Full opinion text
ORDER CLARKE, District Judge. This case is before the Court on remand from the Fourth Circuit Court of Appeals. Plaintiffs, Herbert M. Collins; Dr. H. Marks S. Richard; Barbara C. Parham; William E. Swindell, Jr.; Dr. Milton A. Reid; George Banks; Julian Hazel; and the Norfolk Branch of the National Association for the Advancement of Colored People (NAACP) sued defendants, the City of Norfolk, Virginia; the Norfolk City Council; City Council members Vincent J. Thomas, Joseph A. Leafe, Dr. Mason C. Andrews, the Reverend Joseph N. Green, Jr., Claude J. Staylor, Jr., Robert E. Summers and Elizabeth M. Howell; the City of Norfolk Electoral Board; and Electoral Board members Paul D. Fraim, Martha H. Boone, and Paul M. Lipkin, alleging that Norfolk’s at-large system of electing City Council members improperly diluted black voting strength in violation of the Fourteenth and Fifteenth Amendments to the United States Constitution, Amended Section 2 et seq. of the Voting Rights Act of 1965, 42 U.S.C. § 1973 et seq., and 42 U.S.C. § 1983. I. STATEMENT OF THE CASE The following facts were found by this Court at the original trial and were not disputed on appeal. Plaintiffs, except for the Norfolk NAACP, are black registered voters of the City of Norfolk. Plaintiffs Collins, Reid and Banks, and plaintiffs’ attorney James F. Gay are unsuccessful candidates for election to the Norfolk City Council; Collins and Gay were endorsed in their unsuccessful campaigns by plaintiff Dr. Milton Reid, publisher of the newspaper, The Journal and Guide. They were not, however, endorsed by the Concerned Citizens of Norfolk (CCN), the then dominant black political organization in Norfolk. Plaintiff Swindell is an unsuccessful candidate for the House of Delegates from Norfolk. Plaintiffs Collins, Richard, Parham and plaintiffs’ attorney Gay are former Norfolk City Democratic Committee members who were removed from their positions shortly before filing the instant action. This Court entered judgment for the defendants, Collins v. City of Norfolk, 605 F.Supp. 377 (E.D.Va.1984), holding that plaintiffs had failed to prove their claims under the Fourteenth and Fifteenth Amendments and Section 1983, and further holding that plaintiffs had failed to establish their vote dilution claim under the Voting Rights Act, Section 2 as amended, 42 U.S.C. § 1973. The Fourth Circuit affirmed. Collins v. City of Norfolk, 768 F.2d 572 (4th Cir.1985). Plaintiffs petitioned the United States Supreme Court for a Writ of Certio-rari. One week after its decision in Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986), the Supreme Court, without considering the merits of the case, remanded Collins to the Fourth Circuit “for further consideration in light of Thornburg v. Gingles.” Collins v. City of Norfolk, — U.S. -, 106 S.Ct. 3326, 92 L.Ed.2d 733 (1986). The Fourth Circuit remanded the case to this Court for reconsideration of the issues of racial polarization, minority electoral success, candidate slating, and government responsiveness in the East Ghent redevelopment. Collins v. City of Norfolk, 816 F.2d 932 (4th Cir.1987). Plaintiffs did not appeal the disposition of their constitutional and Section 1983 claims, and these matters are therefore no longer before this Court. Plaintiffs now seek a declaratory judgment that the at-large system of electing Norfolk’s City Council members unlawfully dilutes black voting strength; an injunction prohibiting the holding of future City Council elections under the-at-large system; and the replacement of the at-large system with a system of wards or single-member districts. Except for the issues enumerated here and discussed below, this Court’s findings in 605 F.Supp. 377 were unchallenged or were affirmed on appeal, and are therefore the settled law of the case. A. Issues to be Determined on Remand On remand, this Court must reconsider the following factors in light of Thornburg v. Gingles: 1. Whether analysis of the voting patterns of black and white voters over a period of years reveals legally significant racially polarized voting in the Norfolk City Council elections. 816 F.2d at 936 and 936 n. 3. 2. Whether black electoral success in Norfolk has been so consistent (and nearly proportional) as to overshadow any racially polarized voting and other factors tending to dilute minority access. 816 F.2d at 937. 3. Whether the success of a second black candidate in 1984, after the filing of this lawsuit, resulted from unusual circumstances, such as an attempt by white officials to forestall Voting Rights Act litigation. 816 F.2d at 938. 4. Whether any group of white citizens involved in the Norfolk City Council elections acted as a formal or informal slating organization, controlling access to the ballot, and adversely affecting black citizens’ access to the electoral process. 816 F.2d at 939. 5.Whether the removal of mostly black, low-income families from the East Ghent area of Norfolk was unresponsive to the needs of the minority community. 816 F.2d at 939. B. Settled Factual and Legal Issues The following findings of this Court were not designated for consideration on remand, are now the law of the case, and will not be reconsidered here: 1. Despite past acts of official discrimination, Norfolk’s black citizens today participate in the electoral process equally with white citizens, and have higher registration and turnout rates than do whites. The procedures and facilities of the Registrar’s Office are fair and are equally accessible to all of Norfolk’s citizens. 605 F.Supp. at 385. 2. Norfolk has no unusually large election districts, majority vote requirements, anti-single shot provisions or other voting practices or procedures that might enhance the opportunity for discrimination against the minority group. 605 F.Supp. at 389-90. 3. Norfolk’s black citizens are not hindered in their participation in the electoral process by socio-economic disparities in such areas as education, employment and health. 605 F.Supp. at 392. 4. Norfolk’s political campaigns have not been characterized by either subtle or overt racial appeals. 605 F.Supp. at 393. 5. The at-large system in effect in Norfolk and in a large majority of other cities in Virginia is not the result of a tenuous state or local policy. 605 F.Supp. at 399. Both plaintiffs and defendants having submitted briefs and memoranda, and oral argument having been heard on the matter, this case is now ready for determination. For the reasons set forth below, the Court finds that the voting patterns in Norfolk do not reflect legally significant racially polarized bloc voting within the Gingles standard; that Norfolk's black voters have usually been able to elect their preferred candidates; that black electoral success in Norfolk has been so consistent and nearly proportional as to preclude a finding of vote dilution under Section 2 of the Voting Rights Act and to overshadow any racially polarized voting and other factors, if those factors had been found; that the success in 1984 of a second black candidate after the filing of this lawsuit did not result from unusual circumstances or from any action or inaction by white officials or political leaders tending to forestall Voting Rights Act litigation; that no group of white citizens involved in the Norfolk City Council elections acted as a formal or informal slating organization; that no such group controlled access to the ballot or adversely affected black citizens’ access to the electoral process; and that the East Ghent relocation process was not unresponsive to the needs of the minority community. This Court therefore finds for the defendants on all grounds. All facts found in 605 F.Supp. 377 and not disturbed on appeal are considered as established facts in this opinion. II. DISCUSSION OF ISSUES ON REMAND Much of the evidence overlaps between issues in this case. For the sake of clarity, the Court will not follow the sequence of issues on remand, but will arrange its discussion of issues by topic. A. The 1982 Amendments to Section 2 Of The Voting Rights Act Section 2 of the Voting Rights Act of 1965, as amended, 42 U.S.C. § 1973, prohibits any voting practice or procedure “imposed or applied ... 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....” A violation of this Section “is established if, based on the totality of the circumstances, it is shown that the political processes leading to nomination or election ... are not equally open to participation by members of a [protected] class of citizens ... 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.” See Thornburg v. Gingles, 106 S.Ct. at 2759. B. The Senate Report Factors The Senate Report, S.Rep. No. 417, 97th Cong., 2d Sess. 2, reprinted in 1982 U.S.Code Cong. & Ad.News 177, 179 [the Senate Report] which accompanied the bill amending Section 2, “espouses a flexible, fact-intensive test for § 2 violations.” Gingles, 106 S.Ct. at 2764. At-large voting is not a per se violation of the Act, and the Act explicitly denies that any group is entitled to proportional representation. While the Senate Report provides that “[t]he extent to which members of a protected class have been elected to office” is one circumstance to be considered, the Report also provides specifically “[t]hat nothing in this Section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.” 106 S.Ct. at 2759-60. Proof of discriminatory intent is not required to establish a violation of Section 2. Rather than provide a hard and fast test, the Senate Report enumerates the following “typical factors,” derived from White v. Regester, 412 U.S. 755, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973) and from the Fifth Circuit’s opinion in Zimmer v. McKeithen, 485 F.2d 1297 (1973), aff'd sub nom East Carroll Parish School Board v. Marshall, 424 U.S. 636, 96 S.Ct. 1083, 47 L.Ed.2d 296 (1976) (per curiam): 1. the extent of any history of official discrimination in the state or political subdivision 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: 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, 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. Gingles, 106 S.Ct. at 2759-60 and 2759 n. 4 (quoting S.Rep. at 28-29). A court must assess the impact of a contested structure or practice “on the basis of objective factors,” such as the Senate Report factors. Id. at 2759-60 (citing S.Rep. at 27, 1982 U.S.Code Cong. & Ad.News pp. 177, 205). However, it is not necessary that a certain number (or even a majority) of factors point one way or another. S.Rep. at 29, 1982 U.S.Code Cong. & Ad.News p. 177. C. Gingles v. Edmisten and Thornburg v. Gingles Gingles v. Edmisten, 590 F.Supp. 345 (E.D.N.C.1984) rev’d in part and aff'd in part sub nom Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986), establishes the standard for interpreting Section 2 and the Senate Report factors. In Gingles, black North Carolina voters sued various North Carolina officials, claiming that North Carolina’s legislative districting scheme resulted in impermissible vote dilution and violated their rights under the Fourteenth Amendment to the United States Constitution, and under Amended Section 2 of the Voting Rights Act of 1965, 42 U.S.C. § 1973. Specifically, the Gingles plaintiffs alleged that the plan submerged black voting minorities in multi-member House Districts 36, 39, 23, 21 and 8; and multi-member Senate District 22. They further alleged that the plan fractured a geographical concentration of blacks between more than one Senate District, including Senate District 2, so that there was no effective voting majority of black citizens in any one district. 590 F.Supp. at 349-50. The Gingles three-judge district court relied on the Senate Report factors in determining the statutory claims, applying the “totality of the circumstances” test set forth in Section 2(b). It found that the North Carolina redistricting plan violated Section 2 by diluting the black citizens’ vote in all seven of the challenged districts. 590 F.Supp. at 376. See also 106 S.Ct. at 2760. The Supreme Court affirmed as to six of the districts, but a plurality of six Justices writing in three opinions reversed as to House District 23. Although the district court had found violations of all but one of the Senate Report factors in that District, persistent black electoral success in District 23 precluded the plaintiffs from prevailing in that district. The plurality establishes that proportional representation bars a Section 2 claim. However, nothing in Gingles creates an entitlement to proportional representation, or establishes less-than-proportional representation as proof of a violation. A comparison of the application of the Senate Report factors in Gingles and in Collins will therefore be instructive. 1. Lingering Effects of Past Official Discrimination and Discriminatory Voting Practices The Gingles district court found a long history of official discrimination in North Carolina’s State Senate and House elections, with poll taxes, numbered-seat plans and anti-single shot provisions in effect into the late 1960’s. As late as 1970, a literacy-test requiring black voters to read and write the Constitution was in effect in some North Carolina districts. 590 F.Supp. at 359-60. Since 1915, North Carolina had had a majority vote requirement for primaries, making it less likely that minority-preferred candidates could finally win elections. 590 F.Supp. at 363. As a result of this official discrimination, the district court found that, at the time of trial, “black voter registration remains depressed relative to that of the white majority,” id. at 361, and that lingering effects of discrimination in facilities, education, employment, housing and health care continued to hinder North Carolina’s black voters in their ability to participate effectively in the political process. Id. The court also found that the percentage of black registered voters in North Carolina was far below the percentage of blacks in the population. Id. at 357. In Norfolk, by contrast, there is no majority vote requirement. Black voters exceed white voters in registration and turnout rates. 605 F.Supp. at 385. Despite an early history of discrimination against blacks in Virginia, there are no lingering effects of discrimination in facilities, education, employment and health care affecting their ability to participate in the electoral process. Id. at 392. Norfolk’s black citizens today participate in the electoral process equally with white citizens, and the procedures and facilities of the Registrar’s Office are fair and equally accessible to all of Norfolk’s citizens. Id. at 385. Norfolk has had no discriminatory voting practices at any time relevant to this litigation. Id. at 389-90. 2.Overt or Subtle Racial Appeals Appeals to racial prejudice were prevalent in North Carolina's elections at the time of trial, and had been widespread and persistent for the past thirty years. 590 F.Supp. at 364. Some racial appeals by candidates and supporters were overt and blatant, while others were more subtle and furtive. Id. However, the specific examples of racial appeals in the North Carolina elections “reveal an unmistakable intention by their disseminators to exploit existing fears and prejudices and to create new fears and prejudices on the part of white citizens in regard to black citizens and to black citizens’ participation in the political processes of the state.” Id. No such racial appeals were present in the Norfolk City elections. 605 F.Supp. at 392-93. 3. Whether the State’s Policy is Tenuous The Supreme Court found the policy behind the North Carolina plan to be tenuous. The policy behind Norfolk’s at-large election system is not tenuous. The Gingles redistricting plan, unlike the Norfolk at-large election plan, was put into place after the Voting Rights Act was passed. 590 F.Supp. at 350. The policy behind the original North Carolina plan was that no county should be divided, in accordance with the North Carolina Constitution. However, the revised redistricting plan at issue in Gingles divided some counties. Id. at 351. In that case, where North Carolina had already abandoned the “no-division-of-counties” policy for some of the districts, any policy against dividing the remaining counties was clearly tenuous. “[S]uch a policy obviously could not be drawn upon to justify, under a fairness test, districting which results in racial vote dilution.” Id. at 373-74. Norfolk’s policy, on the other hand, “is consistent with both national and state practices for cities with the council-manager form of government” and is not the product of a tenuous policy. 605 F.Supp. at 399. 4. Legally Significant Racially Polarized Voting Under the Gingles Standard Although all of the Senate Report factors are important, the Gingles district court identified as the “linchpin of vote dilution by districting” a pattern of racial bloc voting consisting of [t]he demonstrable unwillingness of substantial numbers of racial majority to vote for any minority race candidate or any candidate identified with minority race interests. 590 F.Supp. at 355 (citations omitted). The mere existence of an at-large plan, and the mere fact that blacks are a minority in a given multi-member district does not establish that vote dilution has resulted from the districting plan. Id. Nor is a showing that blacks are a voting or population minority in a multi-member district, nor a showing that “blacks have not been elected ... in numbers proportional to their population” sufficient to establish a violation. Id. While all of the Senate Report factors “may be relevant to a claim of vote dilution through submergence in multi-member districts, unless there is a conjunction of the following circumstances, the use of multi-member districts generally will not impede the ability of minority voters to elect representatives of their choice.” 106 S.Ct. at 2765-66 (emphasis added) (footnote omitted). The sine qua non of a plaintiff’s Section 2 vote dilution claim, as set forth in Gingles, is that “a bloc voting majority must usually be able to defeat candidates supported by a politically cohesive, geographically insular minority group.” Id. at 2766 (emphasis in original) (citations omitted). It is “the usual predictability of the majority’s success [that] distinguishes structural dilution from the mere loss of an occasional election.” Id. at 2767 (citations omitted). In making this determination, courts are not to consider whether significant elements in the racial minority disagree with the plaintiffs’ requested judicial remedy of single-member districts in which the racial minority constitutes a majority; whether creating “safe” black majority districts would perpetuate racial ghettos and racial polarization; or whether the remedy creates the risk of disrupting the normal processes of acquiring political power. 590 F.Supp. at 356-57 and 356-57 nn. 17-20, and authorities cited therein. a. Threshold Considerations in Polarized Voting Claims The theoretical basis for a Section 2 claim is that where minority and majority voters consistently prefer different candidates, the majority, by virtue of its numerical superiority, will regularly defeat the choices of the minority voters, and will minimize or cancel out the minority voters’ ability to elect their preferred candidates. 106 S.Ct. at 2765. “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 2766 (footnote omitted). “Second, the minority group must be able to show that it is politically cohesive.” Id. 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.” It is “[i]n establishing this last circumstance, [that] the minority group demonstrates that submergence in a white multi-member district impedes its ability to elect its chosen representatives.” It is the “usual predictability” of the majority’s success that distinguishes vote dilution from the mere loss of an occasional election. Id. In considering these circumstances under the Gingles standard, the Fourth Circuit noted that this Court should not consider causative factors such as backlash and field limitation in determining whether there is racially polarized voting. Rather, this Court should inquire separately into minority and majority voting, to see whether members of the minority usually vote for the same candidates — if so there is the requisite minority political cohesiveness —and to see whether the majority vote is sufficiently homogeneous to cancel the minority vote plus the cross-over vote. These inquiries together determine whether there is a “pattern of racial bloc voting that extends over a period of time,” a critical factor in a claim of vote dilution under § 2. 816 F.2d at 935-36 (emphasis added). As to the first consideration, the Court finds that Norfolk’s black population is sufficiently large and geographically compact to constitute an effective voting majority of 65 percent in no more than two of the proposed wards. See 590 F.Supp. at 358 n. 21; see also Tr. 1661, 1702 (Testimony of William P. Robinson, Jr.). While the Court recognizes that, in Norfolk, black voter registration and participation is far greater than in Gingles, it is noteworthy that in Gingles, defendants’ expert witness testified that a 65 percent black majority was necessary to create a “safe” black district, and that the Justice Department uses 65 percent as a benchmark for their administration of the Voting Rights Act. It is doubtful that establishing a ward system would produce a greater black presence on Norfolk’s City Council than is already there. Plaintiffs’ Exhibit 52 sets forth plaintiffs’ proposals for ward boundaries designed to give blacks three safe seats. The proposal sets 32,614 as the ideal population per ward. This excludes, however, the military population of 38,680 — a larger segment of the population than that contained in any proposed ward. Additionally, none of the proposals provides for a 65 percent or better black population in as many as three wards. It is highly doubtful that a plan excluding military personnel, who are included in census figures, would pass constitutional muster. See Reynolds v. Sims, 377 U.S. 533, 563-68, 84 S.Ct. 1362, 1382-85, 12 L.Ed.2d 506 (1964). If the military were included, none of the five proposals would have any validity. As to the second consideration, the Court finds that the record is clear that Norfolk’s blacks are politically cohesive. See Def.Ex. 470 (A-I) and 1986 Stip. However, the Court finds, for reasons stated below, that Norfolk’s whites do not vote sufficiently as a bloc that they usually defeat the minority’s preferred candidate. The third element is therefore lacking. b. White Bloc Voting: The Test for Legal Significance The Supreme Court in Gingles did not set forth a “lucid explication of all features of the amended § 2,” 816 F.2d at 935, nor did it establish a simple, doctrinal test for legally significant bloc voting. It did make clear that a mere preference for different candidates among voters of different races is not legally significant. The Gingles district court found, based on statistical evidence presented by qualified experts, that blacks and whites generally preferred different candidates, and that the correlation between the voters’ race and choice of candidates was “substantively” significant. 590 F.Supp. at 368. See also 106 S.Ct. at 2768. Under the Supreme Court’s Gingles rule this establishes racially polarized voting, and is sufficient to show the cohesiveness of the black minority. However, after a careful reading of the Supreme Court’s Gingles opinion and the Fourth Circuit’s opinion on remand, it is clear that polarized voting is not legally significant unless the white bloc voting usually defeats minority-preferred candidates despite the combined votes of blacks and white “cross-over” votes. The Supreme Court found that while the Gingles district court did not use the phrase “legally significant,” the record was sufficient to show that the polarized voting in North Carolina met the Gingles test for legal significance. 106 S.Ct. at 2771 and 2771 n. 27. The evidence relied on in Gingles included statistical studies of 53 elections between 1978 and 1982 in which black candidates ran. The Gingles district court found that in no North Carolina election, primary or general, did a black candidate receive a majority of white votes cast, and that an average of 81.7 percent of white voters did not cast any votes for any black candidates. White voters in the general elections consistently ranked black candidates last or next-to-last, except in heavily Democratic areas. There, white voters still rated blacks last or next-to-last among Democrats, if not last or next-to-last among all candidates. Furthermore, the district court found that in North Carolina approximately 66 percent of all white (Democratic) voters would not vote for a black Democrat even after the candidate had won the Democratic primary and the only alternatives were to vote for a Republican or no one. 590 F.Supp. at 368. (Gingles does not indicate that there were any black-preferred candidates who were not black.) Only “some” black incumbents were reelected, and the black incumbents did not receive white majority support even when the election was uncontested. Id. at 368-69. The Gingles court noted that black participation in North Carolina’s political process had recently increased markedly, but that [t]he participatory level of black citizens is still minimal in relation to the overall black population, and ... is largely confined to the relatively few forerunners who have achieved professional status or otherwise emerged from the generally depressed socio-economic status which ... remains the present lot of the great bulk of black citizens. Id. at 372. Similarly, in City of Petersburg, Virginia v. United States, 354 F.Supp. 1021 (D.D.C.1972), aff'd, 410 U.S. 926, 93 S.Ct. 1441, 35 L.Ed.2d 698 (1973), the United States District Court for the District of Columbia analyzed black and white voting patterns in Petersburg’s at-large elections, to determine whether annexation of a predominately white area would impermissibly dilute black votes. In Petersburg, the City sought a declaratory judgment under Section 5 of the Voting Rights Act to determine the validity of the proposed annexation, which would have increased the white population by almost 50 percent, and would have eliminated Petersburg’s black population majority. The district court found that Petersburg had experienced a “dramatic polarization of the races,” and also found “almost total bloc voting by race” in the Petersburg elections. Id. at 1025. In the all-white precincts, no more than 8.5 percent of the total vote was ever cast for black candidates in a given election, while in the black precincts the combined total of all votes cast for white candidates never exceeded 19 percent. In contrast, white vote for white candidates never fell below 91 percent or black vote for black candidates below 81 percent. Id. at 1026 n. 10. Two black candidates had won separate elections while whites held a voting majority, but [w]hen there was a possibility that by re-electing [a black incumbent] and electing another black candidate, the black members would control the council, there was an unusually large turnout of white voters in the white wards, as a result of which [the black incumbent] was defeated and the second black candidate also lost. Id. at 1026 (footnote omitted). Similarly, in City of Rome v. United States, 472 F.Supp. 221 (D.D.C.1979) (three judge court), aff'd, 446 U.S. 156, 100 S.Ct. 1548, 64 L.Ed.2d 119 (1980), a black candidate defeated three whites in the primary but was defeated in the general election when whites, turning out in unusually high numbers, apparently “coalesced to defeat [the black candidate]” whose election appeared imminent. 472 F.Supp. at 227. The Supreme Court found that the black candidate’s defeat in these circumstances was indicative of racial bloc voting. 446 U.S. at 184 n. 20, 100 S.Ct. at 1565 n. 20. See also Butts v. City of New York, 614 F.Supp. 1527, 1545-46 (S.D.N.Y.1985) (in the one election year where a minority candidate appeared on the run-off slate, he lost overwhelmingly to a white candidate in an election unusual for its high voter turnout). The situation in Norfolk is quite different. In Norfolk, as discussed below, the white cross-over vote has combined with the black vote to elect the black-preferred candidates most of the time. Norfolk does not have the extreme and entrenched pattern of bloc voting found in Petersburg, Gingles and the other cases cited above. It is undisputed that the Norfolk City Council elections are nonpartisan, at-large elections under a simple plurality vote system, in which all candidates compete for open seats. The four-year terms for the seven-member council are staggered, so that every two years either three or four seats (and in one case, six) are available. Fields of candidates have ranged from eight to twenty-one, with an average of thirteen. These large fields tend to fragment whatever white bloc voting might occur, and work to the advantage of black voters, who tend to concentrate their votes. See Tr. 1968 (Testimony of Dr. O’Rourke); Def.Ex. 470 (A-I); 1986 Stip. White cross-over voting for a black candidate has been above 20 percent in 70 percent of the Norfolk elections. In half of the elections, including the three most recent, white cross-over voting for a black candidate has exceeded 25 percent. Def.Ex. 474-A; 1986 Stip. White crossover support in elections from 1968 through 1986 has averaged almost 25 percent for the 19 black-preferred candidates (that is, those candidates who received over 50 percent of the black vote in a given election). This is more of the white vote than over half of the white candidates have received. Def.Ex. 470; 474-A; 1986 Stip. This 25 percent white cross-over vote is even more impressive when considered in light of the relative numbers of blacks and whites in the population. For example, a black candidate, such as Reverend Green, who received 92 percent of the black vote in 1978, received approximately 11,000 black votes. His actual 26 percent white cross-over vote means that approximately 5,600 to 5,700 white voters joined with the 11,000 blacks to ensure the success of the black-preferred candidate. See Def.Ex. 470 (A-I); 1986 Stip. Thus, he received approximately one-third of his total vote from white voters. In Norfolk, black registration and turnout percentages have remained at or above the level for whites since 1980. 816 F.2d at 939; 768 F.2d at 574-75; 605 F.Supp. at 385, 402, 404. In Gingles, white cross-over vote was virtually nonexistent, and even those black-preferred candidates who received the most white cross-over votes usually lost. In Norfolk, those candidates have usually won, as will be more fully discussed below. The Court finds, based on the evidence presented, that in Norfolk’s City Council elections, black-preferred candidates usually win, and that white bloc voting does not usually defeat the black voters’ preferred candidates. Plaintiffs have failed to show that the at-large system “operates to minimize or cancel out” their ability to elect their preferred candidates. 106 S.Ct. at 2765 (citations omitted). They have therefore failed to meet the threshold test of Section 2, the establishment of legally significant, racially polarized bloc voting. Plaintiffs' Section 2 claim must fail. 5. Black Electoral Success Under The Gingles Standard Although the election of a few minority candidates does not necessarily foreclose a vote dilution claim, the extent to which members of a protected class have been elected to office is one circumstance which may be considered. 106 S.Ct. at 2779 (citations omitted). This Court’s duty is to conduct an independent consideration of the record and to make “a searching, practical evaluation of the past and present reality in determining whether [Norfolk’s] political processes are equally open to all voters.” Id. (citations omitted). The Supreme Court in Gingles adopted what is in essence a new defense to a Section 2 claim: that roughly proportional minority electoral success will defeat a Section 2 claim even if such factors as racially polarized voting are present. Accordingly, the Fourth Circuit asked this Court to determine [w]hether, as a factual matter, black electoral success in Norfolk has been so consistent and nearly proportional as to overshadow any racially polarized voting and other factors tending to dilute minority political access. 816 F.2d at 937. In making this determination, a comparison of black electoral success in the challenged North Carolina districts with black electoral success in Norfolk will be helpful. a. Black Electoral Success in North Carolina and Norfolk The Gingles court found a significant disparity between the electoral successes of blacks and of whites. At the time of trial (in 1984), North Carolina’s black citizens constituted 22.4 percent of the State’s total population, but held only 9 percent of the city council seats; 7.3 percent of county commission seats; 4 percent of sheriff’s offices; and 1 percent of the offices of the Clerk of the Superior Court. There were 19 black mayors, 13 of whom were from majority black communities. Approximately 40 percent of the 9 percent of city council members who were black were from majority black municipalities or districts. Thus, in Gingles, only 5.4 percent of all city council members were blacks elected in areas where whites outnumbered blacks, although blacks made up 22.4 percent of the population. 590 F.Supp. at 365. In contrast, Norfolk’s City Council is at present, and has been since the 1984 election, 28.5 percent black. This figure is far more nearly proportional to Norfolk’s 35.2 percent black population than the 5.4 percent figure is to North Carolina’s 22.4 percent black population. See 605 F.Supp. at 382. Between 1971 and 1982, black representation in North Carolina’s legislature ranged between 1.6 percent and 3.3 percent of the House seats, and 2 percent and 4 percent of the Senate seats, although the electorate was 22.4 percent black. 590 F.Supp. at 365. By contrast, from 1970 to 1982, Norfolk’s City Council ranged from Í4.3 to 28.6 percent black. More importantly, black voters in Norfolk between 1968 and 1986 elected their first or second choice 100 percent of the time, in fields of as many as 21 candidates. See Def.Ex. 470(A-I); 1986 Stip. From 1970 to the present, Norfolk’s City Council has included at least two, and sometimes three, members who were endorsed by black civic groups and who received majority black support. 605 F.Supp. at 400. All black incumbents who have run have been reelected, a success rate of 100 percent, while the white incumbents’ success rate has been only 80 percent. See Def.Ex. 470(A-I); 1986 Stip. In North Carolina, black-preferred candidates who won Democratic primaries in the challenged districts were three times as likely to lose the general election as were their white counterparts. 590 F.Supp. at 365. No blacks had ever been elected to the State Senate from the area encompassed by House Districts 2, 8, 23 and 39. Id. at 366. Only one black had been elected in this century to the State Senate in Senate District 22, and from the area encompassed by House District 21, although in the latter case the Senator was elected for two terms. Id. at 365-66. Through 1983 there were only one or two blacks at any time in the 50-member State Senate, a mere 2 percent to 4 percent, although blacks then constituted about 22 percent of the total North Carolina population. Id. at 365. In the0 area encompassed by Senate District 22, there had been a total of nine other black candidates, seven for the House, and two for the Senate, all of whom been defeated. Id. In Norfolk, considering all candidates (and not just those who received majority support), 30.43 percent of all black candidates who have run for City Council since 1968 have won, compared with 29.24 percent of all white candidates. See Def.Ex. 470(A-I); 1986 Stip. North Carolina’s black candidates for the State House of Representatives fared only slightly better than candidates for the State Senate. From 1971 through 1982 there were never more than two to four blacks in the 120-member House of Representatives (a mere 1.6 to 3.3 percent of the House members). 590 F.Supp. at 365. No black Representative had ever been elected from House District 8, id. at 366, and only one from House District 36. Id. at 365. In House District 39, only one black Representative had ever been elected before the Gingles lawsuit. Two were elected after the suit was filed. Id. at 366. The court indicated that their election may have resulted from “unusual organized political support by white leaders concerned to forestall single-member districting — ” Id. at 367 n. 27. In House District 21, a black Representative was twice elected to the State House, and was serving at the time of trial. Id. at 366. In House District 23 (Durham), however, one black citizen had been elected every two years since 1973 to the State House, and 25 percent (3 of 12) of the Durham City Council members were black. Durham is 47 percent black. Id. at 366. A plurality of the Supreme Court held that this steady pattern of success foreclosed plaintiffs’ claims of vote dilution, notwithstanding that the black representation in Durham was not mathematically proportional. 106 S.Ct. at 2780 (Brennan, J.); 106 S.Ct. at 2783 (White, J.); 106 S.Ct. at 2793 (O’Connor, J.). Apart from District 23, however, blacks did not fare well in city council elections in the challenged North Carolina districts. Black voters in Charlotte (House District 36; Senate District 32), for example, were 31 percent of the total population, yet were only 5.4 percent of the Charlotte City Council’s at-large elected members before 1975. From 1977 to 1981, blacks won 28.6 percent of the district seats and only 16.7 percent of the at-large seats. 590 F.Supp. at 365. Norfolk’s blacks were 28.3 percent of the population until 1980, and approximately 35 percent since then. See 605 F.Supp. at 382; Def.Ex. 107 at Table P-1. Black-preferred candidates consistently held 28 percent of the City Council seats from 1968 to 1984, and presently hold 41 percent, all elected at-large. See Def.Ex. 470(A-I); 1986 Stip. After the Gingles trial, a black candidate was elected mayor of Charlotte, receiving approximately 38 percent of the vote. 590 F.Supp. at 366. In the area encompassed by House District 21, there had been a one-term black mayor in the early 1970’s, and one black city council member. There was one black on the County School Board, one on the seven-member Board of County Commissioners, and a twice-elected black sheriff. In House District 36 (Senate District 22), one black citizen was elected three times and defeated once for the Board of County Commissioners. 590 F.Supp. at 365-66. Only two blacks had ever been elected to the Board of Education. In House District 39, one black had been elected, then defeated twice, then re-elected to the otherwise all-white, eight-member Board of Education. Another had been elected, then defeated, then re-elected to the Board of County Commissioners. Id. at 366. In House District 8, a similar pattern emerged. There had been only two black members of the Board of County Commissioners in one of the three counties that comprise this district (both elected since 1982), and none in the other two counties. There was only one black Board of Education member, at the time of the Gingles trial, in a community that was 36.5 percent black. Id. at 366. The City of Wilson, North Carolina, located in House District 8, had one black councilman out of six, in a 40 percent black community. Likewise, in the area encompassed by Senate District 2, all black candidates who had run for the Halifax Board of County Commissioners and for the Roanoke Rapids City Council had lost. Id. Based on the record summarized above, the district court in Gingles found that the overall results achieved to date by blacks at all levels of elective office in North Carolina were minimal in relation to the percentage of blacks in the total population. 590 F.Supp. at 367. This is not the case in Norfolk, as the evidence makes abundantly clear. b. Statistical Models of Racial Voting Patterns The City of Norfolk does not maintain voter registration or turnout figures by race, nor does the secret ballot permit observation of how a particular voter of a particular race voted. Tr. 721. Therefore, both plaintiffs and defendants used statistical estimates of voting behavior. 605 F.Supp. at 386. The only information from which these estimates can be drawn is the total number of people who register and vote in a precinct, and the total number of votes cast in a precinct for a particular candidate. These numbers are then applied to the precinct’s racial composition, based on census data, in order to estimate how many blacks and how many whites voted for specific candidates. Tr. at 883-84. The two standard statistical methods used to analyze group voting behavior are the “regression analysis” and the “homogeneous precinct”, or “extreme case” analysis methods. Both were employed at trial, Tr. 1834-36, 1948, and both were acknowledged by the Supreme Court in Gingles as standard techniques for use in voting rights cases. 106 S.Ct. at 2768 n. 20. The Fourth Circuit indicated that it believed that this Court did not admit plaintiffs’ regression analysis into evidence. Plaintiffs’ regression analyses were admitted into evidence as Exhibit P-50-b. Tr. 582. Nevertheless, this Court found that plaintiffs’ studies contained serious methodological flaws which severely undercut their probative value. See 605 F.Supp. at 387. The Court of Appeals for the Fourth Circuit did not challenge this Court’s finding of methodological flaws except to the extent that plaintiffs may have been asked to isolate variables other than race. On remand, the Fourth Circuit only asks this Court to clarify its position on plaintiffs’ methodology. See 816 F.2d at 936. During discovery and trial, plaintiffs presented three statistical reports by Mr. Brace and Mr. Peterson, and one by Dr. Engstrom. See Def.Ex. 99,103; Ex. P-50b. Each employed regression analysis and each yielded different turnout, registration and vote totals for the same elections. For example, in 1978 by a homogeneous precinct analysis black incumbent Joseph Green received 32.2 percent of the white vote. Def.Ex. 470(D). Plaintiffs’ regression analysis yielded a figure of 20.2 percent in the first report, Def.Ex. 99; 18.7 percent in the second report, Def.Ex. 103; and 26.5 percent in the third report. Ex. P-50-b. Again, using homogeneous analysis, Mason Andrews received 16.7 percent of the black vote in 1974. Def.Ex. 470(F). Using plaintiffs’ regression analysis, however, Andrews’ percentage for that same election changed from 17.4 percent in the first report, Def.Ex. 99, to 9.4 percent in the third report. Ex. P-50-b. The first report listed two black candidates as white candidates, contained errors in nomenclature, did not weight the computations to allow for differences in precinct size and excluded 38,000 mostly white, voting-age Navy residents. Tr. 475, 491-92, 495-96, 547 (Testimony of Carl B. Peterson). See Def.Ex. 99. Plaintiffs’ second report, produced to replace the first report, repeated the errors in calculations and the errors in the race of two candidates. This report weighted all computations by voting age population, which plaintiffs’ expert conceded was an error. The report still excluded 38,000 voting-age Navy residents. Def.Ex. 103; Tr. 495-520 (Testimony of Carl B. Peterson). Plaintiffs’ third report, produced four weeks before trial, weighted the data by precinct turnout, giving greater importance to election results in precincts with higher turnouts. Again, the Navy residents were “put in a hypothetical, nonexistent precinct, to eliminate them.” Tr. 475, 543 (Testimony of Carl B. Peterson). The exclusion of approximately 38,-000 predominately white, voting-age Norfolk residents residing on Norfolk’s Navy bases or aboard ships home-ported in Norfolk (a full 17 percent of Norfolk’s total population and a larger population than that of one of plaintiffs' “ideal” proposed wards), improperly skewed the results of their studies, inflating white registration and turnout figures, 605 F.Supp. at 387, and exaggerating the number of whites not voting for blacks. Tr. 527-28, 1842-43. The Fourth Circuit, in its first opinion, noted the impact of leaving out this large, mostly white population. 768 F.2d at 575 n. 6. These voters are eligible to run for office, are counted in the census figures for Norfolk, are considered by the State legislature in drawing House and Senate districts, and cannot properly be excluded from the voting calculations. Tr. 546, 1833; 605 F.Supp. at 387. This error was compounded in both the second and third reports by plaintiffs’ further error of including as “white” the approximately 8,000 to 9,000 Norfolk residents who are neither white nor black. See 605 F.Supp. at 387, Tr. 498-99. One candidate’s race was still wrong in the third report. Tr. 539. Another serious flaw in plaintiffs’ regression analyses was the use of “double regressions”, which led to such impossible results as figures indicating that a candidate received fewer than zero votes, or more than 100 percent of all votes. Tr. 943-48. These impossible results were then “normalized” by the introduction of a a “fudge factor”. Tr. 551-52, 948. Plaintiffs’ expert Dr. Engstrom admitted that he did not understand this “fudging” of figures. Tr. 948. The fourth report, produced two weeks before trial, relied on data in the earlier reports, repeating the erroneous exclusion of the military, and including as white those voters who were neither black nor white. Tr. 942. This last report considered only three elections, Tr. 984, and also contained certain methodological weaknesses. Tr. 1845-50. These methodological flaws are distinct from the issue of causative factors, and go directly to the reliability of the figures derived, and to the reliability of conclusions drawn from those figures. While regression analysis, properly performed, can yield probative results, the Court finds that the methodological weaknesses of plaintiffs’ regression analyses have so distorted the resulting participation and vote percentages for the various candidates that plaintiffs’ studies, and evidence based on these studies, can have only minimal probative value in this matter. This Court will, therefore, base its conclusions on the plaintiffs’ and defendants’ homogeneous precinct analyses, which agree as to the black precincts, and are extremely close as to the white precincts, and which do not contain the fundamental flaws the Court finds in plaintiffs’ regression analyses. The Court should, however, consider all statistical testimony with caution. See, for example, the testimony of plaintiffs’ expert Mr. Brace where he testified that if 100 percent of all white voters cast one of their four votes for a black candidate in a four-seat election, his table would show only a 25 percent white crossover vote. Tr. 800. c. Success of Black-Supported Candidates In Norfolk’s Elections The Supreme Court clearly contemplates that a “candidate of choice” of the black community may be of any race, and need not be black, as it consistently refers to the minority’s preferred candidate, or candidate of choice, rather than the minority candidate. Four of the Justices would have so held. 106 S.Ct. at 2775 (Brennan, J.). Justice White disagreed that the candidate’s race is per se irrelevant, and said it was not necessary to decide that issue. 106 S.Ct. at 2783-84. Justice O’Connor, writing for four Justices, agreed with Justice White that the candidate’s race was not always irrelevant, and also agreed that the Court need not decide the issue. 106 S.Ct. at 2793. No Justice said that a black-preferred candidate must be black. Plaintiffs’ expert Dr. Engstrom concurred. Tr. 992. The black community’s candidate of choice remains the candidate of choice even if he or she received white support, and even if the candidate could not have been elected without that white support. As Justice Brennan wrote for the plurality, “Under § 2, it is the status of the candidate as the chosen representative of a particular racial group, not the race of the candidate, that is important.” 106 S.Ct. at 2776 (emphasis in original). This Court must, therefore, ask who are the black community’s candidates of choice, and then ask whether white racial bloc voting usually defeats the combined black vote and white cross-over vote for these candidates. i. Majority Black Support More than half of all candidates receiving majority black support have been elected in the past eighteen years. Def.Ex. 470(A-I); 1986 Stip. With whites in Norfolk comprising two-thirds of the population, they could win every seat if they were in fact voting as a racially polarized bloc. On the contrary, the record indicates that between 1968 and 1986 white-preferred candidates have won only 25 of the 38 seats contested, or 65.79 percent. See Def.Ex. 470(A-I); 1986 Stip. This success rate corresponds almost exactly with the proportion of whites in the population: 66 percent. The substantial success rate of black-preferred candidates, electing 11 candidates for 38 contested seats (29 percent), is roughly equivalent to the proportion of blacks in the population (28 percent until 1980, and 35 percent after 1980). Def.Ex. 470(A-I); 1986 Stip. Therefore, both races are electing their preferred candidates in numbers roughly proportional to their percentages of the population. More than half of all black-preferred candidates (11 of 19), and 100 percent of black incumbents, have won. In Norfolk City Council elections, whites have voted for black-preferred candidates in numbers sufficient to enable them to be elected. Tr. 1952-53, 1970-71 (Testimony of Timothy J. O’Rourke). In 1986, under homogeneous precinct analysis, black voters elected their first and third choices, while white voters elected their first, second and fourth choices. In 1986, Randy Wright was the white voters’ third choice, under homogeneous analysis, with 60 percent of the white vote. (He was the whites' second choice under the 1986 stipulated regression figures, with 70 percent.) He lost the election because of a lack of black support, despite his preferred position among whites. 1986 Stip. Plaintiffs’ argument that candidates most preferred by the black voters were defeated in three of the five elections between 1974 and 1982 does not stand up to closer examination. The basis for their assertion is that their regression analysis, the defects in which are discussed above, purportedly shows that Mrs. Butts, who lost, received 93.1 percent of black-voter support and Reverend Green, who won, was second with 92 percent. However, plaintiffs’ expert testified at trial that the election estimate figures were “too close to call.” Tr. 917. In other words, the estimates for the two candidates were too nearly identical to afford grounds for plaintiffs’ statement. If plaintiffs’ own homogeneous analysis figures are used, however, Reverend Green is the first choice of black voters, over Mrs. Butts, which would produce three wins and two losses by black voters’ first choice (a 60 percent success rate) for that period. See Ex. P-50-b. Plaintiffs’ reliance on Exhibit P-185 (Table 11), which purports to show that between 1974 and 1982 there have been twelve black candidates for Norfolk City Council, of whom only three were elected, is also misleading. Six of the nine losing candidates, including plaintiffs Reid, Collins, Banks and plaintiffs’ counsel Gay, were rejected by the black community, leaving only three candidates who received a majority black vote but lost the election. These three were actually only two individuals: Mr. Bond in 1974, and Mrs. Butts in 1980 and 1982. Tr. 805-09 (Testimony of Kimball W. Brace). Mrs. Butts lost in 1980 by only 405 votes, and still ran ahead of the white incumbent, Conoly Phillips, who was defeated. Tr. 1978. The black community’s first choice candidate has been elected in six of the eight elections since 1972 (the cut-off date used in Gingles), establishing a success rate of 75 percent for the candidate of first choice (or 78 percent if the 1968 election is included). If the black community’s candidates of first and second choice are considered, the figure rises to a highly satisfactory 100 percent rate of success for first or second choice for the last eight consecutive elections. These results are even more probative of black electoral strength in view of the crowded fields of candidates in each election. See Def.Ex. 470(A-I); 1986 Stip. The elections in Norfolk contrast sharply with the elections in Gingles’ District 23, where half of the blacks elected were unopposed. 590 F.Supp. at 370. The Court finds unpersuasive, and indeed misleading, plaintiffs’ argument that some black-supported candidates “lost in six of the eight City Council elections since 1970 and lost in all three elections held in the 1980’s.” See Plaintiffs’ Memo on Remand at 38, Table 2. The record reflects, and this Court finds, that in each of the eight staggered term elections, at least one first or second choice black-supported candidate was elected; therefore, since 1974 there have been at all times at least two sitting Council members who received black majority support. See Def.Ex. 470(A-F); 1986 Stip. As the Fourth Circuit noted on remand, Unlike the situation in the disputed ... districts at issue in Gingles, ... black voter registration and turnout have recently increased in Norfolk. The district court acknowledged that fact, see e.g., 605 F.Supp. at 402, 404, as did this Court in its first opinion, see 768 F.2d at 574-55. This factor is probative, but not dispositive, of black electoral participation, and there is no reason to believe that the district court viewed it as inordinately important. 816 F.2d 939. At oral argument on remand, plaintiffs’ counsel represented to this Court that the Fourth Circuit had settled the issue of black electoral success in Norfolk as a matter of law. Transcript of December 16, 1987 Oral Argument (“O.A.”) at 59-60. Mr. Parker said, in response to questions from the Court, that the following language, which he quoted, was a ruling of law, binding upon this Court: The myriad numbers before this panel seem to indicate a consistent level of political success by Norfolk blacks that nevertheless is not as conspicuous as that achieved by the black community in North Carolina District 23. 816 F.2d at 937 (footnote omitted). Counsel argued that this language barred this Court from ruling on the matter. By omitting the sentence immediately following the portion he presented, and by omitting footnote 5, counsel seriously misstates the law of the case. The omitted sentence in the quoted passage makes it absolutely clear that the Fourth Circuit is not substituting its judgment on matters of fact for that of the District Court when it says: On remand, the district court must determine whether, as a factual matter, black electoral success in Norfolk has been so consistent and nearly proportional as to overshadow any racially polarized voting and other factors tending to dilute minority political access. Id. (emphasis added). The omitted footnote further undercuts plaintiffs’ interpretation of the Fourth Circuit’s remand opinion: The question is not, however, so simple as how many blacks versus whites were elected. The court must examine the parties’ studies of voting preferences to determine which were the preferred candidates of the majority and minority communities. Id. at n. 5. Plaintiffs’ counsel’s selective citation did not stop there. Counsel also argued that the Fourth Circuit had ruled that several candidates who had received a majority of the black vote in particular elections could not, as a matter of law, be counted as examples of “minority electoral success.” O.A. at 59-61. On this basis, plaintiffs’ counsel would exclude Mrs. Howell and Mr. Staylor as candidates of choice in 1974, when Mr. Bond received more black votes than they did, but lost, and would exclude Mrs. Howell in 1980 when Mrs. Butts received more black votes but lost. Id. at 61. In 1974, with CCN support, Mrs. Howell ran second out of twenty-one candidates among black voters, running ahead of three black candidates among black voters. In 1980, Mrs. Howell, with CCN endorsement, beat Mrs. Butts by over 2,300 votes, and received more than twice the black votes received by plaintiff George Banks. In 1984, Mrs. Howell again beat Mrs. Butts as a result of Mrs. Butts’ decline in black support. Def.Ex. 470(A, C, F). Delegate Robinson confirmed Mrs. Howell’s black support. Tr. 1668. Mr. Staylor, endorsed by CCN, received votes of 56.5 percent of all black voters in 1974, ahead of three black candidates, including plaintiff Milton Reid. Def.Ex. 470(F). Delegate Robinson testified that Mr. Staylor was supported by blacks in 1974, Tr. 1667, and Judge Jordan testified that Mr. Staylor “had the overwhelming support of the black community when he first ran for Council.” Tr. 1535. The evidence strongly indicates that Mrs. Howell in 1974 and 1980, and Mr. Staylor in 1974, were in fact candidates of choice of the black community and that any presumption created by the defeat of the first-choice candidates in those elections has been rebutted. Again, counsel misstates, by selective omissions, the law of the case, quoting that: The mere electi