.

Based on the evidence before it, the district court concluded that the appellants did not prove political cohesion among the black voters in Woodville. That the appellants had the burden of proof on this issue is not disputed. See Overton, 871 F.2d 529, 543 (Jones, J., concurring). Appellants assert that this finding embodies an error of law as well as fact. They note that the district court accepted the parties' stipulation that both whites and blacks in Woodville vote along racial lines. According to appellants, Thornburg v. Gingles, Campos v. City of Baytown, 840 F.2d 1240 (5th Cir.1988), and Citizens for a Better Gretna v. City of Gretna, 834 F.2d 496 (5th Cir.1987), hold that if the district court finds that a minority group votes as a bloc for minority candidates, political cohesion within the minority group is proven. We disagree.

Thornburg does recognize that establishing "that a significant number of minority group members usually vote for the same candidates is one way of proving political cohesiveness...." 106 S. Ct. 2769-70 (emphasis added). Statistical proof of political cohesion is likely to be the most persuasive form of evidence, although other evidence may also establish this phenomenon. Brewer, 876 F.2d 448, 453 (lay testimony from members of the community on political cohesion might be sufficient). Nevertheless, courts must carefully examine statistical evidence of racial bloc voting to determine its relevance and probativeness to a finding of political cohesiveness. Accord Overton v. City of Austin, 871 F.2d 529.

Appellants err by implying that a finding of racial polarization in voting behavior is synonymous with a group's political cohesion. The terms are quite distinct. That a group's voting behavior is racially polarized indicates that the group prefers candidates of a particular race. Political cohesion, on the other hand, implies that the group generally unites behind a single political "platform" of common goals and common means by which to achieve them.7 For example, the black population of a district may vote in a racially polarized manner so as to overwhelmingly favor black candidates, but the group may lack political cohesion if it splits its vote among several different black candidates for the same office. Where the black voters overwhelmingly favor a particular black candidate to the exclusion of others, data on racial bloc voting will be more probative to determining political cohesiveness.8

Political cohesion is required under Thornburg for a fundamental reason. As the Court said, "if the minority group is not politically cohesive, it cannot be said that the selection of a multimember electoral structure thwarts distinctive minority interests." Thornburg 106 S. Ct. at 2767. Put otherwise, if the members of the minority community do not agree among themselves on basic political issues in their subdivision, their interests as a minority are not subverted by an at-large electoral system.

We turn, then, to the district court's finding that Woodville's blacks are not politically cohesive. The ambit of appellate review of this fact is limited. A finding is clearly erroneous when "although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed." Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948)).

If the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.... This is so even when the district court's findings do not rest on credibility determinations, but are based instead on physical or documentary evidence or inferences from other facts.

Anderson, 470 U.S. 564, 573-74, 105 S. Ct. 1504, 1511.

In this case both parties can point to evidence in the record to support their positions. The 1983 elections show a strong degree of racial bloc voting by black residents, which could indicate political cohesion. On the other hand, the more recent results from the 1987 county elections indicate a significant degree of crossover voting for white candidates by black citizens.9 Crossover voting by a minority group does not preclude a finding of political cohesiveness per se, but its presence is relevant. The 1987 data, therefore, weigh against finding political cohesiveness. Dr. Love opined that Woodville's black residents were politically cohesive; Dr. Van Gelder, appellees' expert, disagreed. Testimony from black residents of Woodville also differed on this point.10 Finally, we note that the inability of Woodville's black residents, who comprise over sixty percent of the city's voting age population, to put significant numbers of allegedly black preferred candidates in local office may further signal a lack of political cohesion.11 Therefore, from the record before us, we cannot say that the district court clearly erred in its finding.

The district court found that both white and black residents in Woodville engage in racial bloc voting. The determinative question for a Section 2 claim, however, is not whether whites generally vote as a bloc, but rather, whether such bloc voting is legally significant. In this case, the district court found that the white bloc vote was not legally significant.

Thornburg instructs that "Because ... the extent of bloc voting necessary to demonstrate that a minority's ability to elect its preferred representatives is impaired varies according to several factual circumstances, the degree of bloc voting which constitutes the threshold of legal significance will vary from district to district." Thornburg, 106 S. Ct. at 2769. The Court then identified some factors to consider in conducting this evaluation. Among those are the nature of the allegedly dilutive electoral mechanism, the percentage of registered voters in the district who are members of the minority group, and the size of the district. 106 S. Ct. at 2770. Based on such facts, Thornburg holds that white bloc voting will be legally significant where "the white majority votes sufficiently as a bloc to enable it--in the absence of special circumstances ... usually to defeat the minority's preferred candidate." Thornburg, 106 S. Ct. at 2767.

The district court evidently found it difficult, as do we, to analyze "legally significant" white bloc voting in a jurisdiction where blacks are the numerical majority. The court held:

In the present case, the evidence shows that Blacks constitute the majority of the over-all population, of the voting age population, and of the number of registered voters; the evidence does not support the Plaintiffs' arguments that the white bloc vote usually defeats the minority's preferred candidates. There is no "submergence of the minority" since minority citizens (Blacks) are actually in the majority in the City of Woodville. White candidates win not because of the white bloc vote, but because of the crossover vote. The Plaintiffs may have established a "white bloc vote", but they have not proved legally significant racial bloc voting for purposes of a Section 2 claim. If white bloc voting to a legally significantly degree is not proven, minority voters have not established that the challenged electoral structure interferes with their ability to elect their preferred candidates.

Appellants assert that white bloc voting, rather than black crossovers, is responsible for the historical defeat of Woodville's black candidates. Consequently, they interpret the district court's opinion as essentially precluding Section 2 claims whenever a minority group constitutes a majority of the population. If the district court's opinion so held, it would be wrong. Unimpeachable authority from our circuit has rejected any per se rule that a racial minority that is a majority in a political subdivision cannot experience vote dilution. Zimmer v. McKeithen, 485 F.2d at 1303. Zimmer relied upon the Supreme Court's decision in White v. Regester, 412 U.S. 755, 93 S. Ct. 2332, 37 L. Ed. 2d 314 (1973), which affirmed a finding of racial vote dilution in Bexar County, Texas, even though Mexican-American plaintiffs constituted a numerical majority there. See Graves v. Barnes, 343 F. Supp. 704, 733 (W.D.Tex.1972) (three-judge panel), aff'd sub nom. White v. Regester, 412 U.S. 755, 93 S. Ct. 2332 (1973). The caveat should be added that in Zimmer, at least, the black majority had recently been freed from literacy tests and impediments to voting registration. As de jure restrictions on the right to vote mercifully recede further into the historical past, we should expect it to be increasingly difficult to assemble a Zimmer -type voting rights case against an at-large electoral district where a minority-majority population exists. Such a case is not, however, precluded as a matter of law.

A subtle error plagues the appellants' disagreement with the district court. The quoted portion of its opinion refers not to any general principle of illegal vote dilution but to the specific Thornburg threshold inquiry whether the white bloc vote is legally significant, i.e. whether it usually operates to defeat the black candidates. This narrower issue discussed by the district court requires a more focused attack than that levied by appellants. It seems possible to argue both (1) that a Sec. 2 vote dilution violation may occur even if a minority is more populous in a political jurisdiction and (2) that Thornburg's threshold criterion of legally significant white bloc voting does not deal with such a circumstance. Thornburg repeatedly describes the submergence of black voters by a white majority. See e.g. 106 S. Ct. at 2764, 2765, 2767. The terms "majority" and "minority", in context, refer not only to the relative number of blacks and whites in our general population but to their relative representation in the electoral district being challenged. Discussing its criterion of legally significant white bloc voting, the court explains,

In establishing this last circumstance, the minority group demonstrates that submergence in a white multimember district impedes its ability to elect its chosen representatives. 106 S. Ct. at 2767 (emphasis added).

In light of Thornburg's emphasis when enunciating its threshold standards, that a vote dilution Section 2 claim depends upon a black minority submerged within a white majority, we can readily appreciate the district court's conclusion that Woodville did not experience "legally significant" white bloc voting in part because of the black majority population. Whether this prong of Thornburg was intended to address the case before us is a matter of speculation among several possible interpretations. The issue is, however, ultimately irrelevant because irrespective of Thornburg's meaning in a case like this, Zimmer's holding clearly was not abandoned when Congress amended Section 2.

Because we have already concluded that a Thornburg vote dilution claim is foreclosed here by lack of black political cohesion, and we conclude in the following discussion that a Zimmer totality of circumstances dilution claim was not proven by appellants, we need not opine further on this puzzling aspect of Thornburg.

Responding to appellants' final point, we must review whether the district court was correct in its finding that Woodville's multimember districting scheme does not run afoul of the totality of the circumstances test required by Section 2. In the court's words, the appellants had failed to prove "that the political processes leading to nomination or election in the City of Woodville are not equally open to participation by black voters or that Blacks have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." 688 F. Supp. at 264. The court's misreading of the proper relationship between Thornburg and Section 2 (addressed in Part II of our opinion) is not fatal to its fact finding, because the court proceeded to address each of the Zimmer factors12 and made specific findings as to each. The district court then combined its analysis of the Zimmer factors and the three Thornburg factors to reach its ultimate finding under the "totality" test.

Once again, our review is limited to a determination of whether the district court's findings were clearly erroneous. The district court found that before 1965 discrimination prevented black citizens of Woodville from fully participating in the electoral process; that voting had been polarized along racial lines; that socio-economic disparities existed between Woodville's black and white residents; that few blacks had been elected in the City of Woodville; and that the black population in Woodville was sufficiently large and geographically compact to allow the creation of single-member districts with majority black voting age populations. The court balanced these findings against its other findings, namely: that there was no candidate slating process in Woodville; that overt or subtle racial appeals were absent in political campaigns within the City; that Woodville's elected officials had adequately responded to the needs of the black residents; that Woodville's small size resulted in a multimember district which was not unusually large so as to hinder the opportunity of blacks to elect representatives of their choice; that the anti-single-shot provision had been repealed; that black residents constituted a majority of Woodville's voting age population; that blacks crossed over to vote for white candidates to a large degree; that blacks served in many public offices in the surrounding Wilkinson County; and that the plaintiffs had failed to prove political cohesion among black residents or legally significant white bloc voting. The court also noted that Woodville's black community is very active politically.13 The role of the district court is to make a "searching and practical evaluation" of the election system being challenged. Because its comprehensive individual findings find support in the record, we cannot say that its ultimate finding under the totality of the circumstances test is clearly erroneous.

Thornburg's endorsement of a "functional view" of the political structure of the jurisdiction involved is particularly relevant to this case. Based solely on the observation that black candidates have been largely unsuccessful in past Woodville elections despite their constituting a voting age majority of the community, our prior panel was initially inclined to suspect that the town's electoral structure might violate Section 2. But reliance on a merely visceral response is inadequate and inappropriate in today's Voting Rights Act cases. Examination of the trial record increases our confidence in the district court's decision.

Since "the theoretical basis for this type of impairment is that where minority and majority voters consistently prefer different candidates, the majority, by virtue of its numerical superiority, will regularly defeat the choices of minority voters,"14 appellants, in challenging Woodville's electoral system, are at a distinct disadvantage. "Minority voters who contend that the multimember form of districting violates Sec. 2, must prove that the use of a multimember electoral structure operates to minimize or cancel out their ability to elect their preferred candidates." Thornburg, 106 S. Ct. at 2765. If there truly is a black preferred candidate, or set of candidates, the roughly twenty percent margin of numerical superiority in voting age population should provide the black residents with a sufficient opportunity for success at the polls. At least this court, along with our sister circuits, has so held in fashioning legal remedies.15 Thus, there is now no discernible structural impediment to black success at the polls.

The appellants raise two additional points of error in the district court's decision. First, appellants argue that the court's alternative holding is barred by the law of the case. The district court held alternatively that even if Woodville's electoral structure violated Section 2, no remedy would be proper or necessary because its black residents are the decisive voting-age majority. This statement comes close to flouting the contrary decision of our court in reversing the previous summary judgment awarded to Woodville. We need not address the law of the case doctrine, however, in view of our affirmance of the judgment on the merits of the case.

Second, appellants argue that the district court, on several occasions, ignored substantial evidence contrary to its decision. See Velasquez v. City of Abilene, Texas, 725 F.2d 1017, 1020 (5th Cir.1984). Specifically, appellants claim that the district court ignored evidence that city officials do not repair streets in black neighborhoods as compared to white neighborhoods and that officials at City Hall are not as responsive to the needs of blacks as to those of whites. Having carefully reviewed the record, we agree with the implicit finding of the district court that there was no substantial evidence of these charges at trial. The appellants also claim that the district court overlooked the minimal success of black candidates for public offices in Woodville. On the contrary, the district court noted in its opinion that only one black aldermanic candidate had been successful up to 1985. Having conducted its balancing test mindful of Woodville's unimpressive showing of black candidate success, the district court should not be required to elaborate further on this point.

For the foregoing reasons, the decision of the district court is AFFIRMED.