Bradford County NAACP v. City of Starke, 712 F. Supp. 1523 (M.D. Fla. 1989)

US District Court for the Middle District of Florida - 712 F. Supp. 1523 (M.D. Fla. 1989)
February 27, 1989

712 F. Supp. 1523 (1989)

BRADFORD COUNTY NAACP, et al., Plaintiffs,
v.
CITY OF STARKE, et al., Defendants.

No. 86-5 Civ-J-12.

United States District Court, M.D. Florida, Jacksonville Division.

February 27, 1989.

*1524 *1525 David Lipman, Robert Weisberg, Miami, Fla., for plaintiffs.

Terrence Brown, Starke, Fla., for defendants.

 
FINDINGS OF FACT AND CONCLUSIONS OF LAW

HODGES, Chief Judge.

 
I. PROCEDURAL BACKGROUND

Plaintiffs instituted this action on January 3, 1986, seeking declaratory and injunctive relief against the at-large, city-wide municipal elections held for members of the Starke City Commission. Plaintiffs allege that the at-large election system unlawfully dilutes the voting strength of black citizens and has a discriminatory result in violation of Section 2 et seq. of the Voting Rights *1526 Act, 42 U.S.C. § 1973 et seq.[1] The named Plaintiffs are three black citizens who, as residents and registered voters of Starke, Florida, represent Plaintiffs' class.[2] Additionally, the Bradford County Branch of the National Association for the Advancement of Colored People (NAACP) is also a Plaintiff in the case. The Defendants are the City of Starke, and the four City Commissioners and the Mayor of Starke, and their successors in office, all of whom are sued solely in their official capacities.

The case was tried before the Court without a jury on June 14 through June 23, 1988. Pursuant to the Court's Order of September 12, 1986, the issues of liability and remedy have been bifurcated. Consequently, the trial addressed, and the Court in this decision addresses, the liability issue only. At the conclusion of the trial it was agreed by counsel and ordered by the Court that the respective parties file proposed findings of fact and conclusions of law. The parties have made their submissions and the case is now ready for decision. Accordingly, the Court makes the following Findings of Fact and Conclusions of Law, pursuant to Rule 52(a), F.R.Civ.P.

 
II. OVERVIEW OF LEGAL STANDARDS

Section 2 of the Voting Rights Act prohibits states and their political subdivisions from imposing any voting qualifications or prerequisites to voting, or any standards, practices, or procedures which result in the denial or abridgment of the right to vote of any citizen who is a member of a protected class of racial and language minorities. Thornburg v. Gingles, 478 U.S. 30, 106 S. Ct. 2752, 2762, 92 L. Ed. 2d 25 (1986). As the Supreme Court stated in Gingles, its seminal opinion construing Section 2, "the essence of a § 2 claim is that a certain electoral law, practice or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives." Id. 106 S. Ct. at 2764-65.

The Gingles Court began its analysis of Section 2 by recognizing that a violation of the statute can be proven simply by showing discriminatory effect alone rather than by proving discriminatory intent. Id. at 2759. The Court went on to caution, however, that although "multimember districts and at-large voting schemes may `operate to minimize or cancel out the voting strength of racial [minorities in] the voting population' [citations omitted], ... [such schemes] are not per se violative of minority voters' rights." Id. at 2765. Thus, "minority voters who contend that the multimember form of districting violates § 2, must prove that the use of a multimember electoral structure operates to minimize or cancel out their ability to elect their preferred candidates." Id.

After establishing those general principles, the Court in Gingles reviewed a number of factors, taken from the Senate Judiciary Committee Report accompanying the 1982 amendment of the Voting Rights Act, which are to be considered in evaluating a Section 2 claim. These factors include:

 
*1527 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 a 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.

Id. at 2759-60 quoting S.Rep. 97-417, pp. 28-29, U.S.Code Cong. & Admin.News 1982, pp. 177, 206-207. While all these factors are to be considered, the Gingles Court made it clear that they are not to be given equal weight. Gingles, 106 S. Ct. at 2766 n. 15; Carrollton Branch of NAACP v. Stallings, 829 F.2d 1547, 1555 (11th Cir. 1987), cert. denied sub nom. Duncan v. Carrollton, ___ U.S. ___, 108 S. Ct. 1111, 99 L. Ed. 2d 272 (1988). The most important of the Senate factors to be considered are minority electoral success and racial bloc voting. Id. All other factors are merely "supportive of, but not essential to, a minority voter's claim." Gingles, 106 S. Ct. at 2766 n. 15 (emphasis in original).[3]

Using the Senate factors as a guide, the Court then set out three conditions which must be proven by plaintiffs advancing a Section 2 vote dilution claim. Carrollton, 829 F.2d at 1550. The Gingles conditions *1528 center around the concept of racial bloc voting and, pursuant to them, plaintiffs must demonstrate: first, that the minority group is sufficiently large and geographically compact to constitute a majority in a single-member district; second, that the minority is politically cohesive; and third, that the white majority votes sufficiently as a bloc to enable it to usually defeat the minority's preferred candidate. Gingles, 106 S. Ct. at 2766-67.

Recently, in Solomon v. Liberty County, 865 F.2d 1566 (11th Cir. 1988), the Eleventh Circuit Court of Appeals had the opportunity to examine the analytic framework announced in Gingles. In interpreting Gingles, the Solomon court concluded that a plaintiff prosecuting a § 2 vote dilution claim must prove:

 
(1) that the minority group is under-represented in proportion to its percentage of the total electorate, (2) that the minority group has sufficient geographic and political cohesion to allow the creation of one or more minority controlled single-member districts, (3) that the totality of circumstances, with special emphasis on vote polarization and the extent of past minority electoral success, compels the inference that the current electoral system is driven by racial bias in the community or its political system, and (4) that this same evidence also leads to the conclusion that the challenged electoral system will continue to deny minorities equal access to the political process.

Id., at 1573.[4] With these principles in mind, the Court will now proceed to make its findings of fact and then apply to those findings the applicable law.

 
III. FINDINGS OF FACT 
A. General Political and Demographic Background

1. The City of Starke is in Bradford County and is located in north central Florida. Starke is the largest city in Bradford County and is the county seat.

2. The governmental administration of Starke is organized in a mayor-commissioner form of government consisting of five city commissioners, one of whom acts as mayor, an elected city clerk and an elected chief of police.

3. The term of office for each elected city official is two years. City commission candidates run in one of five groups. At large elections in the odd-numbered groups are conducted in even-numbered years and at large elections in the even-numbered groups are held in odd-numbered years. In all city elections, a candidate must receive a majority of the votes cast to be elected. If no city commission candidate receives a majority of the votes cast within his or her group at the first election, a run-off election is held between the two candidates receiving the greatest number of votes in the initial election.

4. No black person has ever been elected to serve on the Starke City Commission. Similarly, no black has ever been elected to serve in any other elected city office which includes the positions of City Clerk and Chief of Police.

5. Additionally, prior to the implementation of a single member district election system for the Board of County Commissioners of Bradford County and the Bradford County School Board in 1986,[5] no *1529 black had ever been elected to serve in any elective office in Bradford County.

6. According to the 1980 United States Census, the total population of Starke is 5,306. Of that figure, 3,625 residents (68%) of Starke are white and 1,629 residents (31%) are black. Blacks comprise 26% of the total voting age residents of Starke and, as of April 1988, 24.1% (524 out of 2,170) of the registered voters in Starke.

7. Black residents of Starke are concentrated in the northeastern section of the city which numerous witnesses referred to as the "Reno" area. United States Census data from 1980 divides the City of Starke into eight separate enumeration districts. Enumeration District (ED) 56, which is located in the northeastern section of the city, contains a total of 1,164 residents (approximately one-fifth of the total population of Starke) of whom 1,090 (94%) are black. Of the remaining 539 black residents in Starke, 379 live in ED Nos. 57 and 60 both of which are adjacent to ED 56.

 
B. Racially Polarized Voting  
(a) Overview of Statistical Techniques and Expert Testimony

8. Each party presented expert testimony and statistical evidence addressing whether racially polarized voting exists in the City of Starke. Because of the manner in which votes are tabulated in city elections and the manner in which county voting precincts are configured, both experts had to account for several factors in conducting their analyses. First, since all city voters in city commission elections vote at a single location, it is not possible to perform a statistical analysis examining the voting behavior of each voting precinct to determine the correlation between the votes received by a specific candidate and the percentage of registered voters by race in the precinct. Second, in examining the voting behavior of city voters in non-city commission elections, an analysis must address the fact that Bradford County voting precincts are not completely coterminous with the city's boundaries. Consequently, some Bradford County voting precincts which include Starke voters also include non-city, Bradford County voters.

9. Plaintiffs' expert, Dr. Allan Lichtman, Professor of History at the American University in Washington, D.C., examined all fifteen elections contested in Starke since 1968 in which a black candidate ran for city, county, state, or federal office. With respect to ten of those elections Dr. Lichtman performed a statistical ecological regression[6] and an extreme case analysis using voter registration and election data for each voting precinct containing Starke registered voters.[7]

10. Ecological regression analysis was explained as a way to determine how voters in various precincts voted.[8] It is used to measure and estimate the percentage of black citizens casting their vote for a black candidate and the percentage of white citizens voting for a white candidate in each voting precinct. The converse of these findings is also calculated; that is, the percentage of blacks voting for the white candidate and the percentage of whites voting for the black candidate (the crossover vote) is also determined. This analytic technique, therefore, yields data concerning the voting patterns of the two races, including estimates of the percentages of members *1530 of each race who voted for black candidates.[9]

11. The supplementary extreme case analysis performed by Dr. Lichtman relied on a selected series of voting precincts to predict the behavior of all voters. This type of analysis examines the actual behavior of voters in certain precincts without performing any kind of inferential methodology. Dr. Lichtman explained that he utilized voting Precinct No. 7 for his extreme case analysis and that this analysis is useful to verify in a generalized way the ecological regression estimates. However, Dr. Lichtman pointed out that the value of the extreme case analysis was somewhat diminished for elections after 1978 because Precinct No. 7 was no longer comprised of nearly 100% black residents after that date.[10] Even after 1978, though, an extreme case analysis of Precinct No. 7 can serve as a useful tool to corroborate the accuracy of the regression estimates and to determine if the regression estimates of black voting behavior are consistent with what one would expect to find in the majority black precinct.

12. Based on his analysis, Dr. Lichtman concluded that elections in Starke are racially polarized and that black voters in Starke are politically cohesive in their support of black candidates.

13. Dr. Harold Stanley, Professor of Political Science at the University of Rochester, the expert witness for the Defendants, testified that the figures derived from his own regression analysis of voting returns from Starke were essentially the same as those determined by Dr. Lichtman for the same elections, those between black and white candidates. Dr. Stanley further testified, however, that focusing only on races involving black candidates does not present an accurate picture of voting in Starke. Dr. Stanley, therefore, analyzed 93 elections between 1968 and 1988. These elections included not only those in which black candidates ran for office but also some of those in which white candidates ran for office against other white candidates ("white vs. white" elections).

14. Dr. Stanley determined that in 71% of these elections (66 out of 93), the candidate receiving the most[11] votes cast by blacks placed first in the total vote count in Bradford County. The elections analyzed by Dr. Stanley included a variety of national, state and local primary, run-off and general elections. On the basis of his study, Dr. Stanley testified that the overall pattern of voting in Starke indicates that blacks and whites usually vote for the same candidate and that the candidates preferred by blacks often place first.

15. In assessing this conflicting expert testimony, the Court's task is simplified by the fact that the experts' estimates of black and white voting behavior are virtually identical in the elections involving black candidates.[12]

*1531 16. The Court will first review elections involving black candidates for the Starke City Commission; second, the other various elections in which registered voters in Starke could participate and in which black candidates were running for office; third, the elections analyzed by Dr. Stanley; and fourth, miscellaneous related issues.

 
(b) City Commission Elections Involving Black Candidates

17. Three blacks have run for election to office on the Starke City Commission: Ms. Maurice White in 1969, Rev. Leroy Davis in 1970, and Mr. George Holiday, Jr. in 1979. A fourth black candidate, Mr. Ross Chandler, qualified as a candidate in 1986 but withdrew his candidacy over a month prior to the election.

18. In 1969, Ms. Maurice White, a black school teacher and life-long resident of Starke, finished third in the election for City Commission Group No. 4 against two white candidates. Ms. White received 455 of the 1,669 votes cast (27%). According to Ms. White, she became a candidate for the City Commission because she felt qualified, because she believed that blacks were not fully participating in city government, and because she thought that other blacks were afraid to run. Ms. White testified that she campaigned door-to-door throughout Starke. However, while black citizens worked on her campaign and encouraged her candidacy, she testified that she received resistance from whites both to her candidacy and to her door-to-door campaigning. The local newspaper reported that high voter turnout in the election was attributed by some observers to the interest generated by the candidacy of the first black to seek city office.

19. In 1970, Rev. Leroy Davis was the second black citizen to run for City Commission. Rev. Davis ran for the City Commission Group No. 5 position against a white candidate, W.M. Noegel. Rev. Davis received 405 or 31% of the 1,297 votes cast. Testimony from Rev. Davis, Ms. White and others indicates that black citizens overwhelmingly supported Rev. Davis' candidacy. The local newspaper reported that poll watchers stated that "it appeared the city's 387 Negro voters turned out in numbers." Rev. Davis testified that his qualifying fee was paid by black citizens and that, although he campaigned in both white and black communities, only blacks assisted him in his campaign efforts.

20. In 1979, Mr. George Holiday, Jr., a building contractor, became the third black individual to run for the Starke City Commission. In the first primary for the Group 4 seat, Mr. Holiday placed second in a three candidate field garnering 233 votes or 36% of the total vote. Because the leading candidate, incumbent Vernon Silcox, had received 648 votes or only 48% of the total vote, a run-off election was necessary *1532 between Mr. Holiday and Mr. Silcox. In the run-off, Mr. Silcox defeated Mr. Holiday by receiving 799 votes (56%) to Mr. Holiday's 617 (44%).

21. Mr. Holiday and others testified to the extensive involvement of black citizens in his campaign. Even so, Mr. Holiday estimated that he received only approximately 75% of the black vote due to the fact that Mr. Silcox secured considerable support from black voters because of his long time residence and activity in the area as an insurance agent and because a large segment of the black population felt that they were indebted to him for favors he had done for them in the past. Mr. Holiday additionally testified that two white candidates for city office, Mr. Joe Warren, the incumbent Mayor of Starke, and Mr. Harold Epps, a candidate for the office of Chief of Police, openly supported his candidacy. Like Mr. Holiday, both Mr. Warren and Mr. Epps were defeated in their elections.

22. Although the Court is unable to examine precinct election returns for City Commission elections since all citizens vote at a single location, the undisputed available evidence is that black voters in Starke supported each of the three black candidates for City Commission by large margins. The undisputed evidence is that the campaign workers and supporters for each of the candidates were almost all black citizens. Additionally, the witnesses who addressed the matter testified that black citizens in Starke tend to vote for black candidates. While there is evidence that Mr. Holiday did garner some white support for his candidacy, the level of this support was insufficient to enable him to defeat the white candidate, Mr. Silcox.

23. A fourth black candidate, Mr. Ross Chandler, qualified as a candidate for the Starke City Commission in July 1986. Approximately three weeks after qualifying, Mr. Chandler withdrew his candidacy. Mr. Jimmy Scott, President of the Bradford County Branch of the NAACP, and Ms. Maurice White testified that they counseled Mr. Chandler to withdraw from the election because of their concern that the white majority in Starke might support his candidacy simply in an attempt to thwart the NAACP lawsuit (i.e., this case) challenging Starke's at-large election system. These fears were corroborated by the testimony of then-Commissioner Jimmy Crosby who testified that in his opinion some whites would vote for Mr. Chandler simply to undermine this lawsuit. Although he did not agree that an electoral victory on his part would endanger the lawsuit, Mr. Chandler withdrew from the race the day after speaking with Mr. Scott and Ms. White. Mr. Chandler testified that he decided to withdraw because he had been unable to raise sufficient funds and did not have enough time to effectively wage his campaign.

 
(c) Other Elections Contested in Starke Involving Black Candidates

24. Prior to the single member redistricting of the School Board in 1986, two blacks had offered themselves as candidates for county level positions in Bradford County, Mr. Herman Johnson and Rev. Leroy Davis, both of whom ran for at-large positions on the Bradford County School Board.[13] In 1974, Mr. Johnson ran for election to the School Board from District No. 1. In Starke voting precincts in the first primary, Mr. Johnson received 24.1% of the vote and came in third behind two white candidates. In Starke precincts, 94% of blacks voted for Mr. Johnson.[14] In contrast, only 7% of white voters supported him. Four years later in 1978, Rev. Davis ran for a School Board seat from District No. 3. In that election, Rev. Davis received 26.4% of the total vote from Starke precincts. In those precincts he received *1533 99% of the black vote compared to 9% support from whites.[15]

25. City of Starke voters have also participated in several regional elections in which black candidates have run for office. In 1986, two black individuals were candidates in the September Democratic primary elections. Barbara Greadington, a candidate for United States House of Representatives, District 2, the sole black candidate in her race, finished third among a field of five candidates in the Starke voting precincts, receiving 30% of the black vote and 5% of the white vote from those precincts. In the State Senate District 5 election, Michael Moore, the sole black candidate, finished last in Starke voting precincts among the four candidates. In the Starke precincts, 42% of blacks voted for Mr. Moore whereas only 1% of whites did so. Indeed, among black voters, Mr. Moore placed first receiving a plurality of the vote.

26. Finally, black candidates have also run for several offices in statewide campaigns in which Starke citizens could vote. In 1970, Alcee Hastings ran for United States Senate in the Democratic primary. Among black voters in Starke Hastings received 42% of the vote. In contrast, he received only 1% of the vote from white voters. In 1972, Congresswoman Shirley Chisolm offered herself as a candidate in the Democratic Presidential primary. Among Starke's black voters Mrs. Chisolm received 24% of the vote; Mrs. Chisolm received no support from white voters, however. In 1974, Alcee Hastings was again a statewide candidate. In the September 1974 Democratic primary for the Public Service Commission he received 68% of the black vote and 30% of the white vote from Starke residents. In 1976, Judge Joseph Hatchett was the incumbent candidate for the Florida Supreme Court. In the September 1976 election, Judge Hatchett, the black candidate, received 84% of the black vote in Starke and 49% of the white vote. The final black candidate to run statewide in Florida was the Rev. Jesse Jackson who ran in the Democratic Presidential primaries in both 1984 and 1988. In both elections Rev. Jackson received overwhelming support from Starke's black voters and virtually no support from Starke's white voters. Specifically, in 1984 Rev. Jackson received 80% of the black vote and 0% of the white vote. Four years later in 1988, he received 98% of the black vote and 1% of the white vote from Starke residents.

 
(d) Defendants' Racial Polarization Evidence

27. As previously indicated, Defendants' expert, Dr. Stanley, performed an ecological regression analysis on 93 elections which took place between 1968 and 1988. The elections he analyzed included primary, runoff and general contests for positions at the federal, state and county levels as well as some referenda elections.[16] Based on his analysis of the 93 elections, Dr. Stanley concluded that racially polarized voting does not exist in Starke because blacks and whites generally support the same winning candidate (66 of the 93 elections or 71%).

28. In conducting his analysis, Dr. Stanley neither limited the data he considered to elections involving black candidates nor did he place any additional weight on those elections involving black candidates or white candidates who had exhibited any special or unique affinity for the interests of black citizens.

29. The 93 elections analyzed by Dr. Stanley did not represent all elections in which Starke voters could participate between *1534 1968 and 1988. Dr. Stanley testified that he analyzed only those elections which were contested and not "lopsided." Dr. Stanley defined an electoral contest as "lopsided" when the victorious candidate won the election by a three to one (75%) margin or more.[17]

30. A review of election data for all elections in which Starke voters could participate between May 1968 and March 1988 indicates that Dr. Stanley did not adhere to his stated criteria, however. The Court finds that several significant errors are apparent. First, Dr. Stanley erred in the exclusion of certain relevant elections. Between May 1968 and March 1988 there were 76 elections which Dr. Stanley failed to analyze notwithstanding that they were not lopsided based upon his three to one criterion. Thus, nearly one-half (45%) of the elections appropriate for inclusion in the analysis were excluded. Second, Dr. Stanley erred in the inclusion of certain elections in his analysis. Improperly included in his analysis were 23 elections which should have been omitted since they were lopsided. Thus, nearly one-quarter (24.7%) of the elections analyzed by Dr. Stanley should have been omitted from consideration.

31. Although the Court does not question the veracity of Dr. Stanley's testimony that he did not select the elections he analyzed with the intent to bias the results he obtained, the Court must conclude that the sample of elections that Dr. Stanley did analyze were, in fact, skewed in favor of reaching the conclusions that he drew. In a heavily Democratic community such as Starke,[18] it would be expected that in general elections both white and black voters would register as Democrats and would generally vote for the same Democratic candidate. It is therefore relevant that Dr. Stanley included a disproportionate number of general elections, compared to primary elections, in his analysis. The following numerical summary underscores this point. Thirty-three of the 89 candidate elections analyzed by Dr. Stanley were general elections. In these 33 elections, blacks and whites voted for the same Democratic candidate 29 (88%) times. However, in only 33 of the 56 primaries (59%) did blacks and whites vote for the same candidate. Thus, there exists an almost 30% difference between the extent to which whites and blacks vote the same way depending on whether general or primary elections are analyzed.

32. Given the numerical summary as stated, it is highly relevant that of the 76 elections that Dr. Stanley inappropriately excluded from his analysis, 66 were primaries. These are precisely the elections in which one would expect to find blacks and whites more often disagreeing on their candidate of choice. The corollary of this, of course, is that those elections disproportionately included for analysis, the general elections, are those more likely to show blacks and whites voting in the same manner. Therefore, even if the Court were to consider elections in which only white candidates ran for office to determine whether racially polarized voting exists in Starke, a subject discussed infra, the Court could not credit Dr. Stanley's findings, based as they are on a skewed sample of data.

33. Defendants also offered the results of a public opinion survey they had commissioned as evidence that racial polarization does not exist in Starke, that blacks in Starke no longer suffer the effect of historical discrimination, and that blacks in Starke perceive the City Commission as responsive to their needs. Defendants' survey was conducted and an interpretive report was prepared by Ms. Katherine Dowd. According to Ms. Dowd, a Ph.D. candidate at the University of Georgia, the purpose of the survey was "to measure the extent that the candidate's race, and that criterion alone, had an effect on voter decision." Ms. Dowd acknowledged that the work she did on this case was her first job *1535 as a private consultant and further acknowledged that in all of her prior surveying and public opinion polling work she had never acted as a project director or had ultimate responsibility for the survey project. Additionally, Ms. Dowd stated she had not authored any articles, books or book chapters on survey research.

34. Dr. Kenneth Wald, Professor of Political Science at the University of Florida, offered rebuttal testimony directed to Ms. Dowd's survey and report. In addition to teaching graduate level courses on survey research, Dr. Wald has authored books, chapters in books, and published articles using survey research and involving studies of political behavior. Dr. Wald testified that in his judgment the survey instrument and report prepared by Ms. Dowd did not meet high professional standards and that it was his recommendation that Ms. Dowd's report be disregarded by the Court.

35. Dr. Wald offered a number of specific criticisms of the survey questionnaire and report. First, he addressed the phenomenon of social desirability. Dr. Wald stated that when surveying social attitudes regarding race relations, the responses provided by the survey participant may not accurately reveal that individual's behavior but instead the participant may simply conform his answers to reflect the social norms of the country. To illustrate the social desirability problem, Dr. Wald noted that in response to a question asking whether the respondent had voted in the primary election held the previous day, the responses indicated that over 80% of the respondents said they did. In reality, however, the actual election results show that there was only a 60% turnout. Second, Dr. Wald cited confusion within the survey questionnaire. To explain this criticism, Dr. Wald reported that while the introductory instructions to the telephone survey advised the respondent that the survey would focus on the city commission, the first two questions involved county elections and county government. Third, Dr. Wald criticized Ms. Dowd's reliance on the survey questions to "predict behavior." According to Dr. Wald, while the responses to these questions may demonstrate support for the stated principle, it would be "quite naive" to use the specific responses to try to predict behavior. Finally, Dr. Wald criticized the vagueness of many of the survey questions. To illustrate this problem, Dr. Wald cited the question which referred to "past city elections." Dr. Wald stated that because each individual answering that question would consider a different time frame when answering, the question would not be standardized.

35. Because the Court credits the testimony of Dr. Wald, it finds that the report and opinions offered by Ms. Dowd are entitled to virtually no weight. Additionally, as described by Dr. Wald, the broad sweeping conclusions reached by Ms. Dowd as to Starke voters are not supported by the survey data.

 
(e) The Issue of City-County Precincts

36. The voting precincts analyzed by both Dr. Lichtman for the Plaintiffs and Dr. Stanley for the Defendants for the purpose of computing their ecological regression figures included some voting precincts which contained non-City of Starke, Bradford County voters.

37. For a number of reasons, however, the experts' reliance on these voting precincts to measure voting patterns of black and white Starke voters poses no problem for the Court in making its factual findings on racially polarized voting. First, there is no dispute that the voting precinct data used by the experts is the best available data. There was no alternative to each expert relying on the Bradford County voting precinct data despite the fact that certain precincts included non-city voters. Second, the number of non-city registered voters within these precincts is not so great as to create a significant concern. Third, Dr. Lichtman performed an ecological regression analysis on non-city, Bradford County voting precincts in each of the elections involving black candidates in order to determine whether there exists an discernable difference between the voting behavior of City of Starke residents and those voters living outside the city. Based on his analysis, Dr. Lichtman concluded that the *1536 pattern of voting by race in the wholly Bradford County precincts virtually mirrored the voting patterns in the precincts containing Starke voters. Finally, there is no evidence in the record to support a finding that voters in Starke, black or white, vote differently than voters in Bradford County.

 
(f) Conclusions Concerning Racially Polarized Voting

38. The Court finds that in the City of Starke in elections involving a black candidate there is a consistent relationship between the race of the voter and the race of the candidate for which he or she votes. In other words, the Court finds that where there is a black candidate running for office, black and white voters in Starke vote differently with the black voters largely supporting the black candidate while the white electorate consistently votes against the black candidate and for one of the white candidates. Notwithstanding the often overwhelming support black candidates receive from black voters, the percentage of votes received by black candidates from white voters is insufficient, even when coupled with the black vote, to allow the black's candidacy to be successful.

39. In the local elections for City Commission and School Board which occurred in the years 1969, 1970, 1974, 1978 and 1979, the expert and lay testimony reveals a consistent pattern of overwhelming support by black voters for the black candidates coupled with very little support from white voters. The black School Board candidates received 94% and 99% of the Starke black vote in 1974 and 1978 respectively, while they received less than 10% of the vote from Starke's white voters. Similarly, while there is no dispute that each of the black candidates for City Commission received some support from white citizens in Starke, the number of whites who voted for the black candidates when added to the overwhelming black support still proved insufficient for the black candidate to prevail.

40. The Court's finding that in local elections involving black candidates, black and white voters in Starke vote differently is corroborated by an examination of voting patterns in non-local elections. In all the elections involving black candidates in which Starke citizens could vote spanning the two decades between 1968 and 1988, no black candidate has ever received a majority, in head-to-head contests, or a plurality, in multi-candidate contests, of the white vote. Indeed, in only two of the ten elections analyzed by regression analysis did a black candidate even receive over 10% of the white vote. This election data reveals an unmistakably consistent pattern in which the white voters of Starke have provided minimal support for any black candidate. In contrast, black voters in Starke almost always strongly support the black candidate. In five of the ten elections, 80% or more of the blacks supported the black candidate. In the remaining five elections, blacks supported the black candidate at a rate much higher than did white voters, as is evident from an examination of each of the following elections:

a. In the 1986 State Senate primary for District No. 5, Michael Moore, the black candidate, finished first among black voters with 42% of the vote. In contrast, Mr. Moore received only 1% of the white vote.

b. In the 1986 Democratic primary for U.S. Congressional District No. 2, Barbara Greadington, the black candidate, received 30% of the black vote finishing second among five candidates. Ms. Greadington received only 5% of the white vote.

c. In the 1972 Presidential primary, Shirley Chisolm, the black candidate finished second only to former Vice-President Hubert Humphrey among black voters with 24% of the vote. In contrast, Ms. Chisolm received 0% of the white voted and finished last among white voters who supported George Wallace with 71% of the vote.

d. In the final two elections in which a black candidate did not receive 80% or more of the black vote, Alcee Hastings was the black candidate. In the 1974 election for Public Service Commission, he received 68% of the black vote and 32% of the white vote. In his 1970 run for the Democratic nomination for U.S. Senate, he received *1537 42% of the black vote but only 1% of the white vote.

41. In addition to the ecological regression estimates revealing the percentage of support for black candidates by black and white voters, the correlation co-efficient (R2) similarly demonstrates a consistently strong correlation between the percentage of vote for the black candidate, precinct by precinct, and the percentage of blacks in particular voting precincts. Dr. Lichtman found, for example, that in the School Board elections, there were strong correlations in both the 1974 Johnson candidacy (.99) and in the 1978 Davis candidacy (.97).[19] Dr. Lichtman found similarly high R2 values ranging from .71 to .98 in the remaining elections in which there was a black candidate.[20]

 
C. Additional Factors to be Considered  
(a) History of Racial Discrimination

42. The State of Florida has a long and well documented history of discrimination against black individuals.[21] As described by Plaintiffs' expert, Dr. Jerrell Shofner, a number of means were used in the past to prevent blacks from exercising their right to vote including poll taxes, all-white primaries and physical intimidation.

43. Although the City of Starke may be somewhat atypical of Florida communities because it did not employ certain means, such as a poll tax, to discourage black electoral participation, the evidence is clear that black residents of Starke have suffered from pervasive racial discrimination. Perhaps the clearest example of city-sponsored discrimination can be found in the City Charter of 1927. The Charter explicitly empowered the City Council to establish and set aside separate and distinct districts within the city where blacks and whites could reside. Although not enforced in recent years for obvious reasons, this Charter provision empowering the City to segregate the races remains on the books to this day. The modern day City Commission has never acted to delete it.

44. Consistent with the policies of racial segregation embodied in the City Charter, as Starke developed blacks were concentrated and segregated in the northeastern "Reno" section of the city. The lingering effects of the historical racial discrimination are evident by the fact that blacks today continue to be segregated by custom in the northeastern quadrant of the city.

 
(b) Use of Electoral Devices to Enhance the Opportunity for Discrimination

45. The City of Starke's at-large voting system contains several characteristics which enhance the dilutive effects of at-large voting thereby further reducing the opportunity for black citizens to get elected to positions in city government. Starke City Commission candidates run in head-to-head contests for a numbered post. A numbered post system requires a candidate to declare for a particular seat on a governmental body. The candidate then runs only against other candidates who have declared for that position. The voters then have one vote for that seat. The system prevents the use of bullet, or single shot, voting.

46. Additionally, there is a majority vote requirement. A majority vote system requires that in order to win an election, a candidate must receive more than 50% of the vote. If there are more than two candidates, *1538 and no candidate receives more than 50% of the vote, then the top two vote-getters engage in a run-off election.

47. The geographical boundaries of Starke are not so unusually large as to further enhance the dilutive effect of the at-large system or the ability of blacks to participate in the political process. However, the city does not have sub-district residency requirements for City Commissioners. Consequently, the racial segregation of the city, coupled with the complete absence of elected black officials, has resulted in no City Commissioners being elected who lived in the city's black community.

 
(c) Candidate Slating Process

48. There was no evidence of a candidate slating process for the Starke City Commission.

 
(d) Lingering Effects of Past Discrimination

49. Various socioeconomic factors which include income and educational levels indicate that blacks are substantially less well-off in Starke than whites. Black adults in Starke are significantly under-educated as compared to white adults. As of 1980, nearly 60% of black adults had an eighth grade or less education as compared to only 26% of white adults. Moreover, whites completed high school at nearly twice the rate of blacks. Blacks also earn significantly less money than whites in the City of Starke. In 1979, the average median income for white families was $15,840 while the median income for black families was $6,837. Further, as of 1979, 45% of black families in Starke had an income of less than the poverty level while the rate was only 17% for white families.

50. The socioeconomic data clearly demonstrates that a substantial difference in the status of whites and blacks exists in Starke. Despite this disparity, however, the Court does not believe that the effects of past discrimination, as opposed to the actual present discrimination evidenced by the voting patterns discussed supra, inhibits blacks from participating in the political process. This conclusion is supported by the fact that the percentage of black voter registration and the rate of black voter turn-out is similar to that of whites. Further, the disparity in income levels should not inhibit black citizens from participating in City or Starke politics as the testimony reflected that financial expenditures by candidates running for city office were quite small. Finally, two of the Plaintiffs themselves, Mr. Jimmy Scott and Ms. Elizabeth Walker (by deposition) testified that they did not believe that the education, employment or health of blacks hinder their ability to participate in the City's political process.

 
(e) Racial Appeals in Political Campaigns

51. There was no evidence of any overt or subtle racial appeals in any of the political campaigns in the City of Starke.

 
(f) Responsiveness to the Needs of Black Citizens

52. The City presented evidence of significant municipal service improvements such as street paving and water and sewer improvements in the city's black residential areas. While the bulk of these improvements were accomplished through the use of federal funds and not through the expenditure of city revenue, the testimony indicated that the overall level of improvements in city services provided to the mostly black neighborhoods is on par with the level being provided to the mostly white areas of town.

53. While the city's responsiveness to the needs of its black citizens is acceptable with regard to city services, that is not so with regard to upper-level employment opportunities. Blacks who are employed by the city are concentrated in lower paying, more menial positions. Significantly, the undisputed testimony revealed that no black individual has ever served as a department head for any of Starke's numerous departments.

 
IV. CONCLUSIONS OF LAW

54. As was discussed in Section II, supra, the Supreme Court's analysis of the Voting Rights Act in Gingles, as interpreted by the Eleventh Circuit in Solomon, establishes the legal framework which the *1539 Court must use to assess Plaintiffs' claim that the City of Starke's at-large election system violates Section 2.

 
A. Underrepresentation

55. The first element Plaintiffs must prove to establish their § 2 claim is that blacks are underrepresented in proportion to their percentage of the total electorate. Solomon, at 1573. The undisputed evidence established that blacks comprise 31% of Starke's population and that a black candidate has never been elected to any City office. ¶¶ 4 & 6 supra. This evidence conclusively demonstrates that blacks in the City of Starke are underrepresented as a matter of law. See Solomon, at 1573.

 
B. Geographic and Political Cohesion

56. The second element Plaintiffs must prove is that blacks in Starke could constitute a politically cohesive majority in a single-member district. Thus, Plaintiffs must prove both that blacks in Starke are politically cohesive and that they are sufficiently geographically cohesive to form a single-district majority. Solomon, at 1574.

57. With respect to geographic cohesion, it is undisputed that Starke's black population is sufficiently large and geographically compact so as to be able to form a majority in a single-member district. See ¶ 7 supra.

58. With respect to political cohesion, the Court concludes, based upon the statistical evidence compiled and presented by Dr. Lichtman and the testimony adduced at trial, that Plaintiffs have demonstrated that black voters in Starke are politically cohesive. See Campos v. City of Baytown, Texas, 840 F.2d 1240, 1244 (5th Cir.1988) (minority group is politically cohesive if it votes together). As is more fully described above at ¶¶ 24-26 and 39-41, both the R2 correlation coefficient and the ecological regression analyses performed by Dr. Lichtman show that blacks are politically cohesive in the City of Starke. The R2 figures in the ten elections studied range from .71 to .99 with an average of .92, on a scale of 0.0 to 1.0. The high values obtained for R2 indicates that there is a very strong relationship between the percentage of the vote received by the black candidate and the percentage of black registered voters. Cf. Solomon, at 1574-78 (finding average "r" values of .787, .966, and .925 to be "very high" and "probative of black political cohesiveness"). Similarly, the ecological regression analysis revealed that in five of the ten elections studied, 80% or more of blacks voted for the black candidate.[22] Finally, the testimony, particularly that of the black candidates for City Commission, indicated that the black citizens of Starke largely encourage and support black candidates running for office in Starke. Based upon the foregoing, the Court concludes that Plaintiffs have proven the second element of their § 2 claim by establishing that blacks in the City of Starke are politically and geographically cohesive.

59. Two issues concerning the statistical evidence require further discussion.[23] First, because there have been only three elections involving black candidates in the City of Starke and because it was not possible to perform a statistical analysis of those elections, see ¶ 8, the Court has relied, in part, on statistical evidence from non-city elections in reaching its conclusion of black political cohesion. The Court concludes that the election data from non-city elections is probative of whether black voters in Starke are politically cohesive and that the use of this data is appropriate in the circumstances of this case. See Solomon, at 1577; Carrollton, 829 F.2d at 1556.

*1540 60. Second, the only statistical evidence the Court has considered is that which was derived from elections in which a black candidate ran for office. Defendants contend that reliance solely on elections in which a black candidate ran for office is inconsistent with Justice Brennan's plurality opinion in Gingles which states that it is the race of the voter rather than the race of the candidate which is the important concern. Gingles, 106 S. Ct. at 2776. The Court disagrees with Defendants' reading of Gingles and agrees with the Fifth Circuit's interpretation:

 
[W]e conclude that Gingles is properly interpreted to hold that the race of the candidate is in general of less significance than the race of the voterbut only within the context of an election that offers voters the choice of supporting a viable minority candidate. For although the Supreme Court plurality in Gingles emphasizes the race of the voter over the race of the candidate, it upholds the trial court finding of vote dilution based upon analyses of only those elections in which blacks ran. Justice Brennan's plurality opinion is careful not to state that a black candidate is tantamount to the black preference; but implicit in the Gingles holding is the notion that black preference is determined from elections which offer the choice of a black candidate. The various Gingles concurring and dissenting opinions do not consider evidence of elections in which only whites were candidates. Hence, neither do we.

Citizens for a Better Gretna v. City of Gretna, Louisiana, 834 F.2d 496, 503-04 (5th Cir.1987) (emphasis in original).[24] Regardless of the propriety of considering data from "white vs. white" elections in general, the Court would not credit the "white vs. white" evidence offered by Defendants in this case in any event because, as is more fully discussed above at ¶¶ 29-32, that evidence was based on a non-random, skewed sample of data.

 
C. Racial Bias

61. The final two elements Plaintiffs must prove are that the discriminatory effect of Starke's at-large electoral system is caused by racial bias in the community or its political system and that such bias will continue to prevent the Plaintiffs from having equal access to the political process. Solomon, at 1573. To determine if Plaintiffs have met their burden, the Court may consider all the factors set out in Gingles. Under Gingles, however, the most probative indicator of vote dilution is evidence of white bloc voting. If the evidence "clearly establishes that whites, voting as a bloc, will probably defeat blacks voting as a bloc plus white crossovers, then the court may infer, solely from this evidence, both that the current system is driven by racial bias and that this bias is likely to dominate the challenged election." Solomon, at 1572.

*1541 62. Upon review of all the evidence, the Court concludes that legally significant racially polarized voting exists in the City of Starke. The Court draws its conclusion from the statistical evidence presented by Dr. Lichtman which demonstrates that Starke's black voters generally give overwhelming support to black candidates and that Starke's white voters rarely give black candidates more than minimal support. See supra ¶¶ 17-21, 24-26 and footnote 12. On average, in the ten elections studied in which black candidates participated, the white crossover vote (the whites who voted for the black candidate) was 10.3% of the total white vote.[25] In terms of the total vote, white crossover vote averaged 7.6%.[26] Thus, even if the votes of all blacks (26% of the voting age population) were combined with the average white crossover vote of 7.6%, the black candidate will always be defeated by the opposing white bloc vote which represents 66.4% of the total vote. The accuracy of Dr. Lichtman's analysis is borne out by the actual election experience in Starke as a black candidate has never been able to win election to any City office despite receiving overwhelming support from black voters.

63. Although the Court believes its conclusion with respect to racially polarized voting conclusively shows that Plaintiffs have carried their burden with respect to the racial bias element, the Court notes that corroborating evidence of racial bias can be found in examining the other Gingles factors. As is more fully detailed in ¶¶ 42-53 above, no black has ever been elected to office in Starke, past discrimination against blacks by the City has occurred and electoral devices in addition to the at-large voting, such as the majority vote requirement and the absence of single shot voting, reduce the opportunity for black citizens to get elected to positions in city government.

64. This same racial bias evidence, drawn from the experience in the City of Starke over the period of many years, leads the Court to the conclusion that Starke's at-large voting system will continue to deny blacks equal access to the City's political process. Plaintiffs, therefore, have proven the fourth and final element of their § 2 claim.

 
V. CONCLUSION

65. Plaintiffs have proven their § 2 vote dilution claim by establishing that black citizens of the City of Starke are under-represented in proportion to their percentage of the total electorate, that they are sufficiently geographically and politically cohesive to allow the creation of a single-member district in which they would constitute a majority, that the at-large election system currently in place in Starke is driven by racial bias and that the at-large system will continue to deny them equal access to the City's political process. Based on these findings the Court concludes that the at-large election system used to elect the Starke City Commission operates in a manner which violates Section 2 of the Voting Rights Act.

66. The parties are directed to file submissions within thirty (30) days as to how the Court should proceed to determine the appropriate remedy.

IT IS SO ORDERED.

DONE and ORDERED.

NOTES

[1] The amended Voting Rights Act provides:

(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 abridgment of the right of any citizen of the United States to vote on account of race or color ...

(b) A violation of subsection (a) is established if, based on the totality of the 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) in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, That nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.

42 U.S.C. § 1973.

[2] The Court by Order dated May 6, 1986, certified Plaintiffs' class pursuant to Rule 23(b) (2), F.R.Civ.P., to include all black residents of Starke, Florida.

[3] The Gingles Court more fully explained by stating:

Under a "functional" view of the political process mandated by § 2, S.Rep. 30, n. 120, U.S.Code Cong. & Admin.News 1982, p. 208, the most important Senate Report factors bearing on § 2 challenges to multimember districts are the "extent to which minority group members have been elected to public office in the jurisdiction" and the "extent to which voting in the elections of the state or political subdivision is racially polarized." Id., 28-29, U.S.Code Cong. & Admin.News 1982 P. 206. If present, the other factors, such as the lingering effect of part discrimination, the use of appeals to racial bias in election campaigns, and the use of electoral devices which enhance the dilutive effects of multimember districts when substantial white bloc voting existsfor example antibullet voting laws and majority vote requirements, are supportive of, but not essential to, a minority voter's claim.

In recognizing that some Senate Report factors are more important to multimember district vote dilution claims than others, the Court effectuates the intent of Congress. It is obvious that unless minority group members experience substantial difficulty electing representatives of their choice, they cannot prove that a challenged electoral mechanism impairs their ability "to elect." § 2(b). And, where the contested electoral structure is a multimember district, commentators and courts agree that in the absence of significant white bloc voting it cannot be said that the ability of minority voters to elect their chosen representatives is inferior to that of white voters. [citations omitted] Consequently, if difficulty in electing and white bloc voting are not proven, minority voters have not established that the multimember structure interferes with their ability to elect their preferred candidates.

Gingles, 106 S. Ct. at 2766 n. 15 (emphasis in original).

[4] The Solomon court made clear that it was merely restating the procedures a trial court should follow in the wake of the Gingles decision and not formulating a new test to be applied to § 2 claims. Id., at 1571.

[5] Simultaneously with the filing of this action, the named Plaintiffs here also filed Section 2 claims directed against the at-large voting schemes used to elect members of the Bradford County Commission and the Bradford County School Board. See Bradford County Branch of the NAACP, et al v. Bradford County School Board, et al, Case Number 86-4-Civ-J-12 and Bradford County Branch of the Naacp, et al v. Bradford County Commission, et al, Case Number 86-6-Civ-J-14. In July 1986, consent Final Judgments which provided for single-member district elections were entered in both cases. The identical single-member district plan was adopted in each case. Under the new districting plan, citizens in the northeastern section of the City of Starke reside in District Number 1 which is majority black in both population and registered voters. (Final Judgments of July 3, 1986 and July 11, 1986).

[6] Dr. Lichtman described ecological regression analysis as the standard method for inferring the behavior of population groups from data collected for political units. Defendants do not dispute the appropriateness of this methodology; indeed, their expert, Dr. Harold Stanley, testified, "[i]n terms of the method employed, my understanding is that we go about these things in very similar ways, so that my ecological regression and [Dr. Lichtman's] are essentially the same. So, no, there are no disagreements in principles." (Testimony of Dr. H. Stanley, June 20, 1988, Tr. 170).

[7] Dr. Lichtman performed the regression analysis in only ten of the fifteen elections because there was not sufficient voter registration and election return data for him to perform the analysis in the other five elections.

[8] Ecological regression analysis is the methodology which was employed by Plaintiffs and endorsed by the Supreme Court in Thornburg v. Gingles, 106 S. Ct. at 2768 n. 20 (affirming Dr. Grofman's ecological regression and extreme case analysis).

[9] In an effort to determine the reliability of the estimates derived from the ecological regression analysis, Dr. Lichtman employed two supplemental analyses, the R2 correlation analysis which correlates the percentage of blacks registered in a given voting precinct with the percentage of the votes received by the black candidate in that precinct, and a statistical significance analysis.

[10] For example, in 1976 blacks comprised 390 of the 391 registered voters in Precinct No. 7. Following a modification of the Precinct 7 boundaries, blacks comprised between 70% to 80% of the registered voters in that precinct.

[11] A majority of the votes in head-to-head elections or a plurality of the votes in multi-candidate elections.

[12] Percentages of Votes Cast By Black and White Voters for Black Candidates Dr. Lichtman Dr. Stanley A. County Primary Elections WHITE BLACK WHITE BLACK 1. School Board (Johnson 1974) 7% 94% 7% 94% 2. School Board (Davis 1978) 9% 99% 10% 96% B. Regional Primary Elections 3. U.S. Congress District 2 (Greadington) 5% 30% 6% 29%

 
                                  WHITE        BLACK           WHITE       BLACK
4. State Senate
   District 5
   (Moore)                          1%          42%              2%         42%
                            C. Statewide Elections
5. U.S. Senate
   Primary 1972
   (Hastings)                       1%          42%              3%         42%
6. U.S. Presidential
   Primary-1972
   (Chisolm)                        0%          24%              0%         24%
7. Public Service
   Commission Primary
   1974 (Hastings)                 30%          68%             33%         65%
8. Supreme Court
   1976 (Hatchett)                 49%          84%             50%         84%
9. U.S. Presidential
   Primary 1984
   (Jackson)                        0%          80%              0%         76%
10. U.S. Presidential
    Primary 1988
    (Jackson)                       1%          98%              1%         98%

[13] Additionally, Mr. Jimmy Scott and Ms. Carolyn Spooner were candidates for the school board in 1986 after the board had adopted the single member district election system.

[14] The figures for percentage of support by race are taken from Dr. Lichtman's ecological regression analysis. These figures, as well as Dr. Stanley's calculations, are set out in table form, supra, in footnote 12.

[15] No regression analysis was performed on the 1986 black candidacies for School Board. In the case of Mr. Scott's election, this was so because he ran unopposed for the newly created single member District No. 1, a majority black district which includes the northeastern quadrant of the City of Starke. In the case of Ms. Spooner's election, no analysis was performed because she ran for office from a district in which there was an insufficient number of black registered voters in any of the Starke voting precincts on which to adequately perform a regression analysis. The Court does note, however, that in Starke precincts, Ms. Spooner finished third behind two white candidates with 24% of the vote.

[16] Dr. Stanley's universe of data included 89 elections involving candidates and 4 involving referenda.

[17] Dr. Stanley testified that the three to one ratio he employed to determine if an election was lopsided was not a precise yardstick but rather was an "impressionistic" guide.

[18] Voter registration data from 1986 reflect that of the 2,325 registered voters in Starke, 2,093 were registered Democrats.

[19] The correlation coefficient expresses a correlation that can range from 0.0 (no relationship) to 1.0 (perfectly consistent relationship).

[20] The other R2 values computed by Dr. Lichtman are as follows:

 
United States Senate Primary 1970
  Alcee Hastings                         .96
Presidential Primary 1972
  Shirley Chisolm                        .95
Public Service Comm. 1974
  Alcee Hastings                         .71
Supreme Court 1976
  Joseph Hatchett                        .89
Presidential Primary 1984
  Jesse Jackson                          .98
U.S. Congressional Primary 1986
  Barbara Greadington                    .89
State Senate Primary 1986
  Michael Moore                          .91
Presidential Primary 1988
  Jesse Jackson                          .97

[21] See e.g., McMillan v. Escambia County, Florida, 638 F.2d 1239, 1244 (5th Cir.1981).

[22] In the other five elections, black electoral support for black candidates ranged from 68% to 24%. In the three elections in which blacks failed to give a majority or plurality of their support to the black candidate, the black candidate still received a great deal more support from the black voters (ranging from 42% to 24%) than from white voters (5% to 0%). These three elections all had multi-candidate fields with five or more candidates running.

[23] The following discussion concerning the statistical evidence applies with equal force to consideration of both political cohesiveness and racial polarization issues.

[24] While there is no binding precedent directly addressing this issue, the Court's reading of Solomon supports its view that Defendants' evidence of "white vs. white" elections should be excluded from consideration. The Solomon Court, in discussing situations in which there was neither clear bloc voting nor clear non-bloc voting stated,

Between these extremes, however, will fall situations in which evidence of vote polarization and minority electoral failure is either inconclusive, insufficient, or unavailable due to a lack of minority candidacies. In these cases, the courts must rely on the full range of totality of circumstances inquiries to determine whether the challenged electoral system violates the Act.

Solomon, at 1573 (emphasis added). Thus, Solomon focused directly on elections in which minorities offered themselves as candidates. The possible reliance on the "totality of circumstances" refers to the factors drawn from Gingles and the Senate Report and does not include consideration of elections in which just whites ran for office. See supra Section II. It must finally be noted that the Solomon Court did consider evidence of a few elections which did not involve a black candidate. These elections were not simply any "white vs. white" elections in the jurisdiction, however, but were "a group of elections that revolved around racial issues or themes." Id. at 1576. In this case, neither party has presented any evidence of elections involving only white candidates which they contend revolve around "racial issues or themes." In the absence of this special subset of "white vs. white" elections, the Court concludes that Solomon supports the exclusion of elections in which no black candidate ran for office.

[25] In the three City Commission elections in which a black candidate ran for office, Dr. Lichtman projected the average white crossover vote to be 16.8%.

[26] Whites comprise 74% of the City of Starke. 10.3% of 74% is 7.6%.