Carrie Meek and other plaintiffs appeal from the district court's grant of summary judgment in favor of Metropolitan Dade County, Florida ("the county"). The district court held that the plaintiffs failed to establish certain required elements of their vote dilution case, and therefore granted summary judgment for the defendants and denied the plaintiffs' summary judgment motion. Because the district court erred in applying the law regarding vote dilution, we reverse and remand.

The plaintiffs, Black and Hispanic citizens and registered voters of Dade County, brought suit against the county alleging that the at-large scheme for the election of the Board of County Commissioners, which is the legislative and governing body of the county, violated section two of the Voting Rights Act by diluting Black and Hispanic voting strength.1 Although the county has an ethnically diverse population, the three major groups, as identified by the parties and the district court, are: Blacks, Hispanics, and Non Latin Whites. None of the groups constitutes a majority of the total number of registered voters. As the district court noted: "Non Latin Whites comprise approximately 37 percent of the population and 48.67 percent of the registered voters; Hispanics (including Hispanics born in the United States) comprise approximately 43 percent of the population and 32.96 percent of the registered voters; and Blacks comprise approximately 20 percent of the population and 18.37 percent of the registered voters of Dade County." Meek v. Metropolitan Dade County, Florida, No. 86-1820, slip op. at 15 (S.D.Fla. Oct. 5, 1988) [hereinafter Order of October 5, 1988].

The county electoral system for county commissioners is part of a so-called "federated plan" that provides for the allocation of authority between the county and the metropolitan governments. The county commission is composed of a mayor and eight commissioners. Although the mayor may live anywhere in the county, each of the commissioners must reside in a different one of the eight residence districts. Each commissioner, however, must run for election county-wide. Thus, although the county is divided into various residence districts, the elections are conducted at-large and county-wide.

The plaintiffs' vote dilution claim was brought under section two of the Voting Rights Act as amended in 1982.2 The Supreme Court in Thornburg v. Gingles stated that:

[t]he essence of a Sec. 2 claim is that a certain electoral law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives. This Court has long recognized that multi-member districts and at-large voting schemes may " 'operate to minimize or cancel out the voting strength of racial [minorities in] the voting population.' "

478 U.S. 30, 47, 106 S. Ct. 2752, 2764, 92 L. Ed. 2d 25 (1986) (citation and footnote omitted) (interpolation in original). The Court further explicated the requirements a plaintiff must meet in pursuing a section two vote dilution claim:

First, the minority group must be able to demonstrate that it is sufficiently large and geographically compact to constitute a majority in a single-member district. If it is not, as would be the case in a substantially integrated district, the multi-member form of the district cannot be responsible for minority voters' inability to elect its candidates. Second, the minority group must be able to show that it is politically cohesive. If the minority group is not politically cohesive, it cannot be said that the selection of a multimember electoral structure thwarts distinctive minority group interests. Third, the minority must be able to demonstrate that the white majority votes sufficiently as a bloc to enable it--in the absence of special circumstances, such as the minority candidate running unopposed--usually to defeat the minority's preferred candidate. In establishing this last circumstance, the minority group demonstrates that submergence in a white multimember district impedes its ability to elect its chosen representatives.

Gingles, 478 U.S. at 50-51, 106 S.Ct. at 2766-67 (footnotes and citations omitted; emphasis in original).

The Court further explained the causation threshold of Gingles as follows:

The reason that a minority group making such a challenge must show, as a threshold matter, that it is sufficiently large and geographically compact to constitute a majority in a single-member district is this: Unless minority voters possess the potential to elect representatives in the absence of the challenged structure or practice, they cannot claim to have been injured by that structure or practice. The single-member district is generally the appropriate standard against which to measure minority group potential to elect because it is the smallest political unit from which representatives are elected. Thus, if the minority group is spread evenly throughout a multimember district, or if, although geographically compact, the minority group is so small in relation to the surrounding white population that it could not constitute a majority in a single-member district, these minority voters cannot maintain that they would have been able to elect representatives of their choice in the absence of the multimember electoral structure. As two commentators have explained:

"To demonstrate [that minority voters are injured by at-large elections], the minority voters must be sufficiently concentrated and politically cohesive that a putative districting plan would result in districts in which members of a racial minority would constitute a majority of the voters, whose clear electoral choices are in fact defeated by at-large voting. If minority voters' residences are substantially integrated throughout the jurisdiction, the at-large district cannot be blamed for the defeat of minority-supported candidates.... [This standard] thus would only protect racial minority votes from diminution proximately caused by the district plan; it would not assure racial minorities proportional representation." Blacksher & Menefee 55-66 (footnotes omitted; emphasis added).

Gingles, 478 U.S. at 50, 106 S.Ct. at 2766 n. 17.

While the full contours of Gingles have not been agreed upon by this circuit, it is clear that the plaintiff must establish the three core Gingles factors in order to prevail on a vote dilution claim. See Solomon v. Liberty County, 899 F.2d 1012, 1017-21 (Kravitch, J., specially concurring), 899 F.2d at 1037 (Tjoflat, C.J., specially concurring) (11th Cir.1990) (evenly divided en banc court).

The district court's order granting summary judgment in favor of the county held that the plaintiffs had not established that the Non Latin White bloc voting was "legally sufficient" because the plaintiffs had not proved "the existence of a Non Latin White bloc majority that usually defeats the election of the minority's preferred candidate."3 The district court reasoned that where Non Latin Whites did not constitute a majority, Non Latin White bloc voting could not, through numerical superiority, defeat the choice of minority voters.

After the plaintiffs filed a motion to alter and amend the judgment, the district court, albeit conceding that it had erred in its understanding of the statistical data, issued an opinion denying the motion. The district court found that Hispanics were in fact politically cohesive, but nevertheless the district court refused to alter the judgment because the plaintiffs still had not proved the third Gingles factor of legally significant Non Latin White bloc voting that caused defeat at the polls for the minorities' preferred candidates.

The county contends that we lack jurisdiction because the plaintiff-appellants failed to give timely notice of appeal. We disagree. A clerk in the district court provided plaintiffs' counsel with contradictory information concerning the date upon which the order denying the plaintiffs' motion to alter the judgment was entered. The clerk's statement to counsel led him to believe that the time for appeal under Fed.R.App.P. 4(a) had already expired, although in fact counsel still had two remaining days to file a notice of appeal. Had counsel been informed accurately, he would have been able to file a timely notice of appeal. He was led, however, by the district court clerk's statement to believe that his only avenue was through Fed.R.App.P. 4(a)(5), which permits the district court, upon a showing of excusable neglect or good cause to extend the time for filing a notice of appeal upon motion filed not later than 30 days after the expiration of the time prescribed by Rule 4(a). Fed.R.App.P. 4(a) establishes the period for timely appeal in a civil case.

We have noted previously that the addition of the "good cause" language "indicates that the standard should be interpreted flexibly." Davis v. Page, 618 F.2d 374, 378 (5th Cir.1980) (citation omitted), aff'd in part and rev'd in part on other grounds, 640 F.2d 599 (1981) (en banc),4 vacated on other grounds, 458 U.S. 1118, 102 S. Ct. 3504, 73 L. Ed. 2d 1380 (1982). We hold that counsel's justifiable reliance on the district court clerk's erroneous information provides "good cause" under the unique circumstances of this case. Furthermore, this court, rather than "second-guess" the district court, defers to the district court so long as its decision does not constitute an abuse of discretion. Id.; see also Chipser v. Kohlmeyer & Co., 600 F.2d 1061, 1063 (5th Cir.1979). Accordingly, we conclude that the extension of time was properly granted, the notice of appeal was timely filed, and that we have jurisdiction.

The county contends that it was entitled to summary judgment as a matter of law because the plaintiffs failed to "establish the existence of an essential element [of their case], and on which [they] will bear the burden of proof at trial." Celotex Corp. v. Catrett, 477 U.S. 317, 106 S. Ct. 2548, 2552, 91 L. Ed. 2d 265 (1986). The district court agreed that the plaintiffs had failed to satisfy the third Gingles prong.

Generally, in summary judgment we review the district court's legal conclusions de novo. We are mindful that after trial of a vote dilution case, "the clearly erroneous test of Rule 52(a) is the appropriate standard for appellate review of a finding of vote dilution." Gingles, 478 U.S. at 79, 106 S.Ct. at 2781. Nevertheless, we may "correct errors of law, including those that may infect a so-called mixed finding of law and fact, or a finding of fact that is predicated on a misunderstanding of the governing rule of law." Id. (quoting Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 104 S. Ct. 1949, 80 L. Ed. 2d 502 (1984) (citing Pullman-Standard v. Swint, 456 U.S. 273, 287, 102 S. Ct. 1781, 1789, 72 L. Ed. 2d 66 (1982))). In this case, the district court's conclusion was premised on an erroneous application of the law, and we therefore reverse for application of "the governing rule of law."

As noted above, this circuit is divided on the proper application of Gingles. In Solomon, the evenly divided court was united in stating that a plaintiff, at a minimum, must satisfy the three core Gingles factors. The en banc court divided on the question of whether proof of the three core Gingles factors was sufficient and whether a defendant could raise as a defense the lack of racial bias in the community. As the court is of the unanimous opinion that the three core Gingles factors are necessary prerequisites to a vote dilution case, see Solomon, 899 F.2d at 1017-18 (Kravitch, J., specially concurring), 899 F.2d at 1034-36 (Tjoflat, C.J., specially concurring), we will discuss these requirements first as our resolution of the issues concerning the core factors may moot further concerns.

The heart of the dispute relates to the ability of the Non Latin White bloc usually to defeat the minority group's choice. The district court did not decide whether Non Latin White voters constituted a majority of the Dade County registered voters as the district court chose to decide "the case on other grounds...." Meek v. Metropolitan Dade County, Florida, No. 86-1820, slip op. at 9 (S.D.Fla. Nov. 6, 1988) [hereinafter Order of November 6, 1988]. The ground relied upon by the district court was that the Non Latin White voting bloc, the existence of which the district court "cannot deny," id. at 8, did not cause the election losses suffered by Black and Hispanic voters. The court stated that to support the third prong of the Gingles test, "it is imperative that the plaintiffs demonstrate that the Non Latin White community not only vote [sic] as a bloc, but that they are a voting bloc majority." Order of October 5, 1988 at 20.

The district court's overall conclusion, common to both Blacks and Hispanics, that the Non Latin Whites did not cause the loss of candidates of either group rested on differing grounds as to each group. The district court applied different rationales to the vote dilution claim of Blacks and Hispanics.

In its first order, the district court, noted that:

a keen hostility in fact exists between ... [Blacks and Hispanics. The expert's] affidavit shows that without exception, when a Black or Hispanic candidate is running against a Non Latin White candidate, the other minority group will vote for the Non Latin White.... In these elections [where Black candidates lost by a small margin], had there been no hostility between Blacks and Hispanics, the election of the Black over the Non Latin White would have been all but guaranteed.

Order of October 5, 1988 at 24 (record citation omitted).

In its second order, the district court reiterated that "one of the many conceivable reasons why Blacks in Dade County have difficulty in electing their preferred candidates [is that] Hispanics do not vote for them." Order of November 6, 1988 at 8. The district court then attributed Black losses, not to the white voting bloc's refusal to vote for the Black candidate, but to "[o]ther important factors includ[ing] lack of Hispanic support, name recognition, incumbency, low voter turnout, etc." The district court concluded that because in that court's view the plaintiffs had "failed to satisfy the causation element of Thornburg [v. Gingles] 's third prong, their claim of vote dilution must fail." Id.

The district court erred in its application of the causation prong of the Gingles analysis. The district court was aware that the "multi-ethnic" nature of Dade County differed from the simple majority/minority context contemplated by Gingles. The district court, noting the hostility between Blacks and Hispanics, implicitly recognized that the hostility created a permanent anti-minority majority in Dade County, with Blacks siding with Non Latin Whites against Hispanic candidates, and Hispanics siding with Non Latin Whites against Black candidates. The district court, however, did not attribute to this phenomenon the appropriate legal significance.

Keeping in mind the Court's admonition that "[t]he essence of a Sec. 2 claim is that a certain law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representative[,]" we consider the "functional effect" of the existing system. 478 U.S. at 48, 106 S.Ct. at 2765 n. 15. Here the social reality is that Black and Hispanic voters are hostile toward each other in the electoral arena. Similarly, Non Latin Whites are politically cohesive and tend not to vote for Hispanics or Blacks. The district court concluded that because Non Latin Whites by themselves could not block the electoral success of Blacks, Blacks had not succeeded in proving that Non Latin Whites caused the defeat of "minority" voters. The district court erred in failing to recognize that coalitions can form a legally significant voting bloc, and that a coalition of Hispanics and Non Latin Whites could form the relevant majority voting bloc for the purpose of the third Gingles factor.

As to Hispanics in particular, the district court apparently concluded that Hispanic candidates did not lose because they lacked Black support; instead, Hispanic candidates lost because Hispanics were under-registered to vote in proportion to their population. The district court reasoned:

the fact that the Hispanic population in Dade County is actually larger in size than any other group in the county including Non Latin Whites, that Hispanics could also be the largest segment of registered voters with an effective registration drive, and that it is conceivable that Non Latin Whites are not even a majority of the registered voters in Dade County, all combine to defeat the Hispanic defendants [sic] claims that their votes are being impermissibly diluted by a legally significant Non Latin White voting bloc.

Order of November 6, 1988 at 9-10.

In considering the import of the large population percentage of Hispanics, the district court in its initial order noted the authority of 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). In McKeithen, the en banc court reversed the trial court and panel decisions, which had "held that an at-large scheme cannot work a dilution of black voting strength where blacks, though constituting a minority of registered voters, comprise a majority of the total population of the parish." 485 F.2d at 1300. Hence, the district court here properly rejected the county's contention that Gingles could not apply at all in a setting where the Non Latin White bloc did not constitute a majority of the total population. Nevertheless, the district court interpreted the third Gingles prong as requiring the plaintiffs to prove that the Non Latin White bloc in and of itself could defeat the minority choice.

The district court concluded that Non Latin White bloc voting did not cause the defeat of Hispanic candidates; rather, the failure of Hispanics to register to vote was the reason Hispanics lost, i.e., if Hispanics registered and voted, their electoral success would be guaranteed. Critical to the district court's conclusion was its finding that Hispanics could constitute "the largest segment of registered voters with an effective registration drive." Order of November 6, 1988 at 10. There was, however, no evidence before the district court of voting age populations. The only evidence related to total population and numbers of registered voters.5 We must, therefore, conclude that the district court erred in concluding that Hispanic underregistration causes electoral losses as there is no evidence of the relative numbers of voting age persons in the various voting blocs.6

We conclude that Hispanics lose because Non Latin White candidates receive more racially polarized votes than Hispanic candidates. Furthermore, given the polarization of voting and the hostility between Blacks and Hispanics, the fact that Hispanics constitute no more than approximately 32.96% of registered voters7 leads to the conclusion that Non Latin Whites and hostile politically cohesive Blacks form the remaining majority. Non Latin Whites alone are 48.67% of the registered voters.8 In light of the district court's finding that Blacks and Non Latin Whites are politically cohesive and tend to vote against Hispanic candidates, such a setting is ripe for a vote dilution claim, and the Hispanic plaintiffs have demonstrated the third Gingles factor, that they usually are denied the opportunity to elect their preferred representatives because of an opposing voting bloc.

The county argues that the plaintiffs fail under Gingles because they cannot satisfy their burden of demonstrating that the plaintiff minority class will be able to elect more representatives under a single-member district plan than it now elects under the at-large scheme. In support of its claim, the county points to the first Gingles factor, which requires the minority group "to demonstrate that it is sufficiently large and geographically compact to constitute a majority in a single-member district. If it is not, as would be the case in a substantially integrated district, the multi-member form of the district cannot be responsible for minority voters' inability to elect its candidates." 478 U.S. at 50, 106 S. Ct. at 2766 (footnote omitted).

We agree with the county that if the plaintiffs would have no greater opportunity to elect their preferred representatives under a single-member district plan, then they cannot show that the at-large scheme is causing their inability to elect their preferred representatives. Gingles addresses not only a group's ability to elect a satisfactory candidate (that is, a candidate for whom the minority voter is willing to cast a vote), but the group's ability to elect its preferred candidate. The county assumes that if a Black consistently has been elected from district three, then Blacks have been successful in electing their preferred candidate. Further, the county assumes that the only relevant inquiry relates to the ability of Blacks and Hispanics to elect representatives based on the residency populations as those districts are now drawn. Neither assumption, however, is necessarily true.

Relying on existing residence district borders, the county states that Hispanics and Blacks are a majority of registered voters only in districts two and three respectively, that a Hispanic and a Black have represented those respective districts for at least the past seven years, and concludes that neither Hispanics nor Blacks could attain any greater electoral success by switching to single-member districting because Blacks and Hispanics would still only elect one representative each.

There are two hidden premises in the county's argument.

First, the county assumes that no other districts could be drawn that would permit Blacks or Hispanics to elect more representatives than they do now. The county's reliance on the makeup of the existing districts is misplaced. At issue is whether the population distribution could be differently apportioned among redrawn districts such that minority Black and Hispanic voters have greater representation. According to the Supreme Court, "[t]he single-member district is generally the appropriate standard against which to measure minority group potential to elect because it is the smallest political unit from which representatives are elected." 478 U.S. at 50 n. 17, 106 S. Ct. at 2766 n. 17. In Thornburg v. Gingles, the district court did not simply count the votes within the existing districts, but considered whether new districts could be drawn that would enable Blacks to enjoy greater representation more in keeping with their numbers in the population. See Gingles, 478 U.S. at 38, 106 S.Ct. at 2760.

The district court should consider whether single-member districts could be drawn which would permit Blacks and Hispanics equal opportunity to elect representatives of their choice. That is the baseline against which the existing at-large system should be measured. We agree with the observation made by Chief Judge Tjoflat in his now-vacated panel opinion in Solomon that "[o]f course, the resulting single-member electoral system must achieve a more proportional representation of minorities than did the previous multi-member system." Solomon v. Liberty County, 865 F.2d 1566, 1572 n. 5 (11th Cir.1988) (vacated). The district court, however, did not address whether alternative districts could be drawn such that Blacks would have the opportunity to elect one representative from a "safe" district and also have the possibility of occasionally electing a representative from a more marginal "swing" district. Section two does not provide a guarantee of proportional representation. It does, however, require equal opportunity to participate and elect representatives of the minority group's choice.

Second, the county assumes that the Hispanic and Black representatives elected from districts two and three are the "preferred representatives" of the respective minorities. That is not necessarily so.

Whether the candidate elected is the preferred candidate of a particular minority group turns on more than the candidate's race. The at-large scheme could cause the candidate elected from the Black and Hispanic residence district to be someone who is less responsive to the particular needs of an insular community than would be the case under a single-member district structure. Naturally, the less responsive candidate would not be the "preferred" candidate when evaluated against a single-member district baseline and the candidate who would be elected under that system. For example, in an at-large system, the candidates from a Black or Hispanic residence district who hold themselves forth for at-large election may not be the preferred candidate of the minority group, but rather may be a candidate who is perceived as viable given the at-large election scheme. Cf. Gingles, 478 U.S. at 76 n. 37, 106 S. Ct. at 2779 n. 37 (success of Black candidate not always attributable to support in the Black community). Although the representative from district three, by way of example, is in fact Black, that does not mean that particular candidate is the preferred candidate of Black voters; the candidate may be a person for whom Blacks are willing to vote when the alternative would be the election of a White candidate, but that does not mean that the same candidate would be elected in a single-member district plan. In contrast, under a single-member district plan, a candidate who, because he or she is more responsive to Black concerns, is the preferred candidate of Blacks could be elected even though he or she is unacceptable to White voters and would not be elected under an at-large scheme. In short, the at-large scheme may cause the Black candidate to be someone other than the preferred candidate of Black voters. That a candidate is of the race of the group is not dispositive; the legal standard is whether the candidate is the preferred candidate of the minority group.

Whether a given minority candidate who has long enjoyed electoral success is the preferred representative requires appraisal of local facts within the ken of the district court and best left to it.

At oral argument, the county objected to any judicial involvement on the ground that Blacks and Hispanics by joining together have sufficient political clout to change the electoral system by instituting a referendum as permitted by Florida law. Assuming that such a referendum is available, the plaintiffs nevertheless have the right to pursue their judicial remedy as provided by federal statute, rather than embarking on the task of effecting change through the referendum method. Minorities have resorted frequently to the federal courts for vindication of their rights. Moreover, we note that minorities may be less successful at organizing and orchestrating voting campaigns than are other groups, and may therefore prefer the statutory remedial scheme. Finally, the fact that a minority group may perceive itself as less successful in the political arena may cause diminished participation in such a referendum. Regardless of the viability of alternative avenues, the plaintiffs are entitled to enjoy their statutory right to seek redress in the federal courts, and we cannot deny it to them.

In sum, as to the first Gingles prong, we affirm the district court's conclusion that the plaintiffs have satisfied their burden. The second prong is not in dispute. As to the third Gingles prong, we conclude that the district court's legal analysis was premised upon legal error.

As the district court here denied relief based on its erroneous interpretation of the third Gingles factor, we remand for reconsideration in light of the principles enunciated here. See Gingles, 478 U.S. at 78-79, 106 S. Ct. at 2781. Specifically, the district court must decide whether Blacks or Hispanics have thus far usually elected preferred representatives. The court should also consider whether Blacks or Hispanics are impaired in their ability to elect representatives of their choice by the manner in which the voting districts are now drawn. This case does not require us at this juncture to reenter those disputed issues that divided the en banc court in Solomon. Under the Gingles principles enunciated here, the district court could still conclude that the plaintiffs have not satisfied the three necessary Gingles requirements: the district court could find that Blacks and Hispanics have elected their preferred representatives, and that the existing districting plan does not impair the ability of either group to elect representatives of its choice. In that event the issue of affirmative defenses would not be reached. If the district court found the plaintiffs to have satisfied the three Gingles factors, it appears on the facts of this case, where the district court has found the existence of racial hostility between Blacks and Hispanics driving electoral results, that under either view of the Solomon en banc court, the plaintiffs might be entitled to relief. Given the posture of this case, we decline to assume the existence of facts in order to apply the law hypothetically. Further, the Supreme Court has made clear that the district court should analyze the facts in the first instance. We therefore decline to reach issues not necessary to the decision of this case.

REVERSED and REMANDED.