This is a challenge under the Voting Rights Act of 1965, 42 U.S.C. §§ 1971 et seq., to a new system of conducting city council elections adopted by the City of Stockton, California, in 1986. The new electoral system implements, among other features, a change from district to at-large voting in the general election. We are called upon to decide whether plaintiffs’ *886claim, challenging the adoption of this new electoral system on the ground that it reduces the ability of minorities to elect representatives of their choice, may be maintained under section 2 of the Voting Rights Act, 42 U.S.C. § 1973(a). That section prohibits the imposition of any voting practice or procedure which results in “a denial or abridgement of the right of any citizen of the United States to vote on account of race....” Plaintiffs, a group of hispanic and black residents of Stockton, seek a return to the former single-member district electoral system.
The district court held that such a challenge to a new electoral system on the ground that, compared to the previous system, the new system reduced the minority's ability to elect representatives, could be maintained only under section 5 of the Voting Rights Act, 42 U.S.C. § 1973c. That section requires certain jurisdictions to obtain federal authorization (or “preclearance”) before implementing any change in electoral systems, and prohibits any change that results in a reduction or “retrogression” of the ability of protected minority groups to elect representatives as compared with the previous system. See Beer v. United States, 425 U.S. 130, 96 S.Ct. 1357, 47 L.Ed.2d 629 (1976). The district court held that because the City of Stockton is not covered by the preclearance provisions of section 5, which apply only to about 38% of the country, the plaintiffs’ dilution claim is not maintainable under that section.
The district court also held that in order to prevail under section 2 of the Voting Rights Act, the plaintiffs must make the same evidentiary showing that the Supreme Court required in Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986). The plaintiffs in Gingles did not challenge the institution of a new system and request a return to the former system. They instead challenged a long-established at-large electoral system. They asked the district court to create a single-member districting system. In order to justify imposition of a new single-member district system, the Supreme Court required plaintiffs to show that a bloc voting majority was usually able to defeat candidates supported by a politically cohesive, geographically insular minority group. 478 U.S. at 49, 106 S.Ct. at 2766. In applying the Gingles analysis to this case, the district court held that while plaintiffs’ evidence clearly established that hispanics and blacks together could form a majority in a single-member district, the evidence failed to establish that such a combined group of blacks and hispanics would vote in a politically cohesive manner that would guarantee election of a minority representative. It also found that plaintiffs had failed to establish that white bloc voting would defeat the candidate supported by minorities. It further found that plaintiffs failed to show either blacks or hispanics, when viewed as separate groups, voted cohesively-
We do not reject any of the district court’s findings concerning the plaintiffs’ inability to satisfy the standards applied in Gingles. The City’s new electoral system does, however, embody many of the classic devices for reducing a minority’s ability to elect representatives. We conclude that under section 2 as interpreted by the Supreme Court in Gingles, minority plaintiffs must at the very least establish that under the system they wish reinstituted under section 2, minorities exhibited sufficient cohesion to cause the election of their preferred candidates. In this case, because the minority plaintiffs have not demonstrated that minority voters have voted for minority representatives in a cohesive pattern, that condition is not satisfied. We therefore affirm.
I. BACKGROUND
The City of Stockton adopted the challenged electoral system, known as “Measure C,” by referendum in November of 1986. Under the previous system, adopted in 1971, the city was divided into nine districts for purposes of municipal elections. To represent a district on the City Council, a candidate had to reside in that district and receive a plurality of votes from that district. The mayor was elected at-large from sitting council members. Recalls of *887city council members were based upon a recall election conducted solely within the affected district. From the time of its adoption in 1971 until the adoption of Measure C in 1986, the district system resulted in consistent representation of minority members on the City Council in greater proportion than in the city’s population.
Measure C changed the preexisting district system in several material respects. Measure C reduced the number of districts from nine to six, and established a two-step election process. Under Measure C, a primary election is held within each of the six districts. The two candidates receiving the most votes in each district then run in a general election. This general election is conducted at-large, with all voters in the city choosing which of the two nominees from each district will serve on the Council. The mayor is also elected at-large in the general election, without a district nomination process. Unlike the former district system, where vacancies created by recall were filled by an election within the district, under Measure C any vacancy created by a recall election is filled by a person appointed by remaining City Council members.
According to 1980 census data, 10.4% of the population of Stockton is black; 22.1% is hispanic; and 9% are Asian and other minority groups. The 1980 census data indicate that the black and hispanic communities are concentrated in south Stockton, while north Stockton is predominantly white. At the time Measure C was adopted, of the three blacks on the council elected under the 1971 system, two were elected from predominantly white districts. There was one hispanic member.1
Plaintiffs, a group of hispanic and black voters residing in Stockton, filed this action in December of 1987. The action was originally certified as a class action, but the class was later decertified. Plaintiffs claimed that Measure C violates section 2 of the Voting Rights Act, and sought a return to the previous electoral system.
The district court held a bench trial in June of 1989. At the close of plaintiffs’ case-in-chief, the district court granted Stockton’s motion for dismissal pursuant to Fed.R.Civ.P. 41(b). The district court held that the plaintiffs had failed to make out a valid section 2 claim. See 42 U.S.C. § 1973c; 28 C.F.R. pt. 51 app. (1989). The district court applied the section 2 analysis employed in Thornburg v. Gingles, in which the plaintiff class prevailed by showing that the minority group in question could form a voting majority in a single member district, that the group was politically cohesive, and that a white majority consistently voted as a bloc to defeat any candidates preferred by the minority group. See Thornburg v. Gingles, 478 U.S. at 49-51, 106 S.Ct. at 2766-67. The district court found that the plaintiffs had not presented sufficient evidence to establish that blacks and hispanics would vote together or separately as a politically cohesive group or that the white majority would vote as a bloc, overshadowing minority voting strength. This appeal followed.
II. SECTION 2 OF THE VOTING RIGHTS ACT
Section 2 of the Voting Rights Act prohibits a state or political subdivision from imposing any voting “practice[] or procedure ... which results in a denial or abridgment of the right ... to vote on account of race.” 42 U.S.C. § 1973(a).2 A *888violation of section 2 occurs where the “totality of the circumstances” reveal that “the political processes leading to nomination or election ... are not equally open to participation by members of a [protected class] ... 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(b).
During the years following the original enactment of section 2, many actions were filed challenging the constitutionality of allegedly discriminatory practices which had the effect of diluting minority voting strength. Such practices included multi-member, at-large voting systems which in many cases were alleged to have had the effect of making it unfairly difficult for minorities to elect representatives of their own choice. See, e.g., Burns v. Richardson, 384 U.S. 73, 86 S.Ct. 1286, 16 L.Ed.2d 376 (1966) (challenging a multi-member dis-tricting plan); Fortson v. Dorsey, 379 U.S. 433, 85 S.Ct. 498, 13 L.Ed.2d 401 (1965) (challenging at-large election scheme). See also Recent Development, Section 2 of the Voting Rights Act: An Approach to the Results Test, 39 Vand.L.Rev. 139, 140-42 & n. 8 (1986).
In White v. Regester, 412 U.S. 755, 765, 93 S.Ct. 2332, 2339, 37 L.Ed.2d 314 (1973), the Supreme Court upheld the district court’s findings that certain multi-member districts, designed and operated in a Texas county, unconstitutionally discriminated against Mexican-Americans by limiting their ability to participate in the election of state legislative representatives. The Court emphasized that multi-member state legislative districts are not per se unconstitutional, but found that the plaintiffs had met their burden of “producing] evidence to support findings that the political processes leading to nomination and election were not equally open to participation by the group in question — that its members had less opportunity than other residents in the district to participate in the political processes and to elect legislators of their choice.” 412 U.S. at 765-66, 93 S.Ct. at 2339-40. The Court’s analysis focused on a number of historical factors and their negative effect on minorities’ participation and minorities’ ability to elect representatives. Id. at 766-69, 93 S.Ct. at 2339-41.
In City of Mobile v. Bolden, 446 U.S. 55, 100 S.Ct. 1490, 64 L.Ed.2d 47 (1980), however, a plurality of the Court seemed to depart from the Regester analysis, stating that the existing version of section 2 of the Voting Rights Act merely restated the fifteenth amendment and “was intended to have an effect no different from that of the Fifteenth Amendment itself.” 446 U.S. at 60-61, 100 S.Ct. at 1496. The Court held that the plaintiffs’ burden included a showing of intentional discrimination, and therefore rejected plaintiffs’ challenge to the city’s at-large voting system. Id. at 66-67, 100 S.Ct. at 1499-1500.
Congress amended section 2 in 1982 to overrule the Supreme Court’s interpretation of section 2 in Mobile and to make clear that a plaintiff need not prove intentional discrimination to make out a section 2 violation. Congress added subsection (b), using language which harkens back to the White v. Regester opinion, and including in the accompanying Senate Report a list of factors to guide courts in evaluating the “totality of the circumstances,” a phrase used by the White v. Regester Court. See *889S.Rep. No. 417, 97th Cong., 2d Sess. 21, 28-29, reprinted in 1982 U.S.Code Cong. & Admin.News 177, 198-99, 206-07.
Section 2 as amended in 1982 requires courts to look at the effect of the challenged practice, not at the motivations of those who instituted it. See Thornburg v. Gingles, 478 U.S. 30, 43-46, 106 S.Ct. 2752, 2762-64, 92 L.Ed.2d 25 (1986). See generally Recent Development, supra at 154-58. The Senate Report expressly stated that vote dilution claims are to be brought within the scope of section 2 as amended, and that Congress was adopting the “functional view of ‘political process’ used in White rather than the formalistic view espoused by the plurality in Mobile." S.Rep. No. 417, supra at 30 n. 120, 1982 U.S.Code Cong. & Admin.News at 208. Congress thus rejected the implication in Mobile that vote dilution may not be unlawful. “Likewise, although the plurality [in Mobile ] suggested that the Fifteenth Amendment may be limited to the right to cast a ballot and may not extend to claims of voting dilution (without explaining how, in that case, one’s vote could be ‘abridged’), this section without question is aimed at discrimination which takes the form of dilution, as well as outright denial of the right to register or to vote.” Id. Given the language and history of section 2, we must conclude that section 2 encompasses a vote dilution claim like that presented here.
III. THE PLAINTIFFS’ DILUTION CLAIM
Plaintiffs here alleged that the substitution of Measure C for the previous electoral system unlawfully dilutes their voting strength by making it more difficult for them to elect representatives of their choice. Plaintiffs cite the history of single-member district voting in Stockton which resulted in minority representation. They point out that Congress intended to prohibit all devices that “deny or abridge” the right to vote, and that such abridgement is often difficult to quantify. See S.Rep. No. 417, supra, at 28-29, 1982 U.S.Code Cong. & Admin.News at 207-08; Blacksher & Menefee, From Reynolds v. Sims to City of Mobile v. Bolden: Have the White Suburbs Commandeered the Fifteenth Amendment?, 34 Hastings L.J. 1, 50 & n. 318 (1982).
Plaintiffs correctly maintain that Measure C embodies many electoral devices and practices that have been readily identified as common means of diminishing minority voting strength. Congress in enacting the Voting Rights Act recognized that among the most commonly employed of these devices are at-large election systems, runoff requirements, anti-single-shot devices, and decreases in the size of the elected body. See, e.g., S.Rep. No. 417, supra at 29, 1982 U.S.Code Cong. & Admin.News at 206. See also Davidson, Minority Vote Dilution: An Overview, in Minority Vote Dilution 5-9 (C. Davidson ed. 1984) (describing these dilutive devices). As the district court recognized, the devices used in Measure C “enhance the opportunity for discrimination against the plaintiff class.”
The most prominent of Measure C’s dilu-tive mechanisms is its incorporation of an at-large component. In an at-large (or mul-ti-member district) system, all voters elect all representatives, and each voter has as many ballots as there are positions available. This system contrasts with a single-member district plan, under which the entire political jurisdiction is divided into districts roughly equal in population, each of which selects one representative by vote within the district. The potential for vote dilution that accompanies an at-large or multi-member district plan has long been recognized:
In the geographically compact areas where they live, minorities have the votes to determine the outcome of an election in a single-member district or ward system. By virtue of multi-mem-ber or at-large plan, however, they remain a minority in such a system’s larger voting population. The majority elects all the representatives, of course, and therefore candidates preferred by the minority group are consistently defeated.
Bonapfel, Minority Challenges to At-Large Elections: The Dilution Problem, 10 Ga.L.Rev. 353, 355 (1976). As a result, *890the minority group in an at-large system is denied the opportunity to select a representative who is responsive to the minority group’s own views and needs and who is accountable to the group. Id. A switch from an at-large system to single-member districts greatly increases the ability of minority groups to elect candidates of their choice. See Davidson and Korbel, At Large Elections and Minority Group Representation, in Minority Vote Dilution, supra, at 65. See also Gingles, 478 U.S. at 47 & n. 13, 106 S.Ct. at 2764 & n. 13 (stating that at-large voting schemes have long been recognized as dilutive of minority voting strength, citing, inter alia, Bonap-fel, and Davidson and Korbel, supra).
Measure C additionally incorporates a majority vote requirement and the functional equivalent of a runoff election, features which also tend to dilute minority voting strength. These requirements eliminate the chance that a minority will elect its candidate of choice should the majority split its votes among two or more other candidates. “[T]he minority candidate is forced into a runoff against a single white, behind whom the white voters can rally to produce a majority.” Davidson, supra at 6; Bonapfel, supra at 358. In addition, the runoff requirement increases the expense and other burdens of minority candidates by requiring them to run a second time, at-large. The district court correctly found that Measure C’s runoff requirement and majority vote requirement increased the opportunity for discrimination against minorities in Stockton.
Other features of the new system may also do so. Measure C incorporates a so called “anti-single-shot mechanism.” In a typical at-large system, there are several positions to be filled; all candidates compete against each other for those seats, and the candidates with the highest number of votes win. A minority group can “single-shot” by agreeing to vote only for a few preferred candidates and withholding any remaining votes. This concentrates the voting power of the minority group and prevents majority candidates from receiving minority votes. However, Measure C requires a voter to select a candidate for a designated geographic district in the at-large phase of the election, preventing minority vote concentration.
In addition, Measure C reduces the number of districts from nine to six, increasing the size of each district and reducing the number of potential seats on the city council. The increased size of the districts not only waters down minority voting strength within each district, but also increases the costs of campaigning in Stockton, further discriminating against minorities. With fewer districts, there are also fewer seats to fill, increasing competition for the remaining seats and discouraging majority voters from filling one of those seats with a minority. Davidson, supra at 7.
It is thus clear, as the district court in effect found, that the particular mechanisms embodied in Measure C would make it much more difficult for a cohesive minority population to elect representatives of their choice. The mechanisms increase the likelihood of minority voting strength dilution. The adoption of such devices does not of itself, however, mean that such dilution has actually occurred.
We need not decide whether, as appellees contend, plaintiffs in this type of section 2 action must show that minority voters would form a cohesive majority able to elect representatives of their choice in one or more districts under the system they seek to reinstitute. Whether a majority is required, or a dominant plurality is enough, makes no difference in the outcome of this case. In order to maintain a retrogression claim under section 2, plaintiffs must at least be able to demonstrate that minority voters have the ability to elect their preferred candidates under the system they seek. This is the teaching of Gingles, where Justice Brennan stressed that the challenged device must be shown actually to impair the ability of minority voters to elect representatives of their choice. The opinion points out that unless the minority group is politically cohesive, it cannot be *891said that the challenged structure “thwarts distinctive minority interests.” 478 U.S. at 45, 49, 106 S.Ct. at 2763-64, 2766. This is also the view taken by the only other circuit to consider a retrogression claim under section 2. Ketchum v. Byrne, 740 F.2d 1398 (7th Cir.1984), cert. denied, 471 U.S. 1135, 105 S.Ct. 2673, 86 L.Ed.2d 692 (1985). The Seventh Circuit there stated that “retrogression may be defined as a decrease in the new districting plan or other voting scheme from the previous plan or scheme in the absolute number of representatives which a minority group has a fair chance to elect" (emphasis added). Plaintiffs must be able to show that minorities have in the past voted cohesively for minorities and have the potential to elect minority representatives.
It is here where plaintiffs’ claim founders. The district court found that plaintiffs’ testimony, both expert and lay, failed to prove that blacks and hispanics were politically cohesive, either when combined or when considered separately. These are factual findings which must be accepted unless clearly erroneous. The district judge found the plaintiffs’ expert biased and not fully credible, determinations which we have no basis for overturning.
The statistical data before the district court reflected that while in past elections minorities had been elected, those elected were for the most part not elected by minorities. Of the three black representatives in office at the time Measure C was adopted, two were elected from districts that were overwhelmingly (more than 70 percent) white. Prior to 1971, when an at-large system was in effect, hispanics had been elected by white majorities, further indicating that the minority representation before Measure C existed because whites voted for minorities. Although plaintiffs maintain they are now unable to elect minority representatives, the record before the district court could not include any pattern of data from post-Measure C elections which might establish a pattern of cohesive voting sufficient to carry plaintiffs’ burden. See Note 1, supra. We therefore accept the district court’s conclusion that plaintiffs did not establish that the adoption of Measure C resulted in plaintiffs having “less opportunity to elect legislators of their choice” within the meaning of 42 U.S.C. § 1973(b) as interpreted by the United States Supreme Court in Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25.
AFFIRMED.
. Plaintiffs ask us to take judicial notice of the fact that in the 1990 election, conducted after the district court’s judgment, no hispanic and only one black representative was elected, although the City’s first Asian representative was elected. We have held that election results are an appropriate subject of judicial notice. See, e.g., Romero v. City of Pomona, 883 F.2d 1418, 1420 n. 1 (9th Cir.1989). In this case, however, results standing alone, of a single election, are of little value. What is significant is the underlying voting pattern which a single election re-suit, standing alone, does not illuminate. See Gingles, 478 U.S. at 57, 106 S.Ct. at 2769-70. Even if we were to take judicial notice of the result of this election, it would not influence our decision in this case.
. Section 2 states in its entirety:
(a) No voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of *888any citizen of the United States to vote on account of race or color, or in contravention of the guarantees set forth in section 1973b(f)(2) of this title, as provided in subsection (b) of this section.
(b) A violation of subsection (a) of this section is established if, based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of a class of citizens protected by subsection (a) of this section 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. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, 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.