We address, for a third time, the claims of a group of citizens from the city of Westwego, Louisiana that the form of their city's government bars Westwego's black citizens from any meaningful degree of participation in the city's political life. On the basis of the evidence presented to and the facts found by the district court, we conclude that the plaintiffs have prevailed on their claims under the Voting Rights Act. Accordingly, we reverse the district court's dismissal of their claims, render judgment in their favor, and remand for the development and implementation of an electoral scheme which will remedy the vote dilution that attends the current at-large method of electing aldermen.

The procedural history of this case has been an unfortunately prolonged one. Although more fully set out in the two prior opinions of this Court, see Westwego Citizens for Better Government v. City of Westwego, 872 F.2d 1201 (5th Cir.1989) (Westwego I ) and Westwego Citizens for Better Government v. City of Westwego, 906 F.2d 1042 (5th Cir.1990) (Westwego II ), it is useful to summarize that history here. Plaintiffs1 initially filed this action seven years ago, on November 20, 1984, alleging that Westwego's system of electing aldermen on an at-large basis diluted the voting strength of Westwego's black citizens, resulting in the denial or abridgement of those citizens' rights to participate in the political process and to elect candidates of their choice in violation of section 2 et seq. of the Voting Rights Act, 42 U.S.C. § 1973 et seq. Plaintiffs seek 1) a declaratory judgment that Westwego's at-large election system unlawfully dilutes black voting strength, 2) an injunction prohibiting future at-large Westwego aldermanic elections, 3) the replacement of the current at-large system with a scheme of single-member districts, and 4) their court costs and attorneys' fees. At the conclusion of a two-day bench trial, the trial judge dismissed the plaintiffs' action for reasons stated orally from the bench. Plaintiffs timely appealed.

This Court remanded the case for entry of specific findings of fact and conclusions of law. We held that because the district court "failed to explain its reasoning with sufficient particularity, we must remand this case to the district court so that it may make the specific findings of fact and conclusions of law necessary to support a judgment." Westwego I, 872 F.2d at 1204.

Promptly after we issued our mandate, and without any consultation with the parties or any further proceedings, the district court entered Supplemental Findings of Fact and Conclusions of Law. Westwego II, 906 F.2d at 1043. In those supplemental findings and conclusions, the court held that the plaintiffs' case foundered for two reasons: 1) the plaintiffs had failed to demonstrate that blacks would constitute the majority of the voting age population of any single-member aldermanic district, and 2) the plaintiffs had failed to demonstrate that the white citizens of Westwego would vote as a bloc to defeat the candidates preferred by the city's black citizens. Id. In reaching this second conclusion, the district court stated that it was not persuaded that the plaintiffs' evidence of racially polarized voting among the Westwego electorate in exogenous2 elections was sufficient to establish that Westwego's white citizens would vote as a bloc in Westwego aldermanic elections. Id.

Within ten days of the district court's entry of its supplemental findings and conclusions, the plaintiffs moved to amend those findings and conclusions on the basis of evidence of developments in Westwego in the two years since the trial. In particular, plaintiffs wished to introduce evidence that in 1989, for the first time, a black man ran for alderman in Westwego and that the results of that (indigenous) election showed high racial polarization of the Westwego electorate. The plaintiffs also proffered evidence that under certain hypothetical single-member districting plans, a single-member aldermanic district could be created in which the black voting age population would exceed 50%, so that in those districts the black citizens of Westwego would possess the potential to elect candidates of their choice. The district court refused the plaintiffs' proffer and denied their motion. Again, the plaintiffs timely appealed.

This Court reluctantly held that it was necessary to remand the case for a second time. It was error for the district court to refuse to consider the evidence offered by the plaintiffs, and the district court still had not "explain[ed] specifically which evidence it finds credible and which it does not." Id. at 1044. Accordingly, we remanded the case to the district court "for further proceedings, including a hearing at which [the plaintiffs] will have the opportunity to adduce the evidence referred to in their motion." Id. at 1047. We specifically directed the district court to "enter supplemental findings of fact and conclusions of law on the two issues discussed and on any other issues raised by the parties and the evidence...." Id. This panel otherwise retained jurisdiction of this appeal.

On the second remand the district court held a hearing at which the plaintiffs presented their evidence and the defendants responded. Prior to the hearing both sides had submitted to the district court proposed findings of fact, and after both sides had presented their evidence, the district court announced from the bench which of the plaintiffs' and defendants' proposed findings it adopted and which it rejected. The district court did not draw any conclusions of law or enter any new judgment, believing that the limited scope of our remand did not confer jurisdiction on it to do anything but make additional findings of fact. The parties have now returned to this Court, each claiming to have prevailed.

Careful review of the entire record discloses that the district court found or the parties stipulated to the following facts.

The city of Westwego lies just across the Mississippi River from New Orleans. It has an area of one square mile, and is bisected by the West Bank Expressway, a wide highway that runs east-west. Westwego is divided by the Bureau of the Census into three tracts, numbered from south to north as 271, 272, and 273. Tract 271 contains south Westwego, below the West Bank Expressway. Tract 272 contains most of the city's population north of the Expressway, and tract 273 runs along the Mississippi River and contains some population in its western section. The central and eastern sections of Tract 273 are uninhabited. According to the 1990 census, the population of Westwego is 11,218 persons, of whom 81.6% are white and 15.6% are black. The total voting age population of the city of Westwego is 8,072 persons, of whom 85.1% are white and 12.4% are black.

Westwego is governed by a mayor and board of five aldermen. The board of aldermen exercises considerable authority: all permits, zoning changes, land use requests, and licenses to operate a business or sell alcoholic beverages must be approved or granted by the board. Louisiana law authorizes municipalities like Westwego to have from five to nine aldermen, elected at large, by districts, or by a system that combines the two. Westwego has always elected all of its aldermen at-large. There is no requirement that aldermanic candidates run for a specific seat, and there is no "anti-single-shot"3 provision, but Westwego does have a majority vote requirement.4

No black candidate has ever been elected to the board of aldermen or to any other municipal office in Westwego. Prior to the 1989 aldermanic primary election, no black had ever run for alderman in Westwego. In 1989, Glenn Green, a black resident of Westwego and former policeman, ran for a position on the board of aldermen. Of the sixteen candidates in the 1989 primary, Green finished twelfth. He received approximately 89% of the black vote and approximately 16% of the white vote. Green did not qualify for the runoff election and was not elected to office.

B. The Districting Plans Proposed by the Plaintiffs

At trial the plaintiffs proposed three plans for dividing Westwego into single-member aldermanic districts and proposed two more plans at the hearing held on the second remand of this case. These plans were identified as Plans A through E. Two are of interest here. Plan A was the first plan the plaintiffs proposed. It was originally conceived based on 1980 census data, and contains five single-member districts. Data from the 1990 census indicates that one of the districts in Plan A would have a total black population of 59.6% and a black voting age population of 54.0%. At the hearing held on the second remand of this case, the plaintiffs proposed Plan D. Plan D is a variation on Plan A, drawn to account for population changes between 1980 and 1990. It too contains five single-member districts, and according to 1990 census data one of its districts would have a total black population of 59.1% and a black voting age population of 52.8%.

The district court found that plaintiffs' districting Plan A was "an acceptable districting scheme": that is, the plan did not disregard Westwego's historical, political, and geographic boundaries and did not require any unreasonable or objectionable gerrymandering of the city. Moreover, the district court concluded that under any of the districting plans proposed by plaintiffs, including Plans A and D, a black candidate would have a better opportunity to be elected alderman than under the at-large scheme.

C. Racial Polarization of the Westwego Electorate

Both plaintiffs and defendants presented expert testimony on the voting patterns of the Westwego electorate. Dr. Douglas Rose, a professor of political science at Tulane University, testified for the plaintiffs, and Dr. John Wildgen, a professor of political science at the University of New Orleans, testified for the defendants. The district court explicitly found Dr. Rose's testimony to be credible, and although the court did not address the credibility of Dr. Wildgen's testimony on the whole, the court did discredit several of Dr. Wildgen's conclusions.

On many issues the experts were in agreement. Both experts performed bivariate ecological linear regression analyses of results from several elections. Both experts analyzed results from both exogenous and indigenous elections. Both experts testified that the degree of racial polarization of the Westwego electorate should not vary between exogenous and indigenous elections.

Based on the testimony of the experts, the district court concluded that black voters in Westwego vote cohesively. This conclusion was based in part on the fact that the exogenous elections studied by Dr. Rose displayed racial polarization of the Westwego electorate. In particular, the district court found that the results from the 1984 Democratic Party presidential primary election, which included a black candidate--Jesse Jackson--and the results from four recent parish-wide elections--each of which included one black candidate--exhibited racial polarity among the Westwego electorate. The district court further found that in every election which included a viable black candidate, the election results revealed that the Westwego electorate is racially polarized. The court also noted for the record that defendants' expert Dr. Wildgen conceded at trial that no black candidate could be elected alderman under Westwego's current at-large system of electing aldermen.

This concession was borne out by the testimony of plaintiffs' expert Dr. Rose at the hearing held on the second remand of the case. Dr. Rose testified that the results from Glenn Green's electoral bid were consistent with his prior testimony--that is, that Green would not be a viable candidate under the existing at-large system, but that he could win in a single-member district with a black voting majority. After reviewing the results from the 1989 aldermanic primary, Dr. Rose testified that if Green had run for alderman in the black majority district created by any of the plaintiffs' proposed districting plans, Green would have finished first in that district and won a seat on the board of aldermen. The district court did not offer any further remark on this portion of Dr. Rose's evidence beyond its general statement that it found Dr. Rose's testimony to be credible.

The defendants do not dispute that the elections which included black candidates reveal racial polarization among the Westwego electorate. Rather, they argue that elections which included no black candidates exhibit no evidence of racial polarity. The plaintiffs agreed with this aspect of the defendants' case, and the district court found that when there were no black candidates in an election, that election showed no racial polarization of Westwego's voters.

As directed by this Court in our mandates remanding the case, the district court made specific findings with respect to a variety of other factors relevant to a section 2 claim.

All parties stipulated to the trial testimony of Dr. Raphael Cassimere, Jr. in Citizens for a Better Gretna v. City of Gretna, 636 F. Supp. 1113 (E.D.La.1986), aff'd, 834 F.2d 496 (5th Cir.1987), cert. denied, 492 U.S. 905, 109 S. Ct. 3213, 106 L. Ed. 2d 564 (1989). In Gretna Dr. Cassimere testified that there was a long history of racial discrimination in Louisiana and Jefferson Parish. As had the district court in Gretna, the district court in this case found Dr. Cassimere's testimony to be "entirely credible." As the court put it, "[t]he unfortunate history of racial discrimination in Louisiana (and the South in general) is a matter of common knowledge."Dr. Cassimere testified that upon its entry into the Union, the state of Louisiana immediately curtailed the rights of both free blacks and slaves. These severe restrictions lasted until the Civil War; when the post-war Reconstruction era brought some electoral and civil success to Louisiana's black citizens, Louisiana's white citizens responded by boycotting elections, altering vote counts, and killing black elected officials and sympathetic whites. When the federal troops withdrew from Louisiana, the legislature immediately acted to segregate and disenfranchise blacks. The Supreme Court ultimately struck down many of the laws adopted by the state legislature. The registrar of voters in Jefferson Parish, however, continued to require blacks--but not whites--to state their age exactly in terms of years, months and days, and to recite the preamble to the United States Constitution as a condition of registration. Before passage of the Voting Rights Act in 1965 no black had ever been elected to office in Jefferson Parish, and no black has ever been elected to Westwego's board of aldermen or any other municipal office.

The district court also found that blacks in Westwego continue to bear the effects of discrimination. In particular, the district court noted that there was a "significant gap between the socioeconomic status of blacks in Westwego and that of whites," and that "data collected by both parties reveals that housing in Westwego remains largely racially segregated." Both of these findings are well supported by the evidence. Data from the 1990 census shows that almost 50% of all of Westwego's black citizens live in a seven-block area in the central portion of the city. Another large segment of Westwego's black population lives in a small area near the river in the northeastern part of the city.

Moreover, the parties stipulated to several facts that reveal all too clearly that the black citizens of Westwego continue to bear the heavy burdens of past racial discrimination. By almost any measure, the black families of Westwego are less well off than their white neighbors. The median family income for whites residing in Westwego is $15,879; the median family income for blacks residing in Westwego is $8,562. Of all the white families residing in Westwego, 12.8% live below the poverty line; of the black families residing in Westwego, 43.3% live below the poverty line. Of the whites over 16 years of age in the labor force in Westwego, 4.4% are unemployed; of the blacks over the age of 16 in the Westwego labor force, 10.9% are unemployed.

In addition, the statistics indicate that black families are less likely to own their own homes and less likely to own a car. Of the white housing units in Westwego, 41.4% are renter-occupied; of the black housing units, 65.1% are renter-occupied. Of the white occupied housing units in Westwego, 16.1% have no vehicle available; of the black occupied housing units in Westwego, 42.9% have no vehicle available.

Heeding the Supreme Court's directive that section 2 vote dilution claims require a "flexible, fact-intensive" inquiry into the "past and present reality" of the political process, Gingles, 106 S. Ct. at 2764, both plaintiffs and defendants offered evidence of past and present political realities in Westwego. The district court did not discuss, either to credit or discount, any of this evidence. Accordingly, this Court has before it no findings to review. It is appropriate to survey some of the evidence presented, however, as it does bear on other findings made by the district court.

All five of Westwego's current aldermen and its mayor testified at the hearing held on the second remand of this case. Review of their testimony and materials from some of their political campaigns (put into evidence by the defendants) indicates that under the current at-large system, to gain a seat on the board of aldermen a potential candidate must join and be active in certain civic organizations. While the aldermen mentioned several organizations, three appear to be particularly influential: the Volunteer Fire Department, the Young Men's Business Club, and the Westwego Civic Association. While none of these organizations officially endorses candidates, the current aldermen testified that these organizations play a central role in Westwego's political and economic life.5

The testimony of Westwego's current officeholders indicates that these influential civic organizations are not equally open to all of Westwego's citizens. The Young Men's Business Club has no black members. The Westwego Civic Association has no black members. Until 1981, the Volunteer Fire Department's by-laws restricted membership to males of "pure white race." Glenn Green was the first black man admitted to the Volunteer Fire Department, and then only after four years of efforts to obtain membership. The Volunteer Fire Department now has three black members. Green has attempted to join the Westwego Civic Association; to do so, he must convince some current member to sponsor his application and then receive a majority of the votes of the members. So far, he has been unable to find a sponsor.

III. The Voting Rights Act, Vote Dilution, and Thornburg v. Gingles

To prevail on a claim of vote dilution under section 2 of the Voting Rights Act,6 the plaintiffs must first meet certain threshold requirements, and then demonstrate that, under the totality of the circumstances, the political process they are challenging is not equally open to citizens of all races or colors. The threshold requirements were first identified in Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 92 L. Ed. 2d 25 (1986). Under Gingles, the plaintiffs must establish 1) that the minority group is sufficiently large and geographically compact to constitute a majority in a single-member district, 2) that the minority group is politically cohesive, and 3) that the white majority votes sufficiently as a bloc to cause it usually to defeat the minority's preferred candidate. 106 S. Ct. at 2766-67. See also Westwego I, 872 F.2d at 1205-06. If the plaintiffs make these threshold showings, the question then becomes whether, under the totality of the circumstances, the plaintiffs have demonstrated a violation of section 2. Westwego I, 872 F.2d at 1206. We first consider whether the plaintiffs have established each of the Gingles requirements.

A. The First Gingles Factor: The Size and Compactness of Westwego's Black Population

In our prior opinions in this case, we have been unable to determine from the record before us whether the black population of Westwego was sufficiently large and geographically compact to constitute a majority of the voting age7 population of a single member district. Westwego II, 906 F.2d at 1045-46; Westwego I, 872 F.2d at 1205 n. 4. The record is now clear. The district court's findings on the latest remand of this case clearly establish that the plaintiffs have met this requirement. The district court found that under plaintiffs' proposed Plan D, one district (District 1) would have a black majority of 59.1% of the total population and a black voting age majority of 52.8%.8 While these numbers do not indicate that the black citizens of Plan D's District 1 would enjoy an overwhelming majority, they would certainly possess the "potential to elect" an alderman of their choice.9 Indeed, the unchallenged testimony of Dr. Rose indicated that if Glenn Green had run for alderman in Plan D's District 1 he would have won the seat.10 In sum, plaintiffs have demonstrated that under a single-member district scheme Westwego's black citizens would possess the potential to elect a candidate of their choice. Accordingly, the plaintiffs have passed the first Gingles test.11

B. The Second Gingles Factor: The Political Cohesiveness of Westwego's Black Population

Usually, plaintiffs in a vote dilution case will attempt to establish both the second and third Gingles factors with statistical evidence of racial polarization of the electorate. As this Court put it in Westwego I, "[e]vidence of racially polarized voting 'is the linchpin of a section 2 vote dilution claim,' and is relevant to establishing two of the three elements set forth in the Gingles decision--the political cohesiveness of the minority group and the ability of the white majority usually to defeat the minority's preferred candidate." 872 F.2d at 1207 (quoting Citizens For a Better Gretna v. Gretna, 834 F.2d 496, 499 (5th Cir.1987)).12 As noted above, both plaintiffs and defendants in this action presented expert testimony on the degree of racial polarization of the Westwego electorate. There is no dispute that the results from every election which included a viable black candidate exhibit racial polarization of Westwego's electorate. Based on this evidence, the district court concluded that blacks in Westwego vote cohesively. The defendants concede the point. Accordingly, the plaintiffs have established the second of the Gingles factors.

C. The Third Gingles Factor: The Voting Patterns of Westwego's White Majority

Just as evidence of racial polarity in the electorate can satisfy the second Gingles requirement, so can it satisfy the third--that the white majority votes cohesively enough to cause it usually to defeat the candidates preferred by the racial or ethnic minority. Despite its conclusion that the Westwego electorate is racially polarized, however, the district court found at trial, and reiterated on the first remand of this case, that "plaintiffs have not sustained their burden to prove that whites would vote as a bloc to defeat a black candidate for alderman in Westwego." On the second remand of this case the district court stated its belief that our mandate prohibited it from addressing this finding again; accordingly, the district court did not revisit this finding, either to repudiate it or to confirm its continuing validity.

In a section 2 claim, as in federal actions generally, the district court's findings of fact are subject to review only for clear error. Fed.R.Civ.P. 52(a); Gingles, 106 S. Ct. at 2781; Jones v. City of Lubbock, 727 F.2d 364, 371 (5th Cir.1984). A finding of fact is "clearly erroneous" only when although there may be 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. Campos v. City of Baytown, Texas, 840 F.2d 1240, 1243 (5th Cir.1988).13 Despite this deferential standard, this Court must reverse the district court's finding that whites in Westwego do not vote as a bloc (assuming that finding still obtains). After our thorough review of the record, we are firmly and definitely convinced that the district court erred in finding that the white citizens of Westwego would not "vote as a bloc to defeat a black candidate for alderman in Westwego."

First, we note that the district court's finding that whites would not vote as a bloc does not take into account any of the evidence offered by plaintiffs at the hearing on the second remand of this case. In particular, the district court's finding does not take into account any of the evidence regarding Glenn Green's candidacy for alderman. Among other things, that evidence showed that while 89% of Westwego's black voters cast ballots for Green, 84% of Westwego's white voters did not. As a result, Green finished twelfth out of sixteen candidates. Because the district court's finding was entered before this evidence was introduced, and was not reaffirmed after this evidence was presented, it is not entirely clear that the finding has any continuing vitality.

Moreover, even at the time it was entered, the district court's finding was, on the record, at odds with its other findings in regard to racial polarization of the Westwego electorate. The district court found that every election which included a viable black candidate exhibited racial polarization. The court did not explain how, in the face of that evidence, which it fully accepted, it could conclude that whites in Westwego did not vote as a bloc.14 Moreover, the district court did not explain how it could reach such a conclusion in the face of the concession by the defendants' expert that no black could be elected alderman under the current at-large system.

In sum, this Court holds that the evidence in the entire record unmistakably demonstrates that white voters in Westwego do vote as a bloc, so that they usually defeat the candidates preferred by Westwego's black voters. Assuming the district court's finding to the contrary survived the latest remand of this case, its finding is clearly erroneous. The plaintiffs have established the third and last of the Gingles threshold factors.15 The remaining question is whether, under the totality of the circumstances, the plaintiffs have proved a violation of section 2.

The question which the court must answer in a section 2 case is whether "as a result of the challenged practice or structure plaintiffs do not have an equal opportunity to participate in the political processes and to elect candidates of their choice." Gingles, 478 U.S. at 44, 106 S.Ct. at 2763. See also Westwego I, 872 F.2d at 1204. While it is widely recognized that "multimember district and at-large voting schemes may operate to minimize or cancel out the voting strength of racial minorities in the voting population," such schemes are not per se violative of section 2. Gingles, 478 U.S. at 47, 106 S.Ct. at 2765; Westwego I, 872 F.2d at 1204. Rather, the determination whether Westwego's at-large election scheme violates section 2 "depends upon a searching practical evaluation of the past and present reality" and on a "functional view of the political process." Gingles, 478 U.S. at 45, 106 S.Ct. at 2763; Westwego I, 872 F.2d at 1204.

Thus, once the court has established that the plaintiffs have met each of the threshold requirements identified in Gingles, it must inquire in some detail into the past and present reality of the challenged electoral structure and the community which employs it. In general, several factors are relevant to this inquiry. The legislative history of the Voting Rights Act identifies nine.16 Some of these factors are more important than others--the two most important are the extent to which members of the minority group have been elected to public office and the extent to which voting in the jurisdiction is racially polarized, Gingles, 106 S. Ct. at 2765 n. 15--but the inquiry required by section 2 always remains an inquiry into the "totality of the circumstances." No one of the factors is dispositive; the plaintiffs need not prove a majority of them; other factors may be relevant. Gingles, 106 S. Ct. at 2759-60; East Jefferson Coalition for Leadership & Development v. Parish of Jefferson, 926 F.2d 487, 494 (5th Cir.1991); Monroe v. City of Woodville, Miss., 881 F.2d 1327, 1334 n. 12 (5th Cir.1989); Campos v. City of Baytown, Texas, 840 F.2d 1240, 1249-50 (5th Cir.1988); Seastrunk v. Burns, 772 F.2d 143, 149-50 (5th Cir.1985); Jones v. City of Lubbock, 727 F.2d 364, 384 (5th Cir.1984); Zimmer v. McKeithen, 485 F.2d 1297, 1305 (5th Cir.1973).

As noted above, the district court found that Louisiana and the South in general had a long history of racial discrimination. In addition, the district court found "entirely credible" the stipulated testimony of Dr. Cassimere regarding racial discrimination in Louisiana and Jefferson Parish. Further, the district court stated that "Westwego is not an island in itself in the history of Louisiana in terms of discrimination." Nevertheless, at trial the district court found that plaintiffs had not presented any evidence of discrimination in Westwego. The district court added: "I say again as I have said before, that for the last 20 years--and certainly, I confidently say for 15--there is no indication that that [history of discrimination] continues."

In Westwego I this Court addressed the district court's findings and remarks on the issue of the history of discrimination in Westwego.

To the extent that these comments indicate that the district court judge disregarded historical evidence of discrimination, based upon his belief that there has been no racial discrimination in Westwego or in the State of Louisiana in the past 15 to 20 years, the district court may have misconceived the role of historical evidence of discrimination in a section 2 claim. The concern underpinning section 2 "is that a certain electoral law, practice or structure interacts with social or historical conditions to cause an inequality in the opportunity enjoyed by black and white voters to elect their preferred representatives." Gingles, 478 U.S. at 47, 106 S.Ct. at 2765. The first factor listed in the Senate Report is "the extent of any history of official discrimination in the state or political subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process." S.Rep. supra, at 28, 1982 U.S.Code Cong. & Admin.News, at 206 (emphasis added). Thus, Congress was concerned not only with present discrimination, but with the vestiges of discrimination which may interact with present political structures to perpetuate a historical lack of access to the political system. Gingles, 478 U.S. at 69, 106 S.Ct. at 2776. While the district court may decide on remand that the historical evidence of discrimination is not significant in the totality of the circumstances, the court must provide a sounder basis for its conclusion than its personal opinion that discrimination has been eradicated in the recent past.

872 F.2d at 1211-12. In light of these instructions, on the first remand of this case the district court did not reiterate its personal opinion that discrimination has been eradicated in Westwego in the last 15 or 20 years. The district court did persist, however, in its conclusion that the plaintiffs had presented no evidence of discrimination specific to Westwego. On the second remand of this case, the district court did not address itself to the issue of the history of discrimination in Westwego.

There can be no question that the district court clearly erred when it found that the plaintiffs had presented no evidence of a history of discrimination in Westwego. Indeed, such a finding is fundamentally inconsistent with several of the district court's other findings. For instance, the district court found "entirely credible" the testimony of Dr. Cassimere, which recounted in some detail the history of discrimination in Louisiana and Jefferson Parish, with particular emphasis on voting-related discrimination. See Gretna, 636 F. Supp. at 1116-1118. Dr. Cassimere also testified that schools in Jefferson Parish remained segregated until 1971, and that black schools were inferior to white schools. Id. Further, the district court acknowledged the long history of discrimination in Louisiana, and affirmatively stated that, in terms of discrimination, Westwego was "not an island in itself." Moreover, the district court explicitly found that "[t]here is substantial evidence that blacks in Westwego continue to bear the effects of discrimination." The court noted that the evidence showed a "significant gap between the socioeconomic status of blacks in Westwego and that of whites," and that "data collected by both parties reveals that housing in Westwego remains largely racially segregated." Finally, the plaintiffs presented substantial evidence that politics in Westwego continue to suffer from racial discrimination. In particular, the plaintiffs elicited testimony from the current aldermen and mayor of Westwego that certain civic organizations play a central role in Westwego's political life, and that those organizations are not open to blacks. Indeed, until 1981 one of those organizations restricted its membership to males of "pure white race."

On the basis of the record before us, this Court can come to no other conclusion but that the plaintiffs have adequately demonstrated that a history of discrimination in Louisiana, Jefferson Parish, and Westwego has combined with the existing political structures in Westwego to "perpetuate a historical lack of access to the political system." The evidence presented--and indeed, the district court's own findings--admit of no other result. The district court clearly erred when it found that plaintiffs had presented no evidence of a history of discrimination in Westwego.

As discussed above,17 both parties presented extensive evidence, from both indigenous and exogenous elections, demonstrating that Westwego voters vote along racial lines. The district court found, and the defendants concede, that every election which included a viable black candidate exhibits racial polarization of the Westwego electorate. The district court's finding on this issue was well supported by--indeed, compelled by--the evidence before it.

As stipulated to by the parties and found by the district court, Westwego's black citizens, inheritors of a long history of discrimination, continue to suffer significant disabilities. As noted,18 the evidence shows that in Westwego the median income for whites is roughly twice that of blacks; over three times as many black families as white families live below the poverty line; while the white labor force in Westwego is fully employed, the black labor force faces 10.9% unemployment; fewer black families have cars; 50% fewer black families than white families own their own homes; housing remains racially segregated. The district court did not overstate the case when it concluded that the evidence reveals "a significant gap between the socioeconomic status of blacks in Westwego and that of whites."

D. The Extent to Which Blacks Have Been Elected to Office in Westwego

As noted above, the Supreme Court has held that the two most important factors to consider in inquiring into the totality of the circumstances in a section 2 claim are 1) whether the electorate is racially polarized, and 2) whether, under the challenged electoral practice, the minority group has been able to elect candidates of its choice. Gingles, 106 S. Ct. at 2765 n. 15. The plaintiffs have made strong showings on both: the degree of racial polarization is discussed above, and the parties have stipulated that no black has ever been elected to municipal office in Westwego. The only black man to run for alderman, Glenn Green, finished twelfth in a field of sixteen candidates (despite getting 89% of the black vote).

Finally, the legislative history of the Voting Rights Act identifies two factors which may be of limited relevance when inquiring into the totality of the circumstances: 1) the responsiveness of elected officials to the needs of the minority community, and 2) the political subdivision's justification for the challenged electoral practice.19 In this case, the degree of responsiveness of Westwego's elected officials is disputed. The district court found that the plaintiffs had not adequately proved a lack of responsiveness, and after reviewing all of the evidence, we cannot say that the district court clearly erred in so finding.

The district court initially found that the policy underlying Westwego's at-large election scheme was not tenuous. At the trial of this action, and again on the first remand, the district court found that the defendants

presented evidence that the at-large system is justified by concerns of administrative efficiency. The undisputed evidence at trial revealed that Westwego's five aldermen head the city's five departments. Thus, the board of aldermen is functionally better suited to at-large election than election by single-member district. Certainly, then, the policy underlying the city's electoral process cannot be considered tenuous.

Although we expressed some misgivings about this conclusion in our prior opinion, see Westwego I, 872 F.2d at 1212, we need not revisit the issue. The parties now agree that Louisiana law prohibits aldermen from serving as department heads and have stipulated that that practice has ended in Westwego. Accordingly, the district court's initial finding that the at-large system was justified by "concerns of administrative efficiency" no longer finds any support in the evidence. Moreover, the defendants have not argued that there is any other policy or interest served by the at-large system. The district court did not revisit this issue on the second remand of this case. We assume, therefore, that there is no longer any issue in this case with respect to the policy underlying Westwego's at-large system.

The plaintiffs have clearly established the existence of several of the factors relevant to an inquiry into the totality of the circumstances. No blacks have ever been elected to any municipal office in Westwego; the electorate is racially polarized; Westwego, like Jefferson Parish and Louisiana generally, has a long history of discrimination; the black citizens of Westwego continue to bear the effects of that history of discrimination. After reviewing the entire record, this Court is firmly and definitely convinced that the plaintiffs have demonstrated, under the totality of the circumstances, that these conditions interact with Westwego's at-large election scheme to dilute the voting power of Westwego's black citizens in violation of section 2 of the Voting Rights Act.

Having proven that the current at-large electoral scheme impermissibly dilutes the voting strength of Westwego's black citizens, the plaintiffs are entitled to a remedy. Plaintiffs urge this Court to order Westwego to implement a single-member districting plan along the lines of their proposed Plan D, and to order Westwego to hold a special election under that plan. We must decline.

The Supreme Court has made clear that state legislatures have "primary jurisdiction" over legislative apportionment. White v. Weiser, 412 U.S. 783, 795, 93 S. Ct. 2348, 2354, 37 L. Ed. 2d 335 (1973). Thus, "reapportionment is primarily a matter for legislative consideration and determination.... [J]udicial relief becomes appropriate only when a legislature fails to reapportion according to federal constitutional requisites in a timely fashion after having had an adequate opportunity to do so." Id., 412 U.S. at 794-95, 93 S. Ct. at 2354 (quoting Reynolds v. Sims, 377 U.S. 533, 586, 84 S. Ct. 1362, 1394, 12 L. Ed. 2d 506 (1964)). See also Wise v. Lipscomb, 437 U.S. 535, 539, 98 S. Ct. 2493, 2497, 57 L. Ed. 2d 411 (1978) ("The Court has repeatedly held that redistricting and reapportioning legislative bodies is a legislative task which the federal courts should make every effort not to preempt."). These principles apply with equal force to violations of the Voting Rights Act. Mississippi State Chapter, Operation PUSH v. Mabus, 932 F.2d 400, 405-06 (5th Cir.1991) ("[T]he district court properly deferred to the Mississippi Legislature in the first instance to remedy the existing violations [of the Voting Rights Act].").

Moreover, these principles do not apply only to state legislatures: this Court has repeatedly held that it is appropriate to give affected political subdivisions at all levels of government the first opportunity to devise remedies for violations of the Voting Rights Act. See, e.g., East Jefferson Coalition for Leadership & Development v. Parish of Jefferson, 926 F.2d 487, 490 (5th Cir.1991) (affirming district court's decision, after finding Parish's existing system caused impermissible vote dilution, to impose redistricting plan drawn by Parish); Citizens for a Better Gretna v. City of Gretna, 834 F.2d 496, 504 (5th Cir.1987) (affirming district court's decision to afford city first opportunity to propose plan to remedy vote dilution caused by at-large election of board of aldermen); Cook v. Luckett, 735 F.2d 912, 918 (5th Cir.1984) (county supervisors had jurisdiction over county apportionment).

Westwego's board of aldermen has primary jurisdiction over Westwego's electoral system. It must be left to that body to develop, in the first instance, a plan which will remedy the dilution of the votes of the city's black citizens.20 If Westwego fails to develop such a plan in a timely manner, or fails to develop a plan which fully remedies the current vote dilution, the responsibility for devising a remedial plan will devolve onto the federal district court. East Jefferson Coalition for Leadership & Development v. Parish of Jefferson, 703 F. Supp. 28, 29 (E.D.La.1989), aff'd, 926 F.2d 487 (5th Cir.1991) (citing Cook, 735 F.2d at 918) ("Only when the legislative plan is 'uncorrectably invalid' should the court consider a scheme submitted by a private litigant or formulate its own plan.").

Accordingly, this Court will again remand this case to the district court. The district court is directed to give the City of Westwego 120 days to develop and submit to the United States Attorney General, in accord with section 5 of the Voting Rights Act, a plan to remedy the current vote dilution. If and when the plan submitted is precleared, the district court shall review the plan to determine whether it rectifies the current vote dilution and otherwise complies with the law. If Westwego does not develop and have precleared a remedial plan in a timely fashion, or if the district court determines that the plan proposed by Westwego does not comply with section 2 of the Voting Rights Act or other provisions of law, the court shall determine what remedial plan to impose, whether of the plaintiffs' or its own design.

This Court has every confidence that the City of Westwego will promptly and in good faith develop a plan that remedies the current violation of section 2, and that the district court will insure that this Court's mandate is fully and faithfully executed. Of course, should events prove otherwise, plaintiffs may again resort to this Court by means of an appropriate filing, including, if necessary, a petition for a writ of mandamus. This panel shall retain jurisdiction of this case solely to the extent that any future filing with this Court shall be returned to this panel.

Finally, the district court shall also determine on remand whether the plaintiffs are entitled to an award of court costs and attorneys' fees, and, if so, in what amount.

The time has come for Westwego to leave behind its at-large system of electing aldermen, a system which has effectively barred the black citizens of Westwego from any meaningful role in Westwego's government. The judgment of the district court is reversed. Judgment is rendered for the plaintiffs. The case is remanded for the development and implementation of a remedial plan which will remedy the vote dilution caused by the current at-large scheme, and for a determination of what amount, if any, the plaintiffs are entitled to recover in court costs and attorneys' fees.

REVERSED, RENDERED, AND REMANDED.