NELSON, Circuit Judge:
Dolores Cruz Gomez, Patricia Leal, and Waldo Rodriguez (“Appellants”) challenge the district court’s ruling that the City of Watsonville’s at-large mayoral and city council election system does not violate Section 2 of the Voting Rights Act as amended in 1982. 42 U.S.C. § 1973. The district court had jurisdiction pursuant to 28 U.S.C. § 1343(a)(3) & (4). We have jurisdiction pursuant to 28 U.S.C. § 1291. We reverse and remand for implementation of a plan that comports with Section 2.
PROCEDURAL AND FACTUAL BACKGROUND
Appellants, Mexican-American citizens of Watsonville (“the City”) eligible to vote, brought suit for declaratory and injunctive relief under 42 U.S.C. § 1973 (“Section 2”), 42 U.S.C. § 1983, and the fourteenth and fifteenth amendments. They have not appealed the district court’s finding that no constitutional or § 1983 violations occurred. The appellants instead claim that the City’s at-large system of mayoral and city council elections violates Section 2 by lessening the opportunity of Hispanics to participate in the political process and to elect representatives of their choice. Appellants seek implementation of a single-member districting plan to redress the alleged violation, and attorneys’ fees and costs pursuant to 42 U.S.C. §§ 1973i(e) & 1988.
Appellants filed suit on May 21, 1985. On November 3, 1986, they filed a motion to enjoin the mayoral and city council election scheduled for May 1987. The district court denied this motion and held a trial from January 20-26, 1987. The district court found that racially polarized voting exists in Watsonville. The court did not find, however, that the Hispanic community was sufficiently politically cohesive or geographically insular to meet the Section 2 test. The district court issued a decision in favor of the City on January 30, 1987, and affirmed that decision after further briefing on February 20, 1987. The court awarded attorneys’ fees and costs to the City. Appellants timely appealed.
The parties agree on many of the facts in this case. Appellants primarily challenge the district court’s application of the legal standards governing Section 2 claims. The parties stipulated to the following facts, largely drawn from 1950-80 census data.
The City is governed by the city council, comprised of six council members and a mayor. Persons are elected to these positions on an at-large basis, with three council members elected each odd-numbered year. The three candidates with the highest number of votes are elected. Before the citizens voted for an at-large system in November 1952, the City used a district or ward system.
According to the 1980 census, 48.9% of the City’s population is Hispanic, 45.2% Anglo, 5.4% Asian and Pacific Islander, and 0.5% Black. In 1950, Hispanics comprised 1,001 persons in a total population of 11,-373. As of 1980, 11,509 of Watsonville's 23,543 residents are Hispanic. Of those persons eighteen years and older, 40% are Hispanic and 60% non-Hispanic. However, Hispanics comprise only 37.0% of Watson-ville’s citizens because 41.5% of the Hispanics in Watsonville are non-citizens.
No Hispanic had ever been elected as mayor or city council member in Watson-ville under the at-large system prior to the trial. Eight Hispanic candidates ran unsuccessfully for city council positions from 1971 to 1985 and one Hispanic ran for mayor in 1979.1 Twenty-five of the 51 [1410]*1410non-Hispanic candidates ran successfully for city council positions from 1971 to 1985. Hispanic persons have been appointed to City boards and commissions.
According to census data, many more non-Hispanics than Hispanics are employed in managerial and professional jobs in Wat-sonville. A fairly equal number of Hispanics and non-Hispanics are employed as operators, fabricators and laborers. Many more Hispanics than non-Hispanics work in the farming, forestry and fishing industries in Watsonville. The number of Hispanics employed by the City itself has increased by 20% since 1973.
As of 1980, educational attainment levels differed among Hispanics and Anglos. Below are the number of persons 25 years and older completing school in the City.
ANGLO HISPANIC
Elementary 2,739 3,032
High School (1-3 yrs.) 1,564 569
(4 yrs.) 2,792 610
College (1-3 yrs.) 1,728 311
(4 or more yrs.) 1,084 171
Although the parties stipulated to the above facts, they dispute the significance of socioeconomic characteristics of the City's Hispanic community derived from the census data. Appellants maintain that the evidence reveals large discrepancies between the City’s Hispanics and non-Hispanics in educational levels, income, and employment. Appellants stress that the Hispanic population has lower educational attainment levels; are employed in lower status professions, or more commonly are unemployed, and have lower family incomes; and rent instead of own their homes twice as often as other City residents. Appellants also emphasize that of the Watson-ville Hispanics over age 5, 41% speak Spanish at home and approximately one-third do not speak English well or at all.
Appellees interpret the same statistics as showing less disparity between the two groups. They emphasize the increasing level of education among Hispanics during the 1970s and the increasing level of employment of Hispanics by the City itself in recent years. However, appellees fail to compare those figures to the education and employment levels of the City’s non-Hispanics. Appellees also emphasize disparities within the City’s Hispanic population. For example, they note that lower and higher income Hispanics sometimes reside in the same census tracts, and that the Hispanics who do own homes own those of comparable value to those owned by non-Hispanics.
The district court here found that racially polarized voting exists in Watsonville. It relied on appellants’ expert, Dr. Bernard Grofman, who analyzed the census data and concluded that voting is racially polarized in Watsonville. Cf. Thornburg v. Cringles, 478 U.S. 30, 52-61, 106 S.Ct. 2752, 2767-72, 92 L.Ed.2d 25 (1986) (relying on similar testimony from Dr. Grofman in upholding the trial court’s finding of legally significant racially polarized voting). The district court found the evidence of racially polarized voting here “essentially uncontra-dicted” and found Dr. Grofman’s methodology “completely without criticism.”
However, the court ruled against appellants. It found Watsonville’s Hispanic population insufficiently geographically compact to meet the requirements of a Section 2 claim. The district court recognized that the Hispanics have the potential to control two single-member districts, but rejected appellants’ vote dilution claim because the majority of Hispanics would still reside in Anglo-controlled districts in which their vote was ineffective.
The district court also found that appellants failed to demonstrate sufficient political cohesiveness among the Watsonville Hispanics. Although the court found that 95% of Hispanic voters vote for the same candidates, the court considered low voter registration and turnout among Hispanics, and concluded that all Hispanics eligible to vote might not all vote alike. Further, the court found that socioeconomic differences and differences in political opinion supported a conclusion that Hispanics had shown low enthusiasm for past Hispanic candidates and this, in turn, undermined political cohesiveness.
The district court appears to have misunderstood the proper legal inquiry after the 1982 amendments. Because this circuit [1411]*1411has not yet decided a case interpreting the 1982 amendments to Section 2 of the Voting Rights Act, we will begin with an analysis of the legal standards that are to be applied to challenges to at-large electoral schemes under the newly amended Section 2.
STANDARD OF REVIEW
In analyzing a Section 2 claim, we review the district court’s factual findings under the clearly erroneous standard of Fed.R. Civ.P. 52(a). Thornburg v. Gingles, 478 U.S. 30, 79, 106 S.Ct. 2752, 2781, 92 L.Ed.2d 25 (1986). However, “Rule 52(a) ‘does not inhibit an appellate court’s power to 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, 501, 104 S.Ct. 1949, 1959, 80 L.Ed.2d 502 (1984)). Accordingly, the district court’s findings will be set aside to the extent that they rest upon an erroneous view of the law. Pullman-Standard v. Swint, 456 U.S. 273, 287, 102 S.Ct. 1781, 1789, 72 L.Ed.2d 66 (1982). When findings have been set aside on this basis, the normal procedure is to remand the case so that the district court can redetermine the issue using the correct legal standard. See id. at 292, 102 S.Ct. at 1792. However, a remand is not necessary where the record on appeal “permits only one resolution of the factual issue.” Id. at 287, 102 S.Ct. at 1789.
ANALYSIS
I. SECTION 2 VIOLATIONS
In 1982, Congress amended Section 2 of the Voting Rights Act of 1965. The amendments were designed to repudiate the “intent test” described in the plurality opinion in City of Mobile v. Bolden, 446 U.S. 55, 100 S.Ct. 1490, 64 L.Ed.2d 47 (1980), which stated that an electoral scheme could not be challenged under Section 2 absent a showing that the scheme was intentionally designed or maintained for a discriminatory purpose. The 1982 Amendments replace the intent test with a “results test.” Under the new test, a plaintiff in a Section 2 case must show that, based on the totality of the circumstances, the electoral process is “not equally open to participation by the members of a [racial or language minority] in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 42 U.S.C. § 1973.2
The Senate Judiciary Committee Report accompanying the act that amended Section 2 stated that the impact of a challenged electoral device should be judged “on the basis of objective factors.” S.Rep. No. 417, 97th Cong., 2d Sess. 27, reprinted in 1982 U.S.Code Cong. & Admin.News 177, 205. The report listed a number of objective factors that might be probative of a violation under the “totality of the circumstances test”:
1. The extent of any history of official discrimination in the state or political subdivision that touched the right of the [1412]*1412members of the minority group to register, to vote, or otherwise to participate in the democratic process;
2. the extent to which voting in the elections of the state or political subdivision is racially polarized;
3. the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, anti-single-shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority group;
4. if there is a candidate slating process, whether the members of the minority group have been denied access to that process;
5. the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process;
6. whether political campaigns have been characterized by overt or subtle racial appeals;
7. the extent to which members of the minority group have been elected to public office in the jurisdiction.
Id. at 28-29, 1982 U.S.Code Cong. & Admin.News at 206-07.3
The Report emphasized, however, that this list of factors was not a mandatory seven-pronged test; the list was only meant as a guide to illustrate some of the variables that should be considered by the court. As stated in the Report, “there is no requirement that any particular number of factors be proved, or that a majority of them point one way or the other.” Id. at 29, 1982 U.S.Code Cong. & Admin.News at 207; accord id. at 29 n. 118, 1982 U.S.Code Cong. & Admin.News at 207 n. 118 (“[T]he Committee [does not] intend [that these factors] be used as a mechanical ‘point counting’ device. The failure of plaintiff to establish any particular factor is not rebuttal evidence of [no violation].”).
The Senate Committee also noted that, while the basic “totality of the circumstances” test remains the same, the range of factors that would be relevant in any given case will vary depending upon the nature of the claim and the facts of the case. See id. at 28, 1982 U.S.Code Cong. & Admin. News at 206 (“To establish a violation, plaintiffs could show a variety of factors, depending upon the kind of rule, practice, or procedure called into question.”); see also id. at 30, 1982 U.S.Code Cong. & Admin.News at 207 (noting that the proof sufficient to sustain a challenge based upon a series of events or episodes “would not necessarily involve the same factors” that would be relevant in a challenge to a permanent structural barrier).
The Supreme Court first interpreted the 1982 amendments to Section 2 in Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986). Thornburg involved a challenge to the redistricting plan adopted by the North Carolina General Assembly. Several black citizens challenged the plan on the grounds that, inter alia, the plan’s use of large, multimember voting districts had the effect of impermis-sibly diluting the voting strength of black voters, in violation of Section 2. The district court found that the use of multimem-ber districts constituted impermissible vote dilution, and enjoined the state from conducting elections under the plan. The state appealed this judgment to the Supreme Court.
The Supreme Court began by noting that, “[w]hile many or all of the factors listed in the Senate Report may be relevant to a claim of vote dilution through submergence in multimember districts,” only certain factors were “essential ” to a successful claim of vote dilution. 478 U.S. at 48 & n. 15, 106 S.Ct. at 2766 & n. 15. The Court cited many of the provisions of the Report quoted above, emphasizing that the Senate [1413]*1413“list of typical factors is neither comprehensive nor exclusive.” Id. at 45, 106 S.Ct. at 2764. The Court noted that, rather than applying the factors in a mechanical fashion, courts must judge Section 2 claims based on a “searching practical evaluation of the ‘past and present reality’ and on a ‘functional’ view of the political process.” Id.
The Court stated that a plaintiff challenging a multimember district plan under Section 2 must show three things. 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.” Id. at 50, 106 S.Ct. at 2766. The Court reasoned that, unless the minority group could constitute a majority in a single-member district, there is no sense in which “the multi-member form of the district” is responsible for any inability of the minority group to participate equally. Id. (emphasis in original). That is, “[u]nless 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.” Id. at 50 n. 17, 106 S.Ct. at 2767 n. 17. Second, “the minority group must be able to show that it is politically cohesive.” The Court stated that, unless the minority group is politically cohesive, “it cannot be said that the selection of a multi-member electoral structure thwarts distinctive minority group interests.” Id. at 51, 106 S.Ct. at 2767. 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.” Id. at 50-51, 106 S.Ct. at 2766-67 (citations omitted). This showing of racial bloc voting establishes the required casual link between the use of a multimember district and the inability of the minority group “to elect its chosen representatives.” Id. at 51, 106 S.Ct. at 2767. Furthermore, by showing that the majority is usually able to defeat the minority’s candidates, the minority group is able to distinguish a “structural dilution from the mere loss of an occasional election.” Id.
The Supreme Court noted that its analysis put substantially greater emphasis on some of the Senate factors than on others. The Court stated that the most important Senate factors in a case involving a challenge to a multimember electoral scheme are (1) “the ‘extent to which minority group members have been elected to public office in the jurisdiction’ ” and (2) “the ‘extent to which voting in the elections of the state or political subdivision is racially polarized.’ ” 478 U.S. at 48 n. 15, 106 S.Ct. at 2766 n. 15. These two factors overlap substantially with the last of the three elements which the Court stated must be proved. The Court went on to note that while the presence of the other Senate factors might be supportive of a challenge to a multimember voting scheme, they were “not essential to ” such a claim. Id. (emphasis in original). The Court stated that, by recognizing that some of the Senate factors are more important to a multimember district vote dilution claim than others, it was simply “effectuating] the intent of Congress.) Id.
The Court then applied this test to the facts before it, and concluded that, under the applicable clearly erroneous standard, the district court’s finding of a Section 2 violation must be affirmed, except with respect to one district in which there had already been sustained black electoral success. Id. at 77-80, 106 S.Ct. at 2780-82.
As the district court in this case recognized, Watsonville’s at-large scheme is the functional equivalent of the electoral scheme at issue in Gingles. Having enunciated the legal standards spelled out in the Senate Report and in Gingles, we may now apply these standards to the facts of this case.
II. APPLICATION OF GINGLES
A. Geographic Concentration of Hispanics
The first element of a Section 2 challenge to an at-large or other multimember voting scheme is a showing that the minority group “is sufficiently large and [1414]*1414geographically compact to constitute a majority in a single-member district.” Gingles, 478 U.S. at 50, 106 S.Ct. at 2767. Accordingly, we use the single-member district as the measure in evaluating whether Watsonville’s at-large procedure results in vote dilution. See id. at 50 n. 17, 106 S.Ct. at 2767 n. 17 (“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.”).
Dr. Grofman demonstrated that 68.3% of the City’s Hispanic population resides in three of nine census tracts. He analyzed the number of Hispanics eligible to vote, considering age and citizenship requirements, and concluded that they are compact geographically. Appellees have not refuted these statistical findings on appeal. The district court found, however, that Watsonville Hispanics are not sufficiently geographically insular because appellants’ proposed plan includes only two districts that would have contained a majority of Hispanic voters. The district court reasoned that because the plan would alleviate vote dilution for only one-third of the Hispanics eligible to vote, the appellants had not shown geographic insularity.
The district court erred in considering that approximately 60% of the Hispanics eligible to vote in Watsonville would reside in five districts outside the two single-member, heavily Hispanic districts in appellants’ plan. Districting plans with some members of the minority group outside the minority-controlled districts are valid. See, e.g., Ketchum v. City Council, 630 F.Supp. 551, 557 (N.D.Ill.1985); Gingles v. Edmisten, 590 F.Supp. 345, 357-59 (E.D.N.C.1984), aff'd in relevant part sub nom Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986). As the Fifth Circuit stated in Campos v. City of Baytown, Texas, 840 F.2d 1240, 1244 (5th Cir.1988): “The fact that there are members of the minority group outside the minority district is immaterial. All that is required is that the minority group be ‘sufficiently large and geographically compact to constitute a majority in a single member district.’ ” (quoting Gingles, 478 U.S. at 50, 106 S.Ct. at 2766 (emphasis added by Fifth Circuit)). The fact that the proposed remedy does not benefit all of the Hispanics in the City does not justify denying any remedy at all.
The appellants’ plan proposes two districts in which Hispanics would constitute a majority of the voters and would be able to elect representatives of their choice. It is sadly ironic that the district court concluded that because many Hispanic voters would still not be able to elect representatives of their choice under the proposed plan, no Section 2 claim could be maintained, thereby relegating all Hispanic voters to having no political effectiveness. The district court’s finding is premised on a misunderstanding of the applicable legal standard. The Hispanics in Watsonville are capable of constituting a majority in at least one district, and therefore the court’s finding that the group is not sufficiently geographically compact is clearly erroneous. Appellants have demonstrated the first element of the Gingles test.
B. Political Cohesiveness of Hispanics
The second requirement of the Gingles test is that appellants demonstrate that Watsonville Hispanics are politically cohesive. The district court found that 95% of the Hispanic voters in heavily Hispanic precincts support Hispanic candidates and that Hispanic voters ranked Hispanic candidates first or tied for first. Although the district court here found that “in actualpty] those Hispanics who have voted have tended to vote for the same Hispanic candidates,” the district court nonetheless concluded that the Hispanic community is not politically cohesive.
At the trial below, Dr. Grofman testified that Watsonville Hispanics voted the same way in substantial proportions in those elections analyzed. Dr. Grofman testified that this racial bloc voting is the primary factor in evaluating political cohesiveness, but he also pointed to other socioeconomic and cultural differences between Hispanics and non-Hispanics that contribute to the [1415]*1415political cohesiveness of Hispanics in the City. He also cited evidence of single-shot voting (where minority members vote only for certain candidates and do not use their remaining votes) among Watsonville Hispanics to reinforce his conclusion.
Appellees argued below that Watsonville Hispanics are not politically cohesive and that appellants are not representative of Hispanics in Watsonville. The appellee’s expert, Dr. Peter Morrison, and two lay Hispanic witnesses testified to the existence of political differences within the Wat-sonville Hispanic community. Dr. Morrison concluded that he would “not expect ... to find all Hispanics in Watsonville voting alike.” He based this projection on socioeconomic differences among Hispanics and the failure of many eligible Hispanics to register and/or vote. Appellee’s lay witnesses, whom the district court found credible, enumerated reasons why they believed that members of the Hispanic community did not always support Hispanic candidates enthusiastically.
In reviewing all of this evidence, the district court concluded that, with respect to those Hispanics who have actually voted, the evidence favored a finding of political cohesiveness. Nonetheless, the court concluded that, because “no significant number of eligible Hispanics have voted in the elections under consideration,” the Hispanic community as a whole was too apathetic to be politically cohesive.
The district court’s finding on this issue is based on a misunderstanding of what is meant by “political cohesiveness.” The inquiry is essentially whether the minority group has expressed clear political preferences that are distinct from those of the majority. Thus, a “showing that a significant number of minority group members usually vote for the same candidates is one way of proving the political cohesiveness necessary to a vote dilution claim.” Gingles, 478 U.S. at 56, 106 S.Ct. at 2769. Indeed, the Court recognized that one of the reasons for focusing on the Senate factor concerning “racially polarized voting” is that proof that the minority has consistently voted differently helps one “to ascertain whether minority group members constitute a politically cohesive unit.” Id. The Supreme Court’s definition of “political cohesiveness” is further elucidated by referring to the articles that the Court cites as its primary source for the concept. In one passage cited by the Court, see id. at 51, 56, 106 S.Ct. at 2767, 2770, Blacksher and Menefee state that “[wjhether a racial group is politically cohesive depends on its demonstrated propensity to vote as a bloc for candidates or issues popularly recognized as being affiliated with the group’s particularized interests.” Blacksher & Me-nefee, From Reynolds v. Sims to City of Mobile v. Bolden: Have the White Suburbs Commandeered the Fifteenth Amendment?, 34 Hastings L.J. 1, 59 (1982) (emphasis added).
Accordingly, under the “functional” approach mandated by Section 2, see Gingles, 478 U.S. at 48, 106 S.Ct. at 2765 (quoting S.Rep. at 30 n. 120, 1982 U.S.Code Cong. & Admin.News at 208 n. 120), the issue of political cohesiveness is to be judged primarily on the basis of the voting preferences expressed in actual elections. This conclusion finds additional support in the decisions of other circuits. In Collins v. City of Norfolk, Va., 816 F.2d 932, 935 (4th Cir.1987), the Fourth Circuit stated that “the existence of racially polarized voting ... establishes both cohesiveness of the minority group and the power of white bloc voting to defeat the minority’s candidates.” The court then emphasized that “[t]he legal standard for the existence of racially polarized voting looks only to the difference between how majority votes and minority votes were cast.” Id. (emphasis added). Similarly, in Campos, the Fifth Circuit noted that the determination as to whether a minority is politically cohesive “is not an inquiry to be made prior to and apart from a study of polarized voting, ... because the central focus is on voting patterns.” 840 F.2d at 1244. After quoting from Gingles, the Fifth Circuit concluded that “[i]t follows that a minority group is politically cohesive if it votes together.” Id.
Applying these standards to the district court’s decision, it is clear that the [1416]*1416court applied the wrong legal standard. The district court erred by focusing on low minority voter registration and turnout as evidence that the minority community was not politically cohesive. The court should have looked only to actual voting patterns rather than speculating as to the reasons why many Hispanics were apathetic. In fact, there is nothing in the record to support the district court’s apparent conclusion that lack of enthusiasm for Hispanic candidates was responsible for the low rates of voter registration among Hispanics. See United States v. Dallas County Comm’n, 739 F.2d 1529, 1536 (11th Cir.1984).4
The district court expressly found that predominantly Hispanic sections of Watsonville have, in actual elections, demonstrated near unanimous support for Hispanic candidates. This establishes the requisite political cohesion of the minority group.
The district court’s reliance on socioeconomic disparities and differences of political opinion within the Hispanic community as support for Dr. Morrison’s projection was also erroneous. Such differences within the community are only relevant to the extent that they reflect differences in voting behavior among Hispanics. Here, however, the district court found that Hispanics vote alike, coalescing behind Hispanic candidates. Given this degree of cohesion in their voting behavior, the fact that Watsonville Hispanics differ amongst themselves along many other dimensions is irrelevant.
This conclusion is supported by two recent decisions of the Fifth Circuit. In Campos, 840 F.2d at 1245-47, and League of United Latin American Citizens v. Midland Independent School Dist., 812 F.2d 1494, 1500 (5th Cir.), vacated and aff'd on other grounds, 829 F.2d 546 (1987), the Fifth Circuit concluded that groups of Blacks and Hispanics in certain cities were politically cohesive for purposes of Section 2 violations. Thus, despite the many ethnic and cultural differences between Blacks and Hispanics, the two groups were found to have “political goals that are inseparable.” Midland, 812 F.2d at 1500. The defendants in Midland offered a survey indicating that Blacks and Hispanics have mutually exclusive interests and are politically distinct groups. The Fifth Circuit, however, emphasized that “the survey does not prove that the two minority groups would vote differently in any particular election in which a member of one of their groups is a candidate against [a white person].” Id. at 1501 (emphasis added). These decisions teach that a finding that Hispanics overwhelmingly vote for different candidates than do non-Hispanics, as was found by the district court here, is ample support for political cohesiveness.
In the face of the actual voting patterns found by the district court, its finding of noncohesiveness, based on Dr. Morrison’s projection that not all Hispanics are likely to vote alike and on anecdotal lay testimony about the differences within the Hispanic community, was clearly erroneous. Appellants have established the second element of the Gingles test.
C. Racial Bloc Voting Among Majority
Finally, to maintain a Section 2 claim appellants must demonstrate racial bloc voting among the majority. “[A] white bloc vote that normally will defeat the combined strength of minority support plus white ‘crossover’ votes rises to the level of legally significant white bloc voting.” Gingles, 478 U.S. at 56, 106 S.Ct. at 2770; accord Campos, 840 F.2d at 1248-49.
[1417]*1417As noted above, 60% of Watsonville residents over age 18 are non-Hispanic and 41.5% of the Hispanies in Watsonville are non-citizens. The district court found that only an average of 13% of the voters in predominantly Anglo precincts vote for Hispanic candidates while 95% of the Hispanic voters in heavily Hispanic precincts support Hispanic candidates. The court also found that Anglo voters ranked Hispanic candidates last or near last while Hispanic voters ranked Hispanic candidates first or tied for first.
Time is also probative in establishing a pattern of racial bloc voting. The Court in Gingles found persuasive a series of elections in which the minority group had sponsored candidates without electoral success. 478 U.S. at 57, 106 S.Ct. at 2770. The parties here have stipulated that a similar pattern exists in Watsonville. No Hispanic had ever been elected as mayor or city council member under the at-large system prior to the trial. Eight Hispanic candidates ran unsuccessfully for city council positions from 1971 to 1985 and one Hispanic ran for mayor in 1979. Twenty-five of 51 non-Hispanic candidates ran successfully for city council positions from 1971 to 1985. Such a pattern over time of minority electoral failure strongly indicates racial bloc voting. Gingles, 478 U.S. at 57, 106 S.Ct. at 2770.
The district court accepted these statistics and held that racial bloc voting occurs in the City. The court found strong evidence that Hispanies and Anglos supported different candidates. Although the court did not separately find that Anglo bloc voting occurs, it is clear that the non-Hispanic majority in Watsonville usually votes sufficiently as a bloc to defeat the minority votes plus any crossover votes.
D. Other factors
The district court found that the third Senate factor was not present, emphasizing that Watsonville does not employ electoral devices with the potential to dilute minority voting strength such as an anti-single-shot voting requirement or a majority vote requirement. However, the third Senate factor also includes “unusually large election districts” as a disfavored voting procedure. It is not disputed that an at-large system is employed in Watsonville.5 The parties also stipulated to the lack of a district residency requirement for mayoral or city council candidates. The absence of a district residency requirement can hinder the political access of minority groups because all the candidates can come from outside minority neighborhoods. See Rogers v. Lodge, 458 U.S. 613, 627, 102 S.Ct. 3272, 3280, 73 L.Ed.2d 1012 (1978); White v. Regester, 412 U.S. 755, 766 n. 10, 93 S.Ct. 2332, 2340, n. 10, 37 L.Ed.2d 314 (1973). Therefore the third Senate factor was present, at least in part.
The district court correctly found that no evidence of the fourth and sixth Senate factors, exclusion from the candidate slating process or the use of racial appeals in campaigns, had been shown.
The first and fifth Senate factors are, respectively, a history of official discrimination against the minority group and the extent to which the minority continues to bear the effects of discrimination in socioeconomic areas that hinder their ability to participate in the political process effectively. Appellants did not present evidence of historical discrimination against Hispanies other than the socioeconomic differences between Hispanies and non-Hispanics stipulated to by the parties. The appellants did not introduce such evidence because they believed that they had amply demonstrated a Section 2 violation. The district court disagreed and concluded that the first and fifth Senate factors were not present. Indeed, the district court found “no evidence whatever of official discriminatory practices of any sort directed at the Hispanic residents of Watsonville, either presently or historically.” It emphasized that the City government had not been unresponsive to the needs of Hispanies and that the City had been “singularly free of overt discrimination against Hispanies” during [1418]*1418the thirty-year period covered by the evidence. Further, the court found “no evidence that any differences in educational attainment, job opportunity, income or health benefits are the result of past discrimination of any kind, or that past discrimination has caused those factors to impede the ability of the Hispanic community to participate in the [political] process.”
Although we do not believe that the district court committed clear error in finding that the City of Watsonville has not itself engaged in discrimination against Hispanics, we nonetheless remain troubled by the court’s handling of the first and fifth Senate factors. The district court apparently believed that it was required to consider only the existence and effects of discrimination committed by the City of Watsonville itself. This conclusion is incorrect.
The first Senate factor requires consideration of “[t]he extent of any history of official discrimination in the state or political subdivision that touched the right of members of the minority group ... to participate in the political process.” S.Rep. No. 417 at 28, 1982 U.S.Code Cong. & Admin.News at 206 (emphasis added). Arguably, this limitation requires that one consider only electoral discrimination committed by the relevant political subdivision. Such a reading, however, would result in precisely the sort of mechanistic application of the Senate factors that the Senate Report emphatically rejects. The court is required to consider the totality of the circumstances, and given that the enumerated Senate factors are “neither comprehensive nor exclusive,” Gingles, 478 U.S. at 45, 106 S.Ct. at 2764, there is nothing to suggest that courts are forbidden to consider discrimination committed by parties other than the relevant political subdivision. Thus, even if the first Senate factor does embrace only discrimination committed by Watsonville, that does not imply that the district court may not consider any relevant history or effects of discrimination committed by others, such as the state of California.
Furthermore, such a restrictive reading places too much emphasis on the plaintiffs ability to prove intentional discrimination. Section 2 was amended by Congress precisely to relieve plaintiffs of the burden of showing such intent. While any intent to discriminate by Watsonville would indeed be supportive of the plaintiffs’ claim, plaintiffs need only show that, considering the totality of the circumstances, they do not have an equal opportunity to participate in the political process. There is no apparent reason why other forms of discrimination against Watsonville Hispanics may not be considered as factors that contribute to making the Watsonville at-large election scheme a device that impedes Hispanics’ equal participation in the electoral process.
Lastly, the court decisions from which the Senate factors were derived, see note 3, supra, both considered the existence of statewide discrimination as a factor in concluding that at-large elections in particular counties violated Section 2. See White v. Regester, 412 U.S. 755, 766-67, 93 S.Ct. 2332, 2339-40, 37 L.Ed.2d 314 (1973) (referring to statewide and countywide discrimination against blacks in Dallas County, Texas); id. at 767-68, 93 S.Ct. at 2340-41 (noting statewide discrimination against Mexiean-Americans); Zimmer v. McKeithen, 485 F.2d 1297, 1306 (5th Cir.1973) (referring to the effect of statewide racial segregation in education).
These arguments apply with equal force to the fifth Senate factor, which states that courts may consider “the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment, and health, which hinder their ability to participate effectively in the political process.” (emphasis added). Moreover, the literal language of the fifth Senate factor does not even support the reading that only discrimination by Watsonville may be considered; the limiting language describes the people discriminated against, not the discriminator.
The district court does not appear to have considered whether Watsonville Hispanics have suffered from discrimination by parties other than the City of Watson-ville or whether any such discrimination has affected the ability of Hispanics to participate effectively in the city’s electoral process. Thus, while the district court’s [1419]*1419interpretation of the first and fifth Senate factors rested on an erroneous view of the law, the appellants did not present, and the record does not contain, sufficient evidence of historical discrimination against Hispanics to permit this court to find that Watson-ville Hispanics have suffered from such discrimination.
Were it necessary to decide this issue, we would consider the propriety of taking judicial notice of the pervasive discrimination against Hispanics in California, including discrimination, committed by the state government, that has touched the ability of California Hispanics to participate in the electoral process.6 See, e.g., Castro v. State, 2 Cal.3d 223, 231, 466 P.2d 244, 249, 85 Cal.Rptr. 20, 25 (1970) (declaring a California constitutional provision making the ability to read English a prerequisite for voting unconstitutional as applied to those literate in another language). However, we conclude that, even without such a showing, plaintiffs have clearly established a violation of Section 2. As noted earlier, factors other than the three elements discussed above, while supportive of a Section 2 violation, are “not essential to a minority voter’s claim.” Gingles, 478 U.S. at 48 n. 15, 106 S.Ct. at 2766 n. 15. Here, the plaintiffs have shown that Watsonville Hispanics overwhelmingly and consistently have voting preferences that are distinct from those of white voters, that white voters have consistently voted as a racial bloc against such candidates, and that a single-member district system would result in some districts having a Hispanic majority. Considering all of these circumstances, we cannot but conclude that, regardless of the good faith of city officials, Watsonville’s at-large system is an impermissible obstacle to the ability of Hispanics to participate effectively in the political process.
We remand for implementation of a plan that comports with Section 2. The district court has broad equitable powers to fashion relief which will remedy the Section 2 violation completely. See S.Rep. No. 417 at 31, 1982 U.S.Code Cong. & Admin.News at 208.
III. ATTORNEYS’ FEES
Appellants request attorneys’ fees and costs for this appeal pursuant to 42 U.S.C. §§ 1973i(e) & 1988. The district court awarded the costs of suit to the City. “[T]he court, in its discretion, may allow the prevailing party, other than the United States, a reasonable attorney’s fee as part of the costs.” 42 U.S.C. § 1988; see also id. § 1973i(e). Attorneys’ fees can be awarded at the trial or appellate level to prevailing parties under 42 U.S.C. § 1988. See Sotomura v. County of Hawaii, 679 F.2d 152 (9th Cir.1982). Attorneys’ fees should be awarded in vote dilution claims unless special circumstances make such an award unjust. Campaign for a Progressive Bronx v. Black, 631 F.Supp. 975, 980 (S.D.N.Y.1986).
Appellants are clearly the prevailing party in this litigation and no special circumstances prevent an award in this case. We thus reverse the district court’s award of costs to appellees. On remand, the district court should award appellants costs and reasonable attorneys’ fees, including the fees incurred in pursuing this appeal. The court should determine the appropriate amount after implementation of a plan that does not violate Section 2.
CONCLUSION
We reverse the district court’s judgment in favor of the City because of its legal misunderstandings and erroneous findings of insufficient geographical insularity and political cohesiveness. We find that, based on the totality of the circumstances, the at-large scheme of mayoral and city council elections in Watsonville impermissibly dilutes the voting strength of Hispanics. We remand for implementation of a plan that comports with the requirements of Section 2.
REVERSED AND REMANDED.