Thompson v. Glades County Board of County Commissioners

TJOFLAT, Circuit Judge,

dissenting:

In finding a § 2 remedy appropriate for Glades County, Florida, the majority expands our circuit’s Voting Rights Act (or “Act”) jurisprudence into unanticipated territory. Because I believe that the majority’s two holdings shift our circuit precedent afield, I respectfully dissent.

The majority first holds that the district court clearly erred in finding that the plaintiffs failed to satisfy the threshold Gingles requirement for a viable remedy, and second, concludes that the district court clearly erred in failing to state with particularity its reasons for finding that the totality of the circumstances did not support vote dilution. I address each of these holdings in turn.

I.

The majority’s conclusion that the district court erred in its analysis of the first *1268 Gingles requirement turns on the district court’s finding that the illustrative district prepared by the plaintiffs constituted an “influence district” — insufficient as a § 2 remedy — where a bare 50.23% majority of the voting age population would be African American. Under Gingles, a viable § 2 remedy requires a minority population to be “sufficiently large and geographically compact to constitute a majority in a single-member district.” Thornburg v. Gingles, 478 U.S. 30, 50 , 106 S.Ct. 2752, 2766 , 92 L.Ed.2d 25 (1986); see also Negron v. City of Miami Beach, Florida, 113 F.3d 1563 , 1567 (11th Cir.1997) (finding the first Gingles requirement satisfied where Hispanic plaintiffs would constitute 63.77%, 62.24%, and 56.80% majorities in three remedial single-member districts). In rejecting the district court’s factual finding, the majority relies on two premises: (1) even the barest of majorities can satisfy the Gingles threshold requirement; and (2) crossover votes may be included to satisfy that requirement. The former premise places our court in the unpalatable position of mechanically trumping judicial fact-finding on a numerical bright-line basis, and the latter prematurely resolves an issue on which both our circuit and the Supreme Court have recently and purposefully abstained.

A.

Although the language of the Gingles Court suggests that a pure mathematical majority satisfies the first requirement, the Court also articulated that the policy rationale of the first requirement is to afford minority groups an opportunity to elect preferred candidates: “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.” Gingles, 478 U.S. at 51 n. 17, 106 S.Ct. at 2767 n. 17. Common sense militates against conferring judicial safe harbor as a matter of law where a remedial plan achieves a majority-minority district by only a handful of individuals. I submit that the first Gingles requirement cannot be read as satisfied where a district court finds that a slim majority would not give a minority population the “potential to elect” representatives of their choice. In effect, a bare mathematical majority is no shibboleth for a proper § 2 remedy in vote dilution cases.

The evolving law regarding “influence districts” counsels against adopting bare numerical majorities as the mark of a viable § 2 remedy. In Voinovich v. Quilter, 507 U.S. 146 , 113 S.Ct. 1149 , 122 L.Ed.2d 500 (1993), the Supreme Court suggested that the first Gingles requirement might need to be “modified or eliminated” if minority plaintiffs proved that they could harness significant political influence within a given electorate while constituting less than 50% of the relevant voting age population. Id. at 156-60 , 113 S.Ct. at 1157-58 (emphasis added). No definite rule emerged, however, and the Court has deferred the question for well over a decade. 1 Recently, in League of United Latin American Citizens v. Perry, 2 — U.S. *1269 —, 126 S.Ct. 2594 , 165 L.Ed.2d 609 (2006), a plurality of the Court 3 continued to punt on the issue, merely assuming for the purposes of that particular litigation that it would be possible to state a § 2 claim for a racial group that constituted less than 50% of the relevant voting age population. Id. at 2624 (plurality opinion). Ultimately, on the facts of Perry , the Court declined to find the proposed influence district to be an adequate § 2 remedy. See id. at 2625 (“That African-Americans had influence in the district ... does not suffice to state a § 2 claim in these cases. The opportunity ‘to elect representatives of their choice,’ 42 U.S.C. § 1973 (b), requires more than the ability to influence the outcome between some candidates .... ”).

Despite not deciding the issue, the Perry plurality echoed what was already law in our own circuit. In Dillard v. Baldwin County Commissioners, 376 F.3d 1260 (11th Cir.2004), we held that African American citizens in Alabama had no vote dilution remedy where they constituted less than 10% of the electorate at issue and could not comprise a majority of the voting age population in any remedial district. Id. at 1267 . The plaintiff in Dillard sought to persuade the court that numerically small minority groups could create “influence districts” that would suffice under the first Gingles requirement because minority voters would “play a substantial, although not decisive, role in the electoral process.” Id. (quoting Georgia v. Ashcroft, 539 U.S. 461, 482 , 123 S.Ct. 2498 , 156 L.Ed.2d 428 (2003)). However, we rejected that argument in Dillard . In declining to advance influence districts as a cognizable remedy under § 2, the Dillard panel observed that the facts of the case established no form of relief that could “empower the protected minority group with a meaningful opportunity to elect the candidate of its choice.” Id. at 1262 (emphasis added). Where a minority group carries a majority by only a slim margin, a factual *1270 possibility exists that the district’s minority population could “influence” but not have a “meaningful opportunity” to determine the outcome of elections. See Perry, 126 S.Ct. at 2616 , 126 S.Ct. 2594 (“Latinos, to be sure, are a bare majority of the voting-age population in new District 23, but only in a hollow sense, for the parties agree that the relevant numbers must include citizenship. This approach fits the language of § 2 because only eligible voters affect a group’s opportunity to elect candidates.”).

In the instant case, the plaintiffs argue that the district court erred in deeming District 1 of their illustrative plan an impermissible “influence district.” The rub in this case in that the district court extended the definition of an influence district beyond plaintiffs who could not form a majority-minority district to the situation at hand, where the plaintiffs have created a majority, but only by a trace number of individuals. African Americans comprise only 50.23% (775 persons) of the voting age population in the proposed District 1. Moreover, given the already-low population of African Americans in Glades County, the plaintiffs presented an illustrative district with only the slimmest of numerical majorities — a mere five persons. 4 However, although the question of whether a bare mathematical majority can still constitute an impermissible influence district apparently raises an issue of first impression for our circuit, I find the district court’s conclusion in the affirmative to be compatible with the spirit of Gingles and its progeny.

In support of its contention that a 50% numerical majority is sufficient, the majority cites only to eases from other circuits with distinguishable facts. In Valdespino v. Alamo Heights Indep. Sch. Dist., 168 F.3d 848 (5th Cir.1999), the Fifth Circuit applied a 50% bright-line rule, but where the minority population constituted only 47.9% of the voting age population in the proposed remedial district. Id. at 852-53. The majority is likewise on precarious ground in relying upon Cousin v. Sundquist, 145 F.3d 818 (6th Cir.1998). There, the Sixth Circuit not only rejected a sub-50% influence district as an appropriate § 2 remedy, but also found a proposed remedial district deficient where the district contained, similar to the case before us, only a bare majority. Id. at 829 *1271 (“Even in the one district where blacks constitute a voting age population majority, their 50.3% margin is so razor-thin that it does not meet the ‘safe district’ standards of courts that have approved race-conscious realignments in other electoral contexts.”). Pointing to Parker v. Ohio, 263 F.Supp.2d 1100 (S.D.Oh.2003) (three-judge court), aff’d mem., 540 U.S. 1013 , 124 S.Ct. 574 , 157 L.Ed.2d 426 (2003), furthermore, does not help the majority’s case. In Parker, the court merely reaffirmed what our own circuit has stated in the past — that influence districts are not cognizable remedies under the Voting Rights Act. Id. at 1105. The cases cited by the majority only point to the conclusion that influence districts are insufficient § 2 remedies. I would not dispute that conclusion. I would only dispute a mechanical “analysis” of the first Gingles requirement that would automatically find a remedy in a district with a 50.0% minority voting age population but no remedy in a district with a 49.9% minority voting age population. Such inflexibility, I posit, contravenes the searching factual analysis the Supreme Court set in place in Gingles .

Rote application of technical requirements does not comport with the motivating concerns of the Act. Section 2 jurisprudence suggests that Congress did not intend to create pure mathematical requirements for a vote dilution claim. See Perry, 126 S.Ct. at 2615 (“[I]t may be possible for a citizen voting-age majority to lack real electoral opportunity.”); see also De Grandy, 512 U.S. at 1008, 114 S.Ct. 2647 (finding the first Gingles condition satisfied where reasonably compact districts contained “a sufficiently large minority population to elect candidates of its choice”) (emphasis added); Voinovich v. Quilter, 507 U.S. 146, 158 , 113 S.Ct. 1149, 1157 , 122 L.Ed.2d 500 (1993) (“Of course, the Gingles factors cannot be applied mechanically and without regard to the nature of the claim.”) (emphasis added). Although a minority group may technically constitute a statistical majority for purposes of satisfying the first Gingles requirement, if the group does not have a realistic chance to elect its candidates of choice, it fails to present a viable § 2 remedy. To hold otherwise would open the gateway for minority plaintiffs to glean a remedial district from a sparse population that would be a mere simulacrum of a viable remedy. See Dillard, 376 F.3d at 1269 (“To open the door to the inevitable flood of marginal § 2 claims would impose an unwarranted burden on the lower federal courts and an indefensible encroachment into the affairs of state governments.”) (emphasis added). Even if a new District 1 could contain — which it cannot— the 59 remaining African Americans of voting age in Glades County, African Americans would comprise only 54.05% of the voting age population. 5 I cannot conclude that the district court erred in finding a mere five-person, 50.23% buffer ma *1272 jority to be simply a vestigial appendage to a legally insufficient influence district.

B.

The majority’s argument seems to gainsay the validity of the plaintiffs’ proposed District 1. Rather than resting the weight of their argument on the tenuous grounds of the mathematical majority, the majority, on the basis of a 19.2% countywide white crossover voting rate, argues that the white crossover vote in District 1 “should have been statistically significant in determining the sufficiency of the black minority’s ability to elect in District 1.”

In applying the crossover vote at the threshold Gingles stage, the majority quietly carves out a new rule of law for this circuit. The question of whether to apply crossover voting is an issue on which both we and the Supreme Court have recently withheld decision. Dillard, 376 F.3d at 1269 n. 7; see also Perry, 126 S.Ct. at 2624 .

We have yet to answer conclusively the question of whether minority plaintiffs may apply the crossover vote of a separate racial group in order to transform an influence district into a viable remedial district. We specifically declined to address the issue in the Dillard decision that held minority influence districts impermissible as a § 2 remedy: “We leave open the question of whether a § 2 plaintiff can pursue a ‘coalition’ or ‘crossover’ dilution claim, i.e., a claim where ‘members of the minority group are not a majority of the relevant voting population but nonetheless have the ability to elect representatives of their choice with support from a limited but reliable white crossover vote.’ ” Dillard, 376 F.3d at 1269 n. 7 (quoting Rodriguez v. Pataki, 308 F.Supp.2d 346, 376 (S.D.N.Y. 2004)); see also Perry, 126 S.Ct. at 2624 (arguing that if it were possible for influence districts to be a valid § 2 remedy, the plaintiffs must show that they constitute “a sufficiently large minority to elect their candidate of choice with the assistance of cross-over votes”) (quoting Voinovich, 507 U.S. at 158 , 113 S.Ct. at 1149 ). Given the unsettled character of the legal issue, I refuse to consider as clear error the district" court’s decision to decline to account for a speculative white crossover vote in District 1. See Anderson v. City of Bessemer City, N.C., 470 U.S. 564, 574 , 105 S.Ct. 1504, 1511 , 84 L.Ed.2d 518 (1985) (“Where there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.”) (citing United States v. Yellow Cab Co., 338 U.S. 338, 342 , 70 S.Ct. 177, 179 , 94 L.Ed. 150 (1949)).

The plaintiffs themselves, moreover, failed to raise white crossover voting at trial as a basis for the sufficiency of their proposed remedial district — as a result, the record pertaining to white crossover voting in the illustrative district is limited. It is unsurprising, therefore, that in both its order denying the plaintiffs relief after bench trial, as well as its order denying the plaintiffs’ motion to alter judgment, the district court purposefully declined to consider the white crossover vote in coalition with the African American vote in finding District 1 to be an influence district comprised of only the slimmest of literal majorities.

The majority also seems to assume that significant crossover voting would take place in the proposed remedial district. I would indulge in no such assumption. Were the court to adopt the plaintiffs’ plan-replacing at-large voting with single-member district voting — white voters *1273 who might otherwise vote for an African American candidate might not be inclined to do so in light of the fact that such bloc voting could grant African Americans a disproportionate representation on the County Commission and School Board 6 and result in district representatives who would have less incentive to be responsive to white residents in the rest of the county.

I cannot conclude that the district court clearly erred in declining to account for the white crossover vote in its analysis of whether the plaintiffs satisfied the first Gingles requirement. Neither the evidence in the record nor the relevant case law required the court to have addressed any white residents voting as a bloc with the plaintiff African Americans.

II

I turn now to the majority’s argument that the district court failed to set out with particularity its reasoning regarding the totality of the circumstances. Although it is certainly the case that we require a district court to state its reasoning with particularity, we have also held that where a district court possesses a correct understanding of the law, “engaged in a searching and meaningful evaluation of all the relevant evidence,” and the record supports the court’s findings, appellate review is satisfied. Solomon v. Liberty County Com’rs, 221 F.3d 1218, 1228 (11th Cir. 2000) (en banc) (quoting Southern Christian Leadership Conference v. Sessions, 56 F.3d 1281, 1293 (11th Cir.1995) (en banc)). I believe that the district court fully met that standard.

Even assuming that the plaintiffs satisfied the first Gingles requirement, I would still affirm the district court because there was no error in the court’s assessment of the totality of the circumstances. NAACP v. Fordice, 252 F.3d 361 , 374 (5th Cir.2001) (affirming the district court’s conclusion that no vote dilution existed where the district court found that the plaintiffs had satisfied the Gingles requirements, but not the totality of the circumstances test).

As the majority notes, in conducting a totality of the circumstances analysis, a court turn to several factors delineated in the Senate Report on the 1982 amendments to the Act. Gingles, 478 U.S. at 44-45 , 106 S.Ct. 2752 (citing S.Rep. No. 97-417 (1982)). In its 49-page order, the district court went through each of the specific factors, 7 and concluded that the totality of the circumstances failed to support the plaintiffs’ vote dilution claim.

Upon reviewing the record, I cannot say that the district court clearly erred in the *1274 method of its analysis. In reaching its conclusion, the court reviewed the testimony of various experts on the racial history of Florida, particularly with regard to changes to the election process, 8 statistical analyses of Glades County voter behavior, and the background underpinning the plaintiffs’ illustrative plan. See Anderson, 470 U.S. at 575 , 105 S.Ct. at 1512 (“When findings are based on determinations regarding the credibility of witnesses, Rule 52(a) demands even greater deference to the trial court’s findings; for only the trial judge can be aware of the variations in demeanor and tone of voice that bear so heavily on the listener’s understanding of and belief in what is said.”); see also De Grandy, 512 U.S. at 1011 , 114 S.Ct. 2647 (“[T]he ultimate conclusions about equality or inequality of opportunity were intended by Congress to be judgments resting on comprehensive, not limited, canvassing of relevant facts.”). Moreover, the district court reviewed relevant census data, including data relating to the socio-economic disparities between Caucasian and African American households. 9 When considering the historical patterns of election for African Americans; the court noted that only three African Americans had ever sought elected office in Glades County, two of whom had made successful bids. 10 Furthermore, the court invested substantial energy in reviewing several depositions. 11 Those depositions revealed little evidence of racial discrimination or racial appeals during campaigning. Indeed, the testimony of plaintiff Billie Thompson suggests that the opposite is true: despite the vast acreage of Glades County, she campaigned in every major voting area (covering 90% of the county), attended (and was well-received) at the only political rally for the primary, and expressed the opinion that she had a “small but effective” campaign, a fact supported by her having garnered 40% of the vote countywide.

Also, the district court properly considered the special referendum held on June 5, 2001, which Dr. Steven Cole, a plaintiffs’ expert, conceded did not reflect racially polarized voting. In precincts with a 90% or greater African American voting age population, only 52% voted in favor of single-member districts. We have previously considered similar referenda as relevant to the question of whether policies underlying electoral procedures are tenuous, as well as the responsiveness of elected officials to the unique needs of a minority population. See Solomon v. Liberty County Commis *1275 sioners, 221 F.3d 1218 , 1234 (11th Cir. 2000) (noting that a majority of voters in Liberty County, including 60% of African Americans, voted for at-large elections and against single-member districts for school board elections); id. (“Obviously, the fact that black and white voters in Liberty County voted overwhelmingly against single-member districts, tends to undercut any suggestion that the continued policy of maintaining at-large elections is somehow discriminatory or otherwise tenuous.”) (quoting Solomon v. Liberty County, Florida, 957 F.Supp. 1522, 1568 (N.D.Fla. 1997)).

The Supreme Court, furthermore, has agreed that proportionality 12 is a relevant consideration under a § 2(b) totality of the circumstances analysis. Perry, 126 S.Ct. at 2614 ; De Grandy , 512 U.S at 1006-07, 114 S.Ct. 2647 . Case law subsequent to Gingles counsels attention to “whether the number of districts in which the minority group forms an effective majority is roughly proportional to its share of the population in the relevant area.” De Grandy, 512 U.S. at 1006-07 , 114 S.Ct. 2647 ; see Baird v. Consolidated City of Indianapolis, 976 F.2d 357, 358 (7th Cir.1992) (“The district court’s analysis took into consideration the fact that the African-American voting age population, 10% of the relevant voting population (and apparently on a diminishing trend), ought to have the opportunity to elect effectively 20% of the School Board and County Commission (one of the five elected members for each group). 13 Here, 10% of the electorate seeks one out of each of the five seats available on the County Commission and School Board— political representation a full 100% above its percentage among the voting age population of Glades County.”) See Perry, 126 S.Ct. at 2662 (C.J.Roberts, dissenting) (“Latino voters enjoy effective political power 46% above their numerical strength, or, even disregarding District 25 as an opportunity district, 24% above their numerical strength .... Surely these figures do not suggest a denial of equal opportunity to participate in the political process.”); De Grandy, 512 U.S. at 1017 n. 13, 114 S.Ct. at 2660 n. 13 (“When 40 percent of the population determines electoral outcomes in 7 out of 10 districts, the minority group can be said to enjoy effective political power 75 percent above its numerical strength.”). Indeed,’the facts of the instant case are difficult to massage into a vote dilution claim where a 10.1% countywide voting age population of African Americans would seek 20%, respectively, of the five-seat County Commission and School Board. Ultimately, the legisla *1276 tive history of the Voting Rights Act counsels that the purpose of a § 2 remedy is not to ensure a windfall for a minority population. See De Grandy, 512 U.S. at 1017 , 114 S.Ct. at 2660 (“One may suspect vote dilution from political famine, but one is not entitled to suspect (much less infer) dilution from mere failure to guarantee a political feast. However prejudiced a society might be, it would be absurd to suggest that the failure of a districting scheme to provide a minority group with effective political power 75 percent above its numerical strength indicates a denial of equal participation in the political process. Failure to maximize cannot be the measure of § 2.”). Upon review of the record, I cannot find clear error in the district court’s determination that the totality of the circumstances does not support a vote dilution claim in Glades County. 14

III.

For the aforementioned reasons, I would conclude that the district court did not clearly err in finding that District 1 of the plaintiffs’ illustrative plan constituted an influence district insufficient as a remedy under § 2 of the Voting Rights Act. I would likewise conclude that the court did not clearly err in finding that the totality of the circumstances did not support the plaintiffs’ vote dilution claim. In making those determinations, the court properly applied the relevant legal principles and grounded its findings in an accurate understanding of the law. I therefore dissent.

. The Supreme Court declined to decide the issue of whether "influence districts” could constitute a proper § 2 remedy under the first Gingles requirement in, e.g., De Grandy, 512 U.S. at 1008-09 , 114 S.Ct. at 2656 ; Voinovich, 507 U.S. at 154 , 113 S.Ct. at 1149 ; Growe v. Emison, 507 U.S. 25 , 41 n. 5, 113 S.Ct. 1075 , 1084 n. 5, 122 L.Ed.2d 388 (1993); Gingles, 478 U.S. at 46 -47 n. 12, 106 S.Ct. at 2764 n. 12.

. Justice Kennedy wrote the Peny opinion, and delivered the opinion of the Court with respect to Parts II-A (constitutionality of a *1269 partisan gerrymander) and III (vote dilution claim of Hispanics); a plurality opinion with respect to Part I (factual and procedural history) and Part IV (vote dilution claim of African Americans) (joined by the Chief Justice and Justice Alito); an opinion with respect to Parts II-B (political ramifications of congressional redistricting) and II-C (constitutionality of partisan gerrymander); and a plurality opinion with respect to II-D (constitutionality of mid-decade redistricting) (joined by Justices Souter and Ginsburg).

. With regard to the influence district issue, Chief Justice Roberts and Justice Alito joined Justice Kennedy in deeming no § 2 remedy to be available where African-Americans were 25.7% of the citizen voting age population in District 24, the district at issue. The plaintiffs in the case contended that African Americans nonetheless had control ("influence”) over the district, and that a recent redistricting plan diluted the African American vote by removing District 24. Perry, 126 S.Ct. at 2624 (plurality opinion).

For his part, Justice Souter argued that Perry ought to have been the case by which the Court resolved the confusion surrounding influence districts in the Court’s vote dilution jurisprudence. See id. at 2648 n. 1 (“Although both the plurality today and our own prior cases have sidestepped the question whether a statutory dilution claim can prevail without the possibility of a district percentage of minority voters above 50%, ... the day has come to answer it.”). Justice Souter agreed with the plaintiffs' argument that influence districts ought to be viable as a § 2 remedy under the first Gingles requirement by way of analogy to section 5 of the Act, which applies where a political change has the purpose or effect of “denying or abridging the right to vote,” 42 U.S.C. § 1973 (c). Id. at 2648 , 2648 n. 3. Specifically, he would find the first Gingles requirement satisfied where a minority population was a majority in the electorate’s primary or through crossover voting. Id. at 2648.

. If illustrative District 1 were to replace five of the African Americans with five white voting age persons, the African American voting age population would drop to 49.9%.

In a footnote aside, the majority also finds that "the district court also erred in its determination that District 1 would contain only a 50.23% majority because it counted as ‘black’ only non-Hispanic blacks, omitting persons who self-identified on the 2000 Census as both black and Hispanic.” With that in mind, the majority posits that "there may be significantly more than a 50.23% black majority in District 1."

The 2000 Census, however, suggests that the majority’s line of argument may rest on a red herring. The 50.23% majority in the proposed District 1 was calculated by William S. Cooper, the plaintiffs’ expert witness, who prepared statistics under the definition of African Americans as "non-Hispanic blacks.” Although the record does not indicate how many individuals — if any — would self-identify as both Hispanic and black in District 1, the 2000 Census indicates that in Glades County as a whole, there were 834 voting-age individuals who identified themselves — in whole or in part — as black, and 830 voting-age individuals who identified themselves as non-Hispanic black. As a result, there were at most four individuals county-wide who may have been excluded by the calculations presented by Cooper. I find it difficult to presume that the inclusion of any persons who may have self-identified as both black and Hispanic would have resulted in "significantly more than a 50.23% majority” in the plaintiffs’ proposed remedial district.

. The 54.05% calculation (834/1,543) that would include all African Americans of voting age within a single proposed majority-minority district is itself far too generous. The geographical distribution of African Americans in Glades County precludes an easy redistricting to provide for a more substantial majority. According to the illustrative plan, District 1 is bordered by District Two and District Four. Districts Two and Four have a combined African American voting age population of 37 individuals (33 in District Two and 4 in District Four). Therefore, while still improbable, even were those 37 African American voting age individuals to live close enough to the borders of the proposed District 1 to be subsumed within a new majority-minority district, the new calculation (812/1,543) results in only a 52.62% majority voting age population for African Americans.

. Although African Americans represent 10% of the Glades County electorate, they seek 20% (one out of five) of the seats available on the County Commission and School Board.

. In the course of its discussion, the district court found: (1) no evidence that the ability of African Americans in Glades County to participate in the political process continued to be hindered by historical discrimination; (2) no evidence of overt appeals to race during the election process; (3) that only the majority vote requirement, and possibly the size of Glades County, might enhance the potential for discrimination against African Americans; (4) no candidate slating process in Glades County; (5) no socio-economic conditions that severely impair equal footing for African Americans in the political process; (6) no evidence of racial appeals in political campaigns; (7) "some electoral success” by African American candidates for public office; (8) responsiveness to the needs of the African American community; (9) no evidence of tenuous election practices; (10) no districts that contain a majority of minority voting age persons; and (11) that African Americans had not yet achieved representation proportional to their approximately 10% share of the countywide population.

. Mormino found that both the 1947 legislation adopting statewide at-large school board primaries as well as 1900 constitutional amendment adopting at-large elections for county commissioners were motivated by racially discriminatory intent.

. The district court noted that the per capita income for African Americans in Glades County was $6,130, while the per capita income for Caucasians was $12,012. Likewise, while 39.9% of Caucasians over the age of 25 had not completed high school, the same was true for 79% of African Americans.

. Charles Hall was elected to the County Commission in 1976; Beamon Rich was elected to the City Council for Moore Haven in 1998, and plaintiff Billie Thompson made an unsuccessful bid for Glades County School Board in 1998.

.The court reviewed the depositions of: Janet Storey (School Board District 1), Catherine Baxter Peeples (School Board District 2), Tom Gaskins (School Board District 3), Mike Pressley (School Board District 4), Susan T. Shriveler (School Board District 5), K.S. Jones (County Commission District 1), Alvin Ward (County Commission District 2), Franklin Simmons (County Commission District 4), and Holly Whiddon Green (Supervisor of Elections).

. The question of whether the number of majority-minority districts is "roughly proportional" to a minority group's share of the overall population is distinct from the question of whether a minority group is guaranteed proportionate representation. The Act itself resolves the latter question, noting that "nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.” 42 U.S.C. § 1973 (b) (emphasis added). Although the Act precludes any formal guarantee of electoral success to a minority group, § 2 permits an examination of proportionality as an appropriate inquiry into equality of opportunity. See De Grandy, 512 U.S. at 1015 n. 11, 114 S.Ct. 2647 (“[T]he ultimate right of § 2 is equality of opportunity, not a guarantee of electoral success for minority-preferred candidates of whatever race.”).

. We note that the Supreme Court has made clear that a federal court cannot increase the size of elected bodies as a part of remedying a § 2 violation. See Holder v. Hall, 512 U.S. 874, 881 , 114 S.Ct. 2581, 2586 , 129 L.Ed.2d 687 (1994) (“There is no principled reason why one size [of an elected governing body] should be picked over another as the benchmark for [determining whether vote dilution exists].”); Nipper v. Smith, 39 F.3d 1494, 1532 (11th Cir.1994) (same).

. The majority would remand this case for reconsideration of the totality of the circumstances in light of its holding that the district court erred in the threshold Gingles inquiry. If we are to accept the majority’s holding, however, that the district court ought to have applied the Hispanic crossover vote to the African American vote, we are confronted by the awkward question of whether the Hispanics ought not be considered in the totality of the circumstances analysis. Because the plaintiffs failed to provide expert testimony with regard to the Hispanic population, such a question would prove difficult to weigh.