This is a vote dilution case. It involves challenges to the at-large systems used to elect the Dallas County [Alabama] Commission and the Dallas County Board of Education. Plaintiff sued both the Dallas County Commission and the Board of Education under the Fourteenth and Fifteenth Amendments; the Civil Rights Act of 1870, as amended, 42 U.S.C. § 1971(a)(1) (1976); and § 2 of the Voting Rights Act of 1965, as amended, 42 U.S.C. § 1983 (West Supp. 1984).
The case has a complicated history in the district court because of intervening changes in the law. The government brought suit in 1978, and the ease was originally tried in 1979 and 1980 under the then-prevailing standards of the Fifth Circuit as set forth in Nevett v. Sides, 571 F.2d 209 (5th Cir.1978), cert. denied, 446 U.S. 951, 100 S.Ct. 2916, 64 L.Ed.2d 807 (1980), Kirksey v. Board of Supervisors, 554 F.2d 139 (5th Cir.) (en banc), cert. denied, 434 U.S. 968, 98 S.Ct. 512, 54 L.Ed.2d 454 (1977) and Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir.1973) (en banc), affd on other grounds sub nom East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976). Later, before the case was decided, the Supreme Court decided Mobile v. Bolden, 446 U.S. 55, 100 S.Ct. 1490, 64 L.Ed.2d 47 (1980), a case that “completely changed the mode of assessing the legality of electoral schemes alleged to discriminate against a class of citizens.” Jones v. City of Lubbock, 640 F.2d 777, 777 (5th Cir.1981) (Unit A) (Goldberg, J., concurring), modified 682 F.2d 504 (1982). The government had challenged Ala.Code § 16-8-1 (1975), the section under which the Board of Education was elected, only as applied and not on its face,1 so the court held the government *1533would not be permitted to present evidence of discriminatory intent in the enactment of section 16-8-1. The court did allow the government to introduce evidence of discriminatory intent in the enactment of 1901 Ala.Acts 328(6), the section under which the County Commission is elected.
After the evidence was concluded but before the case was decided, Congress amended section 2 of the Voting Rights Act to overrule Bolden. S.Rep. No. 417, 97th Cong., 2d Sess. 15-16, reprinted in 1982 U.S.Codé Cong. & Au.News 177, 192-93 [hereinafter cited as 1982 Senate Report], Under subsection (b) the central inquiry in determining if there has been a violation of subsection (a) is whether it has been shown that the “political processes leading to nomination or election ... are not equally open to participation” by a protected class. A violation is established based on the “totality of circumstance,” and the extent to which members of the class have been elected to office is relevant.
When section 2 was amended the Commission did not request that additional evidence be taken. The parties submitted proposed findings of fact and conclusions of law directed to the standard of new section 2. The court subsequently entered extensive findings of fact and conclusions of law. It found that the government had not proven vote dilution against either the Board of Education or the City Commission. The government appealed.
Amended section 2 now reads:
Denial or abridgement of right to vote on account of race or color through voting qualifications or prerequisites; establishment of violation
(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 any 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.
*1534I. CONSTITUTIONALITY OF AMENDED SECTION 2
On appeal the Commission challenges the constitutionality of amended section 2. The Commission and the Board also assert that vote dilution claims are not cognizable under the Fifteenth Amendment or section 2 of the Voting Rights Act. These issues have been decided adversely to the appellants. U.S. v. Marengo County Commission, 731 F.2d 1546 (11th Cir.1984).
The Commission further contends that amended section 2 should not apply to this litigation because it would have introduced additional evidence had it known that the court would no longer require proof of discriminatory intent. This argument is unpersuasive. “[A] court is to apply the law in effect at the time it renders its decision, unless doing so would result in manifest injustice or there is statutory direction or legislative history to the contrary.” Bradley v. Richmond School Board, 416 U.S. 696, 711, 94 S.Ct. 2006, 2016, 40 L.Ed.2d 476 (1974). Neither the statute nor the legislative history indicates that the statute is not to apply to pending cases. Marengo County, 731 F.2d at 1553. Also, no manifest injustice will occur from applying amended section 2 to this case. As explained above, when the suit was originally tried the parties recognized that the court should consider the Zimmer factors. Bolden did not make these factors irrelevant; rather it provided they could be indicative but not dispositive on the issue of intent. Lodge v. Buxton, 639 F.2d 1358, 1375 (5th Cir.1981) (Unit B), affd sub nom Rogers v. Lodge, 458 U.S. 613, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1982). Under amended section 2 a court should consider the Zimmer factors. See 1982 Senate Report at 28-29, reprinted in 1982 U.S.Code Cong. & Ad.News at 206-07. Consequently the factors relevant in considering a vote dilution claim under amended section 2 were relevant in the period after Bolden but before the amendment, which is the period during which the Commission presented its defense. Furthermore, the Commission did not request an additional hearing after the amendment, as the government had done after Bolden, but instead filed its proposed findings of fact and conclusions of law under the new. standard of amended section 2. Given the ongoing relevance of the Zimmer factors throughout the period of trial and the Commission’s willingness to submit proposed findings and conclusions under the new standard, we find no manifest injustice in application of amended section 2 to this case. Cf. Marengo County, 731 F.2d at 1554 (finding no manifest injustice because “the question is whether the present maintenance of that system meets the standards in effect now”) (emphasis in original).
II. APPLICATION OF SECTION 2
This court explained in Marengo County [t]he language and history of the statute make clear several points. First, discriminatory intent need not be shown to establish a violation. Second, at-large elections are not prohibited per se, nor does a lack of proportional representation automatically require a finding of a violation. At the same time, however, the absence of minority elected officials may be considered an indicium of violation, and an at-large system will violate the statute if it results in a denial of equal participation. Congress noted that some at-large systems diluted black votes, and would be vulnerable under the amended statute. 1982 Senate Report at 6. Third, section 2 focuses not on whether minority groups receive adequate public services but on whether minorities have an equal right to participate in the election process. See id. at 36.
731 F.2d at 1564-65 (emphasis in original and footnotes omitted).
We have pointed out, and Marengo County recognized, that a section 2 case requires examining the totality of the circumstances. The Senate Report refers to nine factors that it considers “typical,” which are to be weighed under the totality of circumstances approach:2
1. the extent of any history of official discrimination in the state or political *1535subdivision that touched the right of the members 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.
Additional factors that in some cases have had probative value as part of plaintiffs’ evidence to establish a violation are:
[8.] whether there is a significant lack of responsiveness on the part of elected officials to the particularized needs of the members of the minority group.
[9.] whether the policy underlying the state or political subdivision’s use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.
1982 Senate Report at 28-29, reprinted in 1982 U.S.Code Cong. & Ad.News at 206-07 (footnotes omitted).
All nine factors listed in the Senate Report appear in this ease. We have considered these nine.3 We conclude that the district court wrongly decided two of them, racial polarization in voting and the structure of the election system. This conclusion requires that the case be reversed and remanded. With respect to a third factor, the existence of past discrimination and its limiting effects, we find that some of the district court’s subsidiary findings are correct and others erroneous. We express no view of the merits of this factor and direct that the court reconsider it on remand. Finally, with respect to others of the eight factors we find that the district court’s conclusions are not plainly erroneous and as to them we affirm.
A. Racially polarized voting
In Marengo County we held that racially polarized voting “will ordinarily be the keystone of a dilution case.” 731 F.2d at 1566. It is “[t]he surest indication of race-conscious politics.” Id. at 1567.
The district court determined that “in those races in which Whites have opposed blacks, there has been evidence of polarization in the voting of both white and black voters.” 548 F.Supp. 794 at 853 (footnote omitted), 905.4 However, the court dis*1536counted racial polarization because some blacks ran as candidates of fringe parties, blacks made little effort to attract white support, whites often supported an incumbent over a black, blacks seldom used the media as extensively as whites, and there was a high level of apathy of both black and white voters. Id. at 853-54, 905-06.
Beginning in 1966 blacks have run in 34 races; they have lost every time. 548 F.Supp. at 848-50. The district court concluded that apathy was the prime cause of black defeat at the polls. In this the court erred. The existence of apathy is not a matter for judicial notice. Kirksey, 554 F.2d at 145; see also Marengo County, 731 F.2d at 1574. The record does not contain evidence supporting the district court’s finding that apathy was a prime cause of black defeat.
The record also does not support the district court’s conclusion that campaigns of blacks as candidates of fringe parties and against incumbents helped explain the polarized voting occurring in Dallas County. In 1970 three incumbent commissioners defeated three blacks running on the National Democratic Party of Alabama ticket (a fringe party). Since 1970 blacks have apparently not run on this or any other fringe ticket.5 Fourteen blacks have lost elections during this time. Additionally, in this period, in two of three races for commissioner a white incumbent has been opposed by a black and a white and has been unseated by the white opponent. Thus, the “fringe party” and incumbency explanations are not supported by the record.
Another explanation for voting polarization was that blacks did not actively seek white votes. As noted in Marengo County, 731 F.2d at 1567, the failure of blacks to solicit white votes may be caused by the effects of past discrimination, an issue that we direct be reconsidered on remand.
We conclude that the court’s finding that racial polarization in voting was not a significant factor is clearly erroneous.
B. Structure of the election system6
The court concluded that the requirement of a majority in the primary plus the significance of the Democratic primary combined to “weigh[ ] in favor of a finding of dilution on this criterion.” 548 F.Supp. at . 908. This finding is correct; the Supreme Court has recognized the potential adverse effects of a majority vote requisite. See City of Rome v. U.S., 446 U.S. 156, 183-84, 100 S.Ct. 1548, 1564-65, 64 L.Ed.2d 119 (1980). The district court also determined that there was no anti-single shot provision and that the requirement that a candidate run for a specific post was neutral. 548 F.Supp. at 908. This is clearly erroneous. Even “though there is no anti-single shot provision, the requirement that candidates run for numbered posts has potential effects that are equally adverse.” Buxton, 639 F.2d at 1380.
*1537Additionally, the court below decided that the requirement that candidates live in specified districts to run for certain seats weighed in favor of finding no dilution. 548 F.Supp. at 908. Courts have interpreted residency requirements as both showing evidence of dilution and showing no dilution. Compare Rogers, 458 U.S. at 627, 102 S.Ct. at 3280 (evidence of no dilution) with City of Rome, 446 U.S. at 185 & n. 21, 100 S.Ct. at 1565 and n. 21 (evidence of dilution). We conclude that the existence of a residency requirement does not, by itself, provide probative evidence on the question of dilution when the election structure, through another mechanism such as numbered posts, already provides for the separation of one contest (e.g., election of county commissioners) into several contests. See City of Rome, 446 U.S. at 185 n. 21, 100 S.Ct. at 1565 n. 21.
Finally, the court found the size and population of Dallas County not unusually large and that this criterion was neutral. The court’s finding and conclusion on this sub-issue are correct.
Considering all these subsidiary factors together, we hold that the district court erred in finding that the combination of a majority requirement in the primary, the significance of the Democratic primary, and the use of numbered posts did not operate to cancel voting strength of minorities.
Given the significance of the two factors we have discussed, polarized voting and the structure of the election system under amended section 2, we reverse and remand for further consideration.
C. Past discrimination and its lingering effects
On this factor some of the district court’s findings are plainly erroneous, others not. This issue should be reconsidered on remand.
The district court found that in the past Alabama required, in order to register to vote, that one pass a literacy test, pay a poll tax, and present a voucher of good character from a registered voter. 548 F.Supp. at 800, 882. The court concluded nevertheless that there were no lingering effects of these past discriminatory requirements. Id. at 846, 909-10. This conclusion was based on several factors, including blacks’ ability to participate in the election process by registering, voting, running for office, and serving as election officials; the progress of integration in the school system; the development of influential black political leaders; and the substantial equality in voter registration of blacks and whites. Id. at 846-47, 854-55, 899, 906.
Schools: The court correctly recognized that Dallas County has made substantial progress in integrating its schools. Blacks nevertheless continue to have lower income and education levels. 548 F.Supp. at 855, 906. As Marengo County explained, “[bjecause blacks are poorer and less educated they have less political influence than whites.” 731 F.2d at 1568. Once lower socio-economic status of blacks has been shown, there is no need to show the causal link of this lower status on political participation.
[T]he court imposed an improper burden of proof of causal relationships. The Supreme Court and this court have recognized that disproportionate educational, employment, income level and living conditions tend to operate to deny access to political life. In this case the court held that these economic and education factors were not proved to have “significant effect” on political access in Hinds County. It is not necessary in any case that a minority prove such a .causal link. Inequality of access is an inference which flows from the existence of economic and educational inequalities.
Kirksey, 554 F.2d at 145; accord Marengo County, 731 F.2d at 1569.
Political leadership: The court found that Dallas County has several influential black political leaders. This finding is not clearly erroneous. However, it is not apparent to us, without additional evidence of their status, power and influence and their impact on political affairs, or further explanation, or both, why the existence of *1538black political leaders tends to show that the effects of past discrimination have dissipated. On remand the court should consider this sub-issue further.
Registration: The district court found that blacks are now registered in substantially equal numbers with whites. This is not clearly erroneous. Using population estimates, the district court determined that in 1976 blacks were approximately 44 percent of the voting age population and 43.8 percent of the registered voters. A difference of .2 percent is not significant. By 1980 blacks were 49.8 percent of the voting age population but only 44.8 percent of the registered voters. It cannot be reasonably expected that registration percentages will mirror voting age percentages; some deviation between the two must be acceptable. We conclude that a difference of 5 percent is not significant.
While blacks do register in approximately the same numbers as whites, evidence suggests that blacks continue to vote at a rate 5 percent to 10 percent lower than whites. 8 Rec. at 1586 (Cottrell). Because the combination of somewhat lower registration rates and lower turnout rates may show that the effects of past discrimination still linger, the district court on remand should determine whether blacks in fact turn out at a lower rate and, if so, consider the significance of this combination.
Subjective evidence: The district court did not extensively discuss subjective evidence of continuing effects of past discrimination. See 548 F.Supp. at 845, 888. A number of government witnesses testified to lingering hostility and fear by blacks of the county courthouse, the site of the Board of Registrars. See, e.g., 4 Rec. at 138-40, 222-23 (Reese); 5 Rec. at 498 (Crum); id. at 689 (Moss); 7 Rec, at 1064-65 (Chestnut). The district court discounted this evidence because it concluded that “[t]he Court House is the seat of county government, and is therefore a proper and appropriate place for the Board of Registrars to conduct its activities.” 548 F.Supp. at 845. This holding does not address the “lingering fear” issue.'
The government introduced evidence of these practices of the Board of Registrars: limited hours, failure to go to the beats to register voters,7 and refusal to appoint deputy registrars. In their consideration of the 1982 amendments to the Voting Rights Act, both the House and the Senate expressed concern over these kinds of barriers to registration. The House Judiciary Committee viewed “inconvenient location and hours of registration ... [and] refusal to appoint minority registration and election officials” as “continued barriers to registration and voting.” H.R.Rep. No. 227, 97th Cong., 1st Sess-. 14 (1981). In a like vein, the Senate Judiciary Committee described such devices as “blatant direct impediments to voting.” 1982 Senate Report at 10 n. 22, reprinted in 1982 U.S. Code Cong. & Ad.News at 187 n. 22. See generally Note, Eradicating Racial Discrimination in Voter Registration: Rights and Remedies Under the Voting Rights Act Amendments of 1982, 52 Ford-ham L.Rev. 93, 110-12 (1983).
In light of Congress’s concerns about these practices, the district court erred in treating their presence in Dallas County as insignificant. The court concluded that the hours were reasonable8 and furthermore affected blacks and whites equally. 548 F.Supp. at 845, 889. While being open during the day may seem reasonable on first consideration, the evidence indicates that such hours are inconvenient to those who work, especially those who work in the rural areas of the county. See 5 Rec. at 499 (Crum); 6 Rec. at 771 (Harvin); id. at 919 (Dixon); 8 Rec. at 1344-46 (Pettway) (explaining that a person only has to go a few miles to vote but may have to go 28 miles to the city and lose a day’s pay in order to register). As Congressman Hyde, ranking minority member of the House Subcommittee, stated after listening to testimony at a hearing on amendments to the Voting Rights Act held in Montgomery, Alabama,
*1539I want to say that I have listened with great interest and concern, and I will tell you, registration hours from 9 to 4 is outrageous. It is absolutely designed to keep people who are working and who have difficulty in traveling from registering.
If that persists and exists, it is more than wrong____
Extension of the Voting Rights Act: Hearings Before the Subcomm. on Civil and Constitutional Rights of the House Comm, on the Judiciary, 97th Cong., 1st Sess. Part 2 at 1584 (1982). There was evidence that the Board had previously held night and Saturday meetings, 14 Rec. at 3217 (Horne), but there was also evidence that these meetings were publicized only at the courthouse, so that most people were unaware of them, 5 Rec. at 418 (Foster). We find the district court erred in not concluding that this practice hindered access to the political process by blacks and continued the effects of past discrimination.
We reach the same conclusion with regard to the Board’s failure, between 1978 and 1982, to go to the beats and register voters. That the county courthouse is an appropriate meeting place for the Board of Registrars does not address the possible need for the Board to take steps to carry on the registration process at other locations. Cf. 1982 Senate Report at 53-55, reprinted in 1982 U.S.Code Cong. & Ad. News at 231-34 (describing positive steps necessary for covered jurisdiction to take under § 4 of the Voting Rights Act in order to qualify for bailout from § 5 of the Act). Some evidence suggests that in 1982 the Board decided to go to the beats to register voters, 14 Rec. at 3205 (Horne), but the record does not reveal whether this was carried out. We therefore conclude that the district court erred in treating the Board’s meeting at only the county courthouse a¡¡^ sufficient in light of the historical disenfranchisement of blacks in Dallas County.
We also find the district court improperly considered as unimportant the failure to appoint deputy registrars. See 548 F.Supp. at 845-46, 889. One of the positive steps Congress suggested for jurisdictions seeking to bail out of § 5 was “appointment of deputy registrars who are present at locations accessible to minority candidates.” 1982 Senate Report at 55, reprinted in 1982 U-S.Code Cong. & Ad. News at 233. While the district court correctly characterized the appointment of deputy registrars as discretionary under the law, the fact that the Board refused to appoint such deputies is evidence of a failure to act to overcome past discrimination.
This issue of lingering effects of past discrimination must be reconsidered by the district court. We express no view on the ultimate merits of the issue.
III. OTHER FACTORS
We have discussed three of the eight factors. Now we turn to the other five and hold the district court correctly decided them.
The court determined that no white groups slate candidates, a fact that supports a finding of no dilution. 548 F.Supp. at 851, 855, 856, 90S, 907, 910. It found that there have been no racial campaigns since 1966 and that this factor weighs in favor of a finding of no dilution. Id. at 855, 907, 916. It held that no blacks had been elected in modern (post-1965) times and that this determination supports a finding of dilution. Id. at 847, 857, 899, 908. All these findings are correct.
The district court’s finding of responsiveness, a subsidiary factor under amended § 2, is also correct. Because the court’s findings concerning the Board of Education and the County Commission rest on different evidence, they are considered separately.
On the issue of responsiveness by the Board of Education, the government and the Board stipulated that the district court could take judicial notice of Lee v. Dallas County Board of Education, 456 F.Supp. 1164 (S.D.A1.1978), a case involving school desegregation and faculty hiring in Dallas County. The district court set out an impressive body of evidence showing significant improvements in Dallas County schools. The government did not include in the record for this appeal the Lee record. *1540Fed.R.App.P. 10(b)(1) and 11th Cir.R. 14(a)(1) require the appellant to include in the record all parts of the transcript of the proceedings not already on file and relevant to findings that it contends are unsupported by or contrary to the evidence. In the absence of a complete record we cannot adequately review the challenged findings on responsiveness by the Board and must affirm the district court on this issue. See U.S. v. Bob Lawrence Realty, Inc., 474 F.2d 115, 126 (5th Cir.), cert. denied, 414 U.S. 826, 94 S.Ct. 131, 38 L.Ed.2d 59 (1973).
The question of the Commission’s responsiveness is more complex.
Roads: A great portion of the trial concerned testimony about paving and maintenance of roads. The district court found that “blacks live scattered across most of the County.” 548 F.Supp. at 892. This finding is not clearly erroneous. Furthermore, the district court correctly concluded that even had it found that blacks tended to live on unpaved and fringe roads, such a conclusion would not show unresponsiveness without other evidence showing discrimination in providing road services. Id.
Appointment Policy: All county boards are integrated except for one that is all black. 13 Rec. at 3121-22 (Jones). While the Commission did appoint two whites to replace deceased commissioners, this action was insignificant since one replacement was the brother of the deceased commissioner and the other was someone who had once run for the office.
Funding: Evidence shows that the Commission has funded or sought federal funding for a variety of projects that have benefitted the black community, including drainage projects, water service, site preparation for industry, a regional comprehensive mental health center, the county health department, and recreational facilities. See generally 548 F.Supp. at 890-98.
After reviewing the entire record, we conclude that the district court’s finding that the Commission is responsive to the needs of blacks is not clearly erroneous.
The other subsidiary factor listed under amended § 2 is the tenuousness of the state policy or political subdivision policy underlying voting procedures. The district court determined that there was a firm and long-standing state policy favoring the at-large election of the Dallas County Board of Education as well as the boards of the other 34 counties covered by Ala. Code § 16-8-1 (1975). 548 F.Supp. at 859. This finding is correct.
The court also concluded that although there was a long-standing policy favoring at-large elections of the County Commission in Dallas County, the policy state-wide was neutral. 548 F.Supp. at 910. The finding and conclusion are correct.
The Senate Report stated that the Zimmer factors were typical but not exclusive. 1982 Senate Report at 28-29, reprinted in 1982 U.S.Code Cong. & Ad. News at 206-07. The district court did not limit its consideration to the Zimmer factors, and we turn to other factors the district court analyzed. The court correctly determined that there was no property requirement for candidates for the Board of Education or the Commission and that this factor supported a finding of no dilution, 548 F.Supp. at 857, 908; that Dallas County had been designated for use of federal registrars and that this factor weighed in favor of dilution, id. at 857; and that there was no evidence of disqualification of black candidates and that this sub-issue supported a finding of no dilution,9 id. at 857, 908.
IV. INTENT
Because we cannot resolve this case on the statutory claim, we must turn *1541to the constitutional challenge to 1901 Ala. Acts 328(6), the section under which the Commission is elected.10 To prove a claim under the Fourteenth Amendment11 the plaintiff must show discriminatory intent in the enactment or maintenance of the statute. Bolden, 446 U.S. at 66, 100 S.Ct. at 1499. The district court determined that the government had failed to show discriminatory intent. 548 F.Supp. at 912-14.
The district court found “as a fact that the adoption of the at-large system was motivated by one human desire as old as the hills: the group which was out of power-wanted to get back in power.” Id. at 913. This conclusion is clearly erroneous. As the Commission’s expert, Dr. Jones, testified, the Democrats were in power in 1900-01 in Dallas County and had no fear of losing their control. 16’ Rec. at 3925-26. In reaching its conclusion the district court may have been thinking of the 1876 change to an appointive system rather than the 1901 change to an elected at-large system. In 1876 the Democrats were returning to power that they had lost to the Republicans after the Civil War; however, the Democrats had solidified their control in Dallas County before 1901. Because the district court erred in finding that 328(6) was enacted because Democrats sought to regain power in the county, the case must be remanded on this issue as well as on the statutory claim.
To aid the district court in its consideration on remand, we note that, in analyzing the intent issue, it stated:
[T]he plaintiff’s witness testified to the opinion that race was the sole or predominant motivation. This court believes the evidence more clearly shows that economic as well as political considerations were dominant factors ____ The Court rejects the conclusions of the Plaintiff’s expert witnesses that disenfranchisement was solely racially motivated.
548 F.Supp. at 912-13. To find a violation of the Fourteenth Amendment plaintiff does not have to prove that racial discrimination was a “dominant” or “primary” motive, only that it was a motive. Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 265-66, 97 S.Ct. 555, 563-64, 50 L.Ed.2d 450 (1977). On remand, if the court reaches the constitutional claim, it should determine whether racial discrimination was a motivating factor.
V. CONCLUSION
Our analysis of the government’s claim under amended § 2 has shown that the district court was clearly erroneous on two of the six principal factors — polarized voting and structure of the election system. We have additionally determined that on remand the district court must reconsider the factor of lingering effects of past discrimination. Finally, we have concluded that the district court erred in finding that the enactment of 1901 Ala.Acts 328(6) was motivated by the Democrats’ desire to regain power. We therefore remand the case for further proceedings consistent with this opinion.
AFFIRMED in part, VACATED in part, REVERSED in part, and REMANDED.
. The Board of Education moved to dismiss the action because the government had not joined necessary and indispensable parties to the action as required by Fed.R.Civ.P. 19. The Board asserted that the governor, state attorney general, secretary of state of Alabama, and every member of a board of education elected under § 16-8-1 was a necessary and indispensible party because the government sought to have the state statute declared unlawful. 1 Rec. at 31-32. The government answered the Board’s motion by explaining that
[djefendant is incorrect in its assertion that "plaintiff has challenged the unconstitutionality [sic] of this state-wide act" (Defendant's Brief, at 3). This lawsuit challenges the at-large method by which the Dallas County Commission and Board of Education are elected. This lawsuit does not challenge the method of election of such bodies in any other county in the state. Although the Dallas County Board of Education is elected pursuant to the same statute that sets forth the election provisions for other county boards of education throughout Alabama, this lawsuit challenges only the election structure in Dallas County, Alabama.
Id. at 46-47 (footnote omitted).
After the close of plaintiffs evidence, the Board of Education moved for an involuntary *1533dismissal under Fed.R.Civ.P. 41(b). 2 Rec. at 270. This motion was still pending when the Supreme Court decided Bolden. After receiving briefs from the parties on the impact of Bolden the district court on July 31, 1981 dismissed the Board from the suit.
[T]he Dallas County School Board persuaded the Court that its arguments in regard to the continued presence of that defendant in the pending case was improper for, among other things, the evidence introduced by the Government in its case in chief and the stipulations of the parties as reflected in the pretrial order, demonstrated that the statute under attack was a statute of state-wide application, that the attack on the statute was only as it applied to Dallas County and that the statute itself had not resulted in discrimination as a whole, but on the contrary, blacks have been elected to a number of positions in various counties and that the Government’s effort to single shot Dallas County without attacking the whole structure was improper. The Court does not limit its findings to what is recited herein, but has directed that the school board prepare and furnish to the Court a [sic] proposed findings of fact and conclusions of law at which time this Court will again consider the matter and enter its determinations that will be dispositive of the school board’s liability-
Id. at 356-57. The court, however, did not enter judgment for the Board of Education until September 3, 1982, when it granted the Board’s Rule 41 motion and determined that final judgment on the merits should be rendered in favor of the School Board defendants and against the plaintiff. 3 Rec. at 538-39.
. We do not consider them exclusive.
. While the Senate Report outlines nine factors to be considered under the "totality of circumstances" approach, we, like the district court, consider together two of the factors, past discrimination in voting procedures and lingering effects from discrimination in health, education, and employment. Consequently, in the remainder of this opinion we refer to eight, not nine, factors.
. Both the government’s expert and the Commission’s expert found a high degree of polarized voting in elections between black and white candidates. Dr. Cottrell, the government’s expert, determined that ”[i]n all of the black/white candidate races save one, the racial polarization index was higher than in white/white races and probably more important to the Court that it ranged in black/white races in [sic] 37 to 75____ [T]he white/white races were all, save one, lower than any of the polarization indices in the black/white races.” 8 Rec. at 1579. Cottrell indicated that a racial polarization index of 40 or higher was significant. Id. at 1576. See also Gov’t’s Exhibit 47 (Dr. Cottrell’s analysis).
Using a different test, the Commission's expert, Dr. Voyles, ascertained that “of the races with black candidates ... there is a correlation figure that is significant when you compare race *1536with voting.” 9 Rec. at 1781. See also Comm’n’s Exhibit 1 (Dr. Voyles’s analysis). These statistics, using either expert's method, show that in every race between a white and a black candidate there was a high degree of racial polarization. This takes on added significance in light of the fact that blacks constitute 44.8 percent of the registered voters.
. Evidence in the record indicates that in 1974 some NDPA candidates were disqualified from the ballot by the county probate judge but ordered back on the ballot by a federal judge.. 7 Rec. at 1180-81 (Chestnut). The discussion in the record of this occurrence does not reveal whether these candidates ran for county commissioner or school board member or whether they ran for other offices not involved in this lawsuit. The record does show that only one black, Meadows, ran for school board member in the 1974 general election. No black ran for county commissioner in this election. 548, F.Supp. at 849. Dr. Cottrell, the government’s expert, testifies that the NDPA ran candidates in Dallas County in 1970 but not after. 9 Rec. at 1639. We therefore assume that the NDPA candidates disqualified by the probate judge in 1974 were not running for either county commissioner or school board member. However, even if Meadows was a NDPA candidate, our analysis would not change, as blacks lost 13 other races after 1970.
. We use the term "structure of election system" to encompass a variety of factors, including use of a majority requirement, significance of the primary, size and population of the district, use of numbered posts, existence or not of a residency requirement, and ability or not to use single-shot voting.
. In 1978 the law governing the Board of Registrars was amended. Prior to the change the Board was required to go to the beats periodically. Under the new law the Board is permitted, but not required, to go to the beats.
. The Board opens at either 8:30 or 9:00 a.m., closes at noon, reopens at either 1:00 or 2:00 p.m. and closes at 4:00 or 4:30 p.m. 5 Rec. at 402-03 (Foster); 14 Rec. at 3222 (Horne).
. There is some evidence that suggests that in 1974 the Probate Judge disqualified several black candidates from appearing on the ballot whom the federal court ordered be put back on. 7 Rec. at 1180-81 (Chestnut). However, questioning by counsel for the County Commission suggested that the black candidates had not met state requirements on qualifications and ballot procedures. Id. at 1181. In another incident a black who should have been disqualified because he was a convicted felon was allowed to run. 548 F.Supp. at 857. Because the government did not challenge the district court finding on this criterion and the record does not provide sufficient information about the 1974 incident, we conclude the district court’s determination that there had been no evidence of disqualification of blacks is not clearly erroneous.
. As explained supra, the court did not permit the government to introduce evidence of discriminatory intent in the enactment of § 16-8-1 because the government had challenged the statute only as applied.
. On appeal the government challenged only the district court’s determination that it had not proven a violation of the Fourteenth Amendment. We therefore do not reach the issue whether the Fifteenth Amendment proscribes vote dilution, which.is an open question. See Jones v. City of Lubbock, 727 F.2d 364, 370 (5th Cir.1984).