Wright v. Sumter Cnty. Bd. of Elections

301 F. Supp. 3d 1297

• Court: District Court, M.D. Georgia
• Decided: March 17, 2018
• Docket: CASE NO.: 1:14–CV–42 (WLS)
• Status: Published

Opinion

W. LOUIS SANDS, SR. JUDGE

*1302This case is a challenge to the method of electing members of the Board of Education in Sumter County, Georgia. (Doc. 1.). The plaintiff, Mathis Kearse Wright, Jr., contends that the current election plan's two at-large seats and high concentration of African-American voters in Districts 1 and 5 dilute African-American voting strength in violation of Section 2 of the Voting Rights Act of 1965, as amended, 52 U.S.C. § 10301. (Id. )

The Court held a four-day bench trial on December 11-14, 2017. (Docs. 144; 145; 146; 148.) In issuing these findings of fact and conclusions of law, the Court has considered the evidence presented at trial, the Parties' written closing arguments (Docs. 161; 162; 163), their proposed findings of fact (Docs. 169; 172), and their trial briefs. (Doc. 170; 171.)

FINDINGS OF FACT

Plaintiff Mathis Kearse Wright, Jr. is an African-American resident and registered voter in Sumter County.¹ (MUF at ¶ 1.) Defendant Sumter County Board of Elections and Registration was established by state law in 2001 and is responsible for conducting elections for members of the Sumter County Board of Education. 2001 Ga. Laws 3865.

I. County Demographics

Sumter County has a total population of 31,070 people.² Of those, 12,399 (39.9%) are non-Hispanic white and 16,122 (51.9%) are non-Hispanic black; 13,095 (42.1%) are white and 16,159 (52.0%) are black. (Doc. 164-1 at 1, 3). Most-23,541-of those people are voting-age. (Doc. 164-1 at 5-10.) Their demographics are similar to the general population: 10,991 (46.7%) are white and 11,652 (49.5%) are black. Id. Sumter County has 15,683 total active registered voters, 7,327 (46.7%) are white and 7,604 (48.5%) are black. (Doc. 166 at 2.)

The socioeconomic disparities between black and white residents of Sumter County are striking. Only 13.6% of white residents lack a high school diploma. The rate is over double-29.9%-for African Americans. (Doc. 164 at 4.) White residents are over three times more likely to have a bachelor's degree or higher-30.9% versus 8.8% of African Americans. (Id. ) The educational differences are reflected in employment numbers as well. Among those in the workforce aged sixteen years or over, the unemployment rate is 7.1% for white residents and 18.2% for African Americans. (Id. ) Only 15.3% of white residents live in poverty compared to an astonishing 46.2% of African Americans. (Id. at 5.) Three in four African American households receive Supplemental Nutrition Assistance Program benefits. (Id. ) The number is reversed for white residents: only one in five households receive the same benefits. (Id. ) The median African American household earns $22,736, less than half of the median $48,672 for white households. (Id. )

*1303These disparities result in decreased political participation. (See Doc. 157 at 110:18-111:25.) Despite African Americans outnumbering white residents in population, voting-age population, and registered voters, white voters have outnumbered black voters in school-general elections by an almost two-to-one margin since 1996. (Docs. 153-38-153-60.)

II. School Board

A. Historical Composition

Before passage of the Voting Rights Act, members of the Sumter County Board of Education were appointed by the Sumter County grand jury. See Edge v. Sumter Cty. Sch. Dist. , 775 F.2d 1509, 1510 (11th Cir. 1985). In 1964, the General Assembly reorganized the Board to consist of seven members elected from four single-member districts, one two-member district, and one member elected at-large. See Edge v. Sumter Cty. Sch. Dist. , 541 F.Supp. 55, 56 (M.D. Ga. 1981). The composition has changed some times since. In 1973, it moved to at-large elections for the entire Board after a federal judge concluded the prior districts were unconstitutionally apportioned. 1992 Ga. Laws 5171; see Edge , 541 F.Supp. at 56. The United States Attorney General found the at-large system would "have a racially discriminatory effect," but it continued nonetheless until 1981. (Doc. 153-62); see Edge , 541 F.Supp. at 56. That year, a three-judge panel found the system violated the Voting Rights Act and enjoined its further use. Edge , 541 F.Supp. at 56.

The Board struggled to make a permissible change. It first proposed six single-member districts and one at-large seat, but the Attorney General found that the evidence "suggests that the submitted plan was designed with the purpose of minimizing minority voting strength in the school district." (Doc. 153-63 at 2.) The Board proposed another "six-one" plan, but again the Attorney General objected. (Doc. 153-64.) The district court eventually proposed its plan for all single-member districts, but the United States Court of Appeals for the Eleventh Circuit vacated because the district court had failed to consider whether it violated Section 2 of the Voting Rights Act. Edge , 775 F.2d at 1510. Eventually, in 1986, all parties involved settled on a six-one plan with three majority-black districts. (Doc. 153-65.)

The composition was short-lived: following the 1990 census, the Georgia General Assembly adopted a new election plan consisting of seven single-member districts. (Doc. 153-81.) Four years later, the Assembly upped the count to nine single-member districts. (Doc. 153-83.) The Board stuck with nine single-member districts after the 2000 census, though the district borders changed. (Doc. 153-84.) Under the updated plan, four districts were majority African American in voting-age population. (Doc. 153-23 at 13.)

B. Recent Changes to the Board

The Board had five white members and four black members in 2010. (Doc. 153-61.) That year, it began discussing redistricting and downsizing. In June, it approved a plan-with a 5-3 vote along racial lines-to reduce the size of the Board to five or seven members with details to be worked out later. (Docs. 153-67 at 3; 159 at 39:10-18.) The nine-member School Board at the time was the largest in the state, despite Sumter County's relatively small population. (Doc. 159 at 14:3-18.) The Southern Association of Colleges and Schools (SACS), the organization which accredits Sumter County's schools, said on several occasions that the School Board was large. (Id. at 15:1-16.) Defendants assert that the Board's size put Sumter County's accreditation at risk. (See Doc. 172 at ¶ 45.) This assertion is not credible. Michael Busman, *1304the School Board's chairman and whose testimony Sumter County relies on, testified that SACS accredits the schools and that a loss of accreditation would be detrimental to the students. (Doc. 159 at 15:1-11.) He did not testify that SACS's observation regarding the Board's size would have any impact on the accreditation process.

It is somewhat unclear the justification for landing on the number seven, however. The five single-member districts mirror the five single-member Board of Commissioners districts-a reason cited by the School Board when it submitted the new plan for preclearance. (Doc. 153-23 at 2.) But the Board of Commissioners doesn't have any at-large districts. (Doc. 159 at 16:8-15.) Busman testified only that it was "easier" to go from nine to seven members rather than down to five. (Id. at 16:10-15.) The Court infers, based on the testimony, that the smaller shift was "easier" because fewer incumbent seats were put at risk.

On November 2, 2010, an African-American candidate, Kelvin Pless, defeated a white incumbent, Donna Minich, in District 3. (Doc. 153-61 at 3.) For the first time, the Board had an African American majority.

On December 9, 2010, before Pless was installed, the Board unanimously approved a resolution calling for the legislature to move it to a five-two plan-five district seats and two at-large seats. (Docs. 153-67 at 3; 154-11; 159 at 17:14-16.) African American board members testified that they did not knowingly vote to support the addition of two at-large seats. (Doc. 158 at 154:5-10, 174:2-10.) The Court does not find them credible on this point. The resolution was covered extensively in both newspapers and on the radio. (Doc 159 at 40:6-10.) Moreover, the resolution was reviewed by the members before the vote. (Id. at 17:3-13.) The Court does not believe that responsible Board members would vote in support of a resolution to change the Board composition with no knowledge of what the change entailed.

The General Assembly adopted the change, and the Governor subsequently signed it into law. 2001 Ga. Laws 4020.

On July 31, 2011, the incumbent in District 7, who was not African American, resigned his seat. (Doc. 153-24 at 44.) Michael Lewis, an African-American, was appointed to fill the seat, bringing the racial makeup of the Board to six African American members and three white members. (Doc. 158 at 143:21-144:8.)

Shortly after that, the General Assembly redrew the new district boundaries based on the 2010 census. 2011 Ga Laws. 280. The changes were submitted to the Department of Justice for preclearance under Section 5 of the Voting Rights Act. (Docs. 153-23; 153-24.) The Department found the information submitted insufficient for its analysis and requested additional information from the Board. (Doc. 153-23.) The Board refused. On January 12, 2012, the Board voted to move from the five-two plan to seven single-member districts, and on January 18, 2012, it voted to withdraw its request for preclearance. (Docs. 153-23 at 43; 159 at 22:20-23:7.)

The decision to withdraw the request for preclearance left the nine single-member districts configuration in place, but following the 2010 census, they were malapportioned. In June 2012, this Court granted the plaintiff's motion for a preliminary injunction and thereby canceled the 2012 elections for members of the Board of Education. See Order, Bird v. Sumter County Board of Education , 1:12-cv-76-WLS (M.D. Ga. June 21, 2012). Board members whose terms were set to expire after the 2012 elections were held over-maintaining the six-three black majority.

*1305On June 25, 2013, the majority's strategy backfired. The Supreme Court struck down Section 4 of the Voting Rights Act in Shelby County v. Holder , 570 U.S. 529, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013). Preclearance was no longer required, and the General Assembly's post-census plan went into immediate effect. See Order, Bird v. Sumter County Board of Education , 1:12-cv-76-WLS, 2013 WL 5797653 (M.D. Ga. Oct. 28, 2013). The General Assembly-perhaps out of an abundance of caution-then readopted the same five-two plan through House Bill 836 on February 17, 2014. The bill also moved school board elections from the November general election to the nonpartisan general election held in May. It also adopted a transition procedure: a special election would be held for Districts 1, 2, 4 and 6 as those districts existed under the nine-member plan. Members elected at that special election were to serve only until December 31, 2014, when the new plan would be put in place. (Doc. 153-22); 2014 Ga. Laws 3503.

That brings us to today. The Board still has its five-two composition. Elections take place in May of even-numbered years, with candidates running on a staggered four-three basis. One at-large seat is filled each election, and a majority vote is required for all Board members. African Americans constitute a majority of the voting-age population in two of the five existing school-board districts. (MUF ¶ 8.) They represent 62.7% of the voting-age population in District 1 and 70.6% of the voting-age population in District 5. (Doc. 153-87 at 4.)

C. Elections Under the Current Plan

Wright relies on a racial bloc voting analysis of Sumter County elections performed by his expert, Dr. Frederick G. McBride, and contained in McBride's supplemental report. (Doc. 153-87.) Dr. McBride has a doctorate in Political Science from Clark Atlanta University. His work has focused on quantitative and qualitative research in redistricting and voting rights. He has drawn and evaluated redistricting plans, performed racially polarized voting studies, performed demographic analysis, and presented at redistricting hearings for over 100 jurisdictions in twenty-two states, and the District of Columbia. (Docs. 153-1; 153-87 at 2; 157 at 23:6-28:10.)

McBride's analysis in this case used three statistical methods to estimate the voting patterns of black and white voters in Sumter County: (1) homogeneous precinct analysis; (2) bivariate ecological regression analysis; and (3) Ecological Inference (EI). (Doc. 153-87 at 8-9.) All three methods have been accepted by the courts as reliable for use in voting cases, and their reliability is not at issue here. See, e.g., Thornburg v. Gingles , 478 U.S. 30, 53 n.20, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986) (discussing ecological regression and homogeneous precinct analysis). There is also no dispute that the EI method is currently the "gold standard" for use in racial bloc voting analyses, so the Court will include only the results of Dr. McBride's EI analyses unless otherwise noted. (Docs. 157 at 40:3-10; 159 at 207:24-208:3.)

The data for Dr. McBride's analysis consisted of precinct election returns and racial turnout data from the Georgia Secretary of State. (Doc. 153-87 at 10.) There is no dispute as to the data used in Dr. McBride's analysis. (Doc. 159 at 209:2-210:4.) Dr. McBride analyzed a total of twelve Sumter County elections in his supplemental report. These included all three general elections and one runoff election held under House Bill 836 for the at-large seats on the Board of Education, plus a variety of other local elections. (Doc. 153-87 at 11.)

Dr. McBride's analysis shows that African-American voters have been highly cohesive in ten of the twelve elections analyzed *1306by Dr. McBride. The two exceptions are first, the District 4 race on May 20, 2014, when 53.9% of black voters supported Rick Barnes and 46.6% of black voters supported Gary Houston. Both Barnes and Houston are white. The second race was the March 18, 2014, District 6 election. There, 68.8% of black voters supported Sarah Pride while 52.1% of black voters supported Michael Mock. Pride is an African American; Mock is white. Of the remaining races, the lowest level of support for the black-preferred candidate among black voters was an astonishingly high 85.3%.

Of particular note is the November 2, 2004 race for Sheriff. There, Nelson Brown-the only African American candidate in the race and the black-preferred candidate-won the support of 96.5% of black voter despite being a write-in candidate.

The results of McBride's analysis are shown below. First, the Court shows the general demographic and turnout data provided by McBride:

Table 1. Demographics of the Sumter County School Board's Districts.

Seat Population % White % Black % 18+ White % 18+ Black 1 6,432 31.1% 65.9% 34.4% 62.7% 2 6,654 56.4% 34.6% 62.2% 30.3% 3 6,546 54.1% 38.8% 57.8% 36.2% 4 6,679 44.8% 47.7% 49.1% 43.9% 5 6,508 24.1% 72.8% 27.0% 70.6% At-Large3 31,070 41.2% 51.0% 46.7% 49.5%

[Editor's Note: The preceding image contains the reference for footnote³ ].

Table 2. Voter Turnout Data for the Elections Analyzed.

Election Date Seat White Turnout Black Turnout May 20, 2014 1 9.4% 8.6% May 20, 2014 2 20.7% 4.4% May 20, 2014 3 12.9% 5.4% May 20, 2014 4 6.9% 4.1% May 20, 2014 5 6.7% 9.3% May 20, 2014 At-Large #1 11.1% 6.7% May 20, 2014 At-Large #2 11.2% 6.7% July 22, 2014 At-Large #1 8.2% 4.7% May 24, 2016 At-Large #2 14.3% 8.4% November 2, 2004 Sheriff 29.1% 18.4% November 2, 2010 3 12.9% 10.8%

Next, the Court lists the elections analyzed *1307by McBride, categorized by election date. African American candidates are denoted with an asterisk.

1. May 20, 2014 Elections

Table 3. District 1 Results.

Candidate Overall Support White Support Black Support Alice Green* 52.8% 15.0% 94.2% E. Lockhart 11.6% 21.2% 1.2% Allen Smith 35.4% 66.8% 1.1%

Table 4. District 2 Results.

Candidate Overall Support White Support Black Support Everette Byrd 28.9% 30.1% 23.3% Meda Krenson 48.7% 59.0% 0.0.% Sarah Pride 22.2% 5.8% 99.3%

Table 5. District 3 Results.

Candidate Overall Support White Support Black Support W. Fitzpatrick* 30.5% 4.6% 92.3% J.C. Reid 69.4% 95.0% 8.5%

Table 6. District 4 Results.

Candidate Overall Support White Support Black Support Rick Barnes 54.4% 54.7% 53.9% Gary Houston 45.5% 44.9% 46.6%

Table 7. District 5 Results.

Candidate Overall Support White Support Black Support Edith Green* 55.4% 13.9% 85.3% Mark Griggs 44.5% 86.4% 14.3%

Table 8. At-Large Seat # 1 Results.

Candidate Overall Support White Support Black Support Michael Coley* 36.7% 4.1% 89.1% David Kitchens 20.4% 32.7% 0.0% Sylvia Roland 36.4% 53.0% 9.7% Patricia Taft* 6.3% 6.5% 6.2%

Roland is a career public school educator having served as an English teacher in middle and high school in Arkansas for twelve years and a literacy coach in middle school in Florida for six years. (Doc. 159 at 43:15-22.) She moved to Sumter County in *13082012 and became a school improvement specialist in Americus High School in Sumter County. (Id. ) Roland never had a child in Sumter County schools and had never voted in a city election. (Doc. 159 at 54:10-17.)

Coley has lived in Sumter County for almost his entire life. (Doc. 158 at 33:23-37:11.) He served on the Sumter County Board of Education from 1996 until 2005, but never worked in Sumter County schools. (Id. at 42:1-8.) Coley's three children all graduated from Americus High School. (Id. at 38:3-10.)

Table 9. At-Large Seat # 2 Results.

Candidate Overall Support White Support Black Support Michael Busman 59.8% 94.4% 3.0% Kelvin Pless* 40.1% 5.8% 96.7%

Busman has lived in Americus, Georgia for over 19 years and is a family medicine and sports medicine physician. (Doc. 159 at 8:19-20, 9:2-4.) He is the volunteer team physician for the high school, and he performs free physicals for the school athletes and Special Olympics athletes. (Id. at 9:15-22.) Busman has four children-one graduated from Americus Sumter High School, and the other three are now homeschooled. (Doc. 159 at 28:11-20.)

Pless has lived in Americus, Georgia his whole life. (Doc. 158 at 60:23-24.) He was elected to the School Board as the representative for District 3 in 2010. (Id. at 65:25-66:4.) Pless has a degree in education, though he has never worked for the Sumter County schools. (Id. at 66:22-23.)

2. July 22, 2014 Elections

Table 10. At-Large Seat # 1 Runoff Results.

Candidate Overall Support White Support Black Support Michael Coley* 41.0% 7.5% 99.5% Sylvia Roland 58.9% 92.4.% 0.0%

3. May 24, 2016 Elections

Table 11. At-Large Seat # 1 Results.

Candidate Overall Support White Support Black Support Michael Coley 44.5% 15.3% 93.6% Sylvia Roland 55.4% 84.7% 6.2%

4. Other Elections

Table 12. November 2, 2004 Sheriff Election Results.

*1309Candidate Overall Support White Support Black Support Pete Smith 40.3% 54.6% 0.0% James Driver 32.3% 39.9% 6.3% Nelson Brown* (Write In) 27.3% 4.4% 96.5%

Table 13. November 2, 2010 School Board District 3 Results

Candidate Overall Support White Support Black Support Donna Minich 44.4% 76.7% 6.3% Kelvin Pless* 55.3% 22.9% 94.0%

Table 14. March 18, 2014: School Board District 6 Results

Candidate Overall Support White Support Black Support Michael Mock 71.0% 85.1% 52.1% Sarah Pride* 28.9% 28.8% 68.0%

In addition, there were other races presented during the trial that were not analyzed by Dr. McBride. The Court notes the races in which there was a contested choice between an African American and a white candidate.⁴

In the 2012 general election, Barack Obama (African American) defeated Gary Johnson (white) and Mitt Romney (white) in Sumter County for President. (Doc. 154-10 at 21; Doc. 159 at 103.) Sanford Bishop (African American) defeated John House (white) in Sumter County for a United States House of Representatives seat. (Doc. 154-10 at 21; Doc. 159 at 102-03.) Kevin T. Brown (African American) defeated Michael Arthur Cheokas (white) for a State House of Representatives seat. (Doc. 154-10 at 21; Doc. 159 at 104.) George R. Torbert (white) defeated Tangalia Robinson (African American) in Sumter County for a County Commission seat, but Andrea F. Brookes (African American) defeated Carey Harbuck for a second seat. (Doc. 154-10 at 22; Doc. 159 at 105.)

In the 2014 general election, Sanford Bishop (African American) defeated Greg Duke (white) in Sumter County for a United States House of Representatives seat. (Doc. 154-10 at 25; Doc. 159 at 102.) Kevin T. Brown (African American) again defeated Michael Arthur Cheokas (white) for a State House of Representatives seat. (Doc. 154-10 at 25; Doc. 159 at 104.)

That makes six wins for African American candidates over white candidates, and one win for a white candidate over an African American candidate in the 2012 and 2014 general elections. Of the races where the entire county voted in a single *1310race, African American candidates were five for five in defeating white candidates.

In the 2016 primary election, Cortisa Barthell (African American) defeated C. Cromer (white) and M. Harry (white) in Sumter County to be the Democratic nominee for the Clerk of Superior Clerk. (Docs. 154-10 at 27; 159 at 89.) In that election, African Americans made up 81.9% of the electorate. (Doc. 157 at 218:12-14.) Barthell did not face an opponent in the general election. There were no examples of races in the 2016 general election, based on the evidence presented, where the Court can find an African American candidate went against a white candidate. (Doc. 159 at 91-97.)

III. Discrimination in Sumter County

Georgia's history of discrimination "has been rehashed so many times that the Court can all but take judicial notice thereof. Generally, Georgia has a history chocked full of racial discrimination at all levels. This discrimination was ratified into state constitutions, enacted into state statutes, and promulgated in state policy. Racism and race discrimination were apparent and conspicuous realities, the norm rather than the exception." Brooks v. State Bd. of Elections , 848 F.Supp. 1548, 1560 (S.D. Ga. 1994), appeal dismissed and remanded sub nom. Brooks v. Georgia State Bd. of Elections , 59 F.3d 1114 (11th Cir. 1995). The Parties' have stipulated to this sordid history in both Georgia generally and Sumter County more specifically. (See Doc. 155 ("Georgia and Sumter County have a long and extensive history of voting discrimination against African Americans.") ) Given the stipulation, the Court declines to make the litany of factual findings about Georgia requested by Wright. (See Doc. 169 at ¶¶ 174-383.) However, the Court does make the following findings specific to Sumter County to provide better context for this challenge.

In 1967 in Bell v. Southwell , the court set aside an election in Sumter County because of "gross, unsophisticated, significant, and obvious racial discrimination," including segregated voting lists and polling booths, intimidation of black voters by whites, and the arrest of black voters attempting to vote in white polling booths. 376 F.2d 659, 660-61, 664 (5th Cir. 1967).

In 1981 in Edge v. Sumter County School District , this court noted that "[o]n July 13, 1973 the Attorney General interposed an objection to the change [to at-large elections for the board of education]. In spite of this objection the at-large system has been utilized for Board elections up to the present time." 541 F.Supp. 55, 56 (M.D. Ga. 1981), aff'd, Sumter County School District v. Edge , 456 U.S. 1002, 102 S.Ct. 2287, 73 L.Ed.2d 1297 (1982). In a later ruling, the Eleventh Circuit Court of Appeals noted that "[n]o black person has ever served on the county school board" and that "[i]n 1964, prior to the Voting Rights Act, Georgia law provided that the Sumter County grand jury appoint school board members." 775 F.2d 1509 (11th Cir. 1985).

While its worst days may be behind it, Sumter County remains a largely segregated community, with separate neighborhoods, civic organizations, and churches. (Docs. 158 at 39:7-41:6, 64:1-65:20, 107:4-108:15, 116:3-13, 125:15-133:22, 167:21-168:7, 211:5-212:4; 159 at 54:24-55:22.) Explicitly racist incidents are still not unheard of. Wright ran for a seat on the county commission in 2006 and described several such incidents from his campaign: "on this one occasion this-this white family sicced their German shepherd on the-on one of my daughters during one of the times. And then there were other times when, you know, they just basically said, you know, sorry, but, you know, we don't vote for-and they said the N word. And then there was a couple of incidents where *1311they said don't come on my property." (Doc. 158 at 215:19-216:1.)

IV. Illustrative Remedial Plan

Wright has proposed an illustrative remedial plan which he asserts would remedy the alleged Section 2 violation. It is as follows:

Table 15. Plaintiff's Illustrative Remedial Proposal.

Dist. Total Pop. Voting Age White % White Black % Black Pop. (VAP) VAP VAP VAP VAP 1 4,663 3,290 1,083 32.92 2,120 64.44 2 4,686 3,636 2,446 67.27 957 26.32 3 4,772 3,605 1,975 54.79 1,490 41.33 4 4,675 3,575 1,924 53.82 1,476 41.29 5 4,703 3,279 717 21.87 2,472 75.39 6 4,677 3,797 1,999 52.65 1,457 38.37 7 4,693 3,336 1,293 38.76 1,818 54.50

(Doc. 153-87 at 6.) Sumter County challenges the districts for reasons related to Section 2, but does not allege that the proposals are unfaithful "to Georgia's traditional redistricting principles of compactness, contiguity, minimizing the splits of counties, municipalities, and precincts, recognizing communities of interest, and avoiding multi-member districts." Larios v. Cox , 314 F.Supp.2d 1357, 1369 (N.D. Ga. 2004) (footnote omitted); (see generally Doc. 176).

To estimate how the proposed districts would vote, McBride applied the conceptual framework set forth by Bernard Grofman, Lisa Handley, and David Lublin in Drawing Effective Minority Districts: A Conceptual Framework and Some Empirical Evidence , 79 N.C. L. Rev. 1383 (2001). (Doc. 153-87 at 22.) The framework uses cohesion, crossover voting, and turnout to determine how a proposed district would vote. (Doc. 157 at 136:12-18.) The authors are well respected in the field of political science, (Id. at 137:3-6; Doc. 159 at 222:7-20), and their methods have been cited in Georgia v. Ashcroft , 539 U.S. 461, 483, 123 S.Ct. 2498, 156 L.Ed.2d 428 (2003) and League of United Latin Am. Citizens v. Perry , 548 U.S. 399, 488, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006) (Souter, J., concurring in part and dissenting in part).

McBride's analysis shows that the proposed District 1 and District 5 would allow African American voters to elect their preferred candidate. (Doc. 153-87 at 23.) Proposed District 7, however, is a close call. The percentage of the voting-age population needed for a minority-preferred candidate to be elected in Sumter County has ranged from 44.1% to 77.8% in the current districts. (Id. ) At 54.5%, proposed District 7 would be sufficient in some cases but not others. Determining whether African American voters in the proposed district could elect the candidate of their choice is "guesswork," but McBride testified that based on the Grofman framework, he believed it was sufficient. (Doc. 157 at 198:8-199:8.)

The County's expert, Dr. Karen Owen, did not express any opinions on McBride's illustrative districts or his analysis of the viability of the districts. (Doc. 159 at 221:16-222:6.)

*1312DISCUSSION

I. Gingles and Senate Factors

In this case, Wright claims that the Sumter County Board of Education's composition, five members from single-member districts and two at-large members, violates Section 2 of the Voting Rights Act of 1965. Section 2 prohibits an election plan that

[d]ivid[es] the minority group among various districts so that it is a majority in none may prevent the group from electing its candidate of choice: If the majority in each district votes as a bloc against the minority candidate, the fragmented minority group will be unable to muster sufficient votes in any district to carry its candidate to victory.

Colleton Cnty. Council v. McConnell , 201 F.Supp.2d 618, 633 (D. S.C. 2002). Section 2 also prohibits "packing" where minority voters are all placed in one single member district.⁵ Id. "[T]he critical question in a § 2 claim is whether the use of a contested electoral practice or structure results in members of a protected group having less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." Thornburg v. Gingles , 478 U.S. 30, 63, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986) (citations omitted).

In Thornburg v. Gingles , the United States Supreme Court set forth three preconditions that a plaintiff must prove for a Section 2 claim to go forward. Gingles , 478 U.S. at 50-51, 106 S.Ct. 2752 ; see also Nipper v. Smith , 39 F.3d 1494, 1524 (11th Cir. 1994) (holding that a "plaintiff cannot obtain relief unless he or she can establish" each of the three Gingles preconditions). The three Gingles preconditions are: (1) the minority group "is sufficiently large and geographically compact to constitute a majority in a single-member district"; (2) the minority group "is politically cohesive"; and (3) "the white majority votes sufficiently as a bloc to enable it ... usually to defeat the minority[-]preferred candidate." Id. (citations omitted).

Each of the three Gingles preconditions must be established before a reviewing court can proceed to consider the "Senate Factors," a non-exhaustive and non-exclusive list of factors set forth in a Senate Judiciary Committee Majority Report that accompanied an amendment to Section 2, which aid courts in assessing the totality of the circumstances surrounding challenged voting schemes. Id. at 37-38, 106 S.Ct. 2752 (citing S. Rep. No. 97-417 (1982) ). Some of the Senate Factors may have a direct bearing on the three Gingles preconditions, but none of the Senate Factors must be present in order to satisfy the Gingles threshold. However, "they must be examined when determining whether, considering all of the circumstances in the case, the plaintiffs are entitled to section 2 relief." Nipper , 39 F.3d at 1526-27. The Senate Factors include:

the history of voting-related discrimination in the State or political subdivision;

the extent to which voting in the elections of the State or political subdivision is racially polarized;

the extent to which the State or political subdivision has used voting practices or procedures that tend to enhance the opportunity for discrimination against the minority group, such as unusually large election districts, majority vote requirements, and prohibitions against bullet voting;

*1313the exclusion of members of the minority group from candidate slating processes;

the extent to which minority group members bear the effects of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process;

the use of overt or subtle racial appeals in political campaigns;

and the extent to which members of the minority group have been elected to public office in the jurisdiction.

Gingles , 478 U.S. at 44-45, 106 S.Ct. 2752 (citing S. Rep. No. 97-417 at 28-29) (formatting altered).

A. The First Gingles Factor: Numerosity and Compactness

As to factor one, Wright must show that the minority group is sufficiently large and geographically compact to constitute a majority in a single-member district. "[T]he first Gingles condition requires the possibility of creating more than the existing number of reasonably compact districts with a sufficiently large minority population to elect candidates of its choice." Johnson v. De Grandy , 512 U.S. 997, 1008, 114 S.Ct. 2647, 129 L.Ed.2d 775 (1994). The Court already granted summary judgment as to this factor; there is no need to revisit it now. (Doc. 125 at 16.) African Americans currently hold two of the seven School Board seats. McBride has demonstrated a plan which would permit African Americans to elect three members of their choice.

B. The Second Gingles Factor: Minority Political Cohesiveness

As to the second factor, Wright must show the minority group is politically cohesive. "A showing that a significant number of minority group members usually vote for the same candidates is one way of proving the political cohesiveness necessary to a vote dilution claim ... and consequently establishes minority bloc voting within the context of § 2." Gingles , 478 U.S. at 56, 106 S.Ct. 2752 (citations omitted). Gingles does not require that the minority group always vote for the same candidate but does require that the minority group usually or consistently vote for the same candidate, a standard which this Court finds demands more frequency than a more often than not standard but less frequency than an always standard. Id. at 48, 56, 106 S.Ct. 2752.

The Court finds that the second Gingles factor is also satisfied. Of the twelve elections with reliable data before the Court, in ten of them the overwhelming majority of African Americans voted for the same candidate. In one of the two where they did not, the District 4 race on May 20, 2014, both candidates were white. While still relevant, elections without a black candidate are less probative in evaluating the Gingles factors. Davis v. Chiles , 139 F.3d 1414, 1418 n.5 (11th Cir. 1998). The Court is particularly struck by the November 2, 2004, race for Sheriff. In that race, African American candidate Nelson Brown received nearly every single black vote (96.5%) despite being a write-in candidate when running against two white candidates. Write-in candidates face obvious structural barriers that make their election in the American political system rare. To see African American voters demonstrate that level of cohesion for a write-in campaign is extraordinary. When buttressed by the other nine cohesive elections, it is clear factor two is also satisfied.

Sumter County makes a few arguments in opposition. First, it points to McBride's original report which showed a lower level of cohesion in the same elections than does his supplemental report. For those elections analyzed in both McBride's original report and his supplemental report, the *1314only analytical change was to use actual black and white turnout data rather than the estimated turnout data McBride had to rely on originally. Sumter County argues that "[t]here is virtually no correlation between Dr. McBride's [turnout] estimates and the actual numbers." (Doc. 170 at 15.) True enough. The estimates do vary wildly from the actual turnout data. (See id. at 15-16 (Defendant's comparison chart).) But even Defendant's expert testified that were she reanalyzing an election where she had originally used voting age population data for turnout, she would use actual turnout data was it to become available. (Doc. 159 at 209-210.) The shift, then, only reflects poorly on McBride's original turnout estimates and not his final analysis.

Second-and a slight variation on the first argument-Sumter County argues McBride's original analysis demonstrates a low level of cohesion amongst black voters. (Doc. 170 at 12.) The County's arguments cut against themselves. The Court agrees that McBride's original turnout estimates were unreliable and vary wildly from the actual turnout numbers. The Court does not find those results credible and therefore does not consider any of the original results-good or bad-in evaluating factor two.

Third, the County argues McBride's analysis is unreliable because he eliminated three elections from his original analysis when producing his supplemental report. (Id. at 11-12.) McBride explained that he did not reanalyze those three elections, the 2002 Board of Education District 3, the 2006 Board of Education District 3, and the 2008 Board of Education District 1 races, because he did not have voter turnout data in those elections available to him, and therefore he had "nothing new" to report from his original report. (Doc. 157 at 60:25-61:9.) McBride testified that he stood by the results of his original analysis. (Id. at 61:10-12.) The Court, however, does not find any of the original analysis credible after seeing how drastically different the voter turnout numbers were from McBride's original predictions. But even if it did, and even assuming those three elections show a lack of cohesion, they would only bring the total to ten cohesive elections and five non-cohesive elections. The Court finds those results would still satisfy factor two.

The Court also does not accept that McBride's selection criteria introduced bias into his results. His selection criteria-recent elections for which he had reliable turnout data-is entirely reasonable. Sumter County was free to run its own analysis on additional elections to show how McBride's results were unreliable. It chose not to do so.

Fourth, the County argues McBride's analysis is not credible because in some cases his estimates of voter preferences total over 100%, a logical impossibility. (Doc. 170 at 14.) Dr. McBride explained that ecological inference establishes bounds between zero and 100 for the estimates of black or white support for an individual candidate, but it does not constrain the sum of those estimates to 100%. (Doc. 157 at 94:9-95:20, 162:13-164:12.) Gary King, the creator of ecological inference, identified this problem and proposed that researchers could either use an algebraic expression to bring the estimates within a 100% bound or leave them as is. (Doc. 160 at 128:12-129:23.) McBride chose to leave them as-is to avoid altering the results. (Id. at 130:13-19.)

The County's expert, Dr. Karen Owen, testified in her deposition that the EzI program-written by Gary King and used by McBride to conduct his EI analysis-could give a sum of over 100% for the estimates of white voter support or black voter support. (Id. at 57:6-10.) At trial, Owen testified that a sum of over 100%

*1315would call into question the data inputted because "Gary King wanted to ensure we were getting estimates between a bound of zero and 100 percent." (Doc. 160 at 55:14-20.) Owen has never used EzI (id. at 53:14-17), did not independently analyze the elections in Sumter County (see generally Doc. 154-9), did not use EI in her dissertation or research (Doc. 160 at 59:8-16), did not publish any results using EI until 2015 (id. at 59:17-24), and could not identify a source for her claim that a sum over 100% calls into question the accuracy of the estimates. (id. at 58:20-59:7.) The Court does not find her criticism credible. Rather, it accepts McBride's testimony that EI can give sums exceeding 100% and that such a result does not call into the question the reliability of McBride's analysis.⁶

C. The Third Gingles Factor: Majority Bloc Voting

Finally, "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." Gingles , 478 U.S. at 51, 106 S.Ct. 2752 (citation omitted). "[T]he degree of racial bloc voting that is cognizable as an element of a § 2 vote dilution claim will vary according to a variety of factual circumstances. Consequently, there is no simple doctrinal test for the existence of legally significant racial bloc voting." Id. at 57-58, 106 S.Ct. 2752.

"[P]laintiffs seeking to establish the third Gingles factor 'must show not only that whites vote as a bloc, but also that white bloc voting regularly causes the candidate preferred by black voters to lose; in addition, plaintiffs must show not only that blacks and whites sometimes prefer different candidates, but that blacks and whites consistently prefer different candidates.' " Johnson v. Hamrick , 296 F.3d 1065, 1074 (11th Cir. 2002) (quoting Johnson v. Hamrick , 196 F.3d 1216, 1221 (11th Cir. 1999) ).

The elections analyzed in this case fall into three general categories. First are those seven races where a white candidate faced an African American candidate. In six of those seven races, as detailed in factor two, there was a clear candidate preferred by African Americans. An average of 88.3% of white votes cast in those races went to the white candidate. African American candidates only won two of those races: the 2014 District 5 race and the 2010 District 3 race. In District 5, over 70% of the voting-age population is black. In the previous District 3, the population was approximately half white and half black. African Americans constituted 48.4% of the voting-age population. (Doc. 154-6.) In sum, in six of the seven races, African Americans and whites preferred different candidates. The Court excludes the seventh race-the March 2014 District *13166 election-because, without a black-preferred candidate, it cannot meaningfully consider whether white voters are usually able "to defeat the minority's preferred candidate." Gingles , 478 U.S. at 51, 106 S.Ct. 2752. In four of the six races in this category, the black-preferred candidate lost. African Americans had only one true success: the District 3 race. The District 5 win was in a predominantly African American district. See Old Pers. v. Cooney , 230 F.3d 1113, 1122 (9th Cir. 2000) (considering minority group success in minority-majority districts under the totality of the circumstances, but not under Gingles factor three because "[t]o do otherwise would permit white bloc voting in a majority-white district to be washed clean by electoral success in neighboring majority-[minority] districts").

The second type of race is where there are multiple candidates facing a black-preferred candidate. The Court counts four such races: the 2014 elections in District 1, District 2, and At-Large Seat # 1, and the 2004 sheriff race. The Court discounts the race for sheriff. There, 96.5% of African Americans wrote in Nelson Brown, demonstrating an incredible level of political cohesion. While only 4.4% of whites voted for Brown, a write-in candidacy is a special circumstance which does not shed light on whether there is "racial bias in the voting community." Nipper v. Smith , 39 F.3d 1494, 1524 (11th Cir. 1994). The Court can only speculate as to whether white voters were aware that Brown was running as a write-in candidate and, if they did, whether they would have voted for him. In the District 1 case, between 85% and 88% of white residents voted against the black-preferred candidate. The African American candidate was still able to win the race, however, as 62.7% of the voting-age population in that district is black. In the District 2 race, between 89.1% and 94.2% of white residents voted against the black-preferred candidate. The African American candidate was defeated and the two white candidates advanced to a run-off election. In the At-Large Seat # 1 race, between 92.2% and 95.9% of white residents voted against the black-preferred candidate. Although the black-preferred candidate won the plurality and was able to advance to a run-off, white voters then coalesced around a single candidate and defeated him. Discounting the majority-black district, African Americans had no real successes in these types of elections.

Finally is the one election where two white candidates faced each other-the 2014 District 4 election. The race had no clear black-preferred candidate, nor a clear white-preferred candidate. The Court discounts that race because of the lack of a clear preference and the lack of an African American running. See Johnson , 196 F.3d at 1221.

Sumter County argues the results from the at-large elections should be discounted because Busman and Roland had worked in the Sumter County schools, but Coley and Pless had not. (Doc. 170 at 20.) The election of the white candidates, it implies, is thus a preference for better-qualified candidates and does not reflect "entrenched voting patterns." (Id. at 21.) The Court disagrees. The Ninth Circuit has rejected any attempt "to scrutinize the qualifications of minority candidates who run for public office in jurisdictions with historically white-only officeholders." Ruiz v. City of Santa Maria , 160 F.3d 543, 558 (9th Cir. 1998). There does not appear to be any binding Eleventh Circuit law holding the same. However, the Court finds that even if it can examine candidate quality, doing so would not discount the importance of the at-large elections. While Coley had never worked in the school district, he had been a school board member for almost a decade. Conversely, his opponent Roland had impeccable education credentials *1317but knew very little about the community and had never had a student in the public school system. Voters could easily decide either was more qualified. Likewise, Busman volunteered as a team physician and is an upstanding member of the community, but homeschooled three of his children rather than send them to the school district he oversees. His opponent, Pless, had experience on the School Board and a background in education. Voters could reasonably select either.

Sumter County points out that African Americans have had success in November general elections. (Doc. 170 at 18.)⁷ Neither side has presented a statistical analysis of these races. There is thus no evidence of whether there was a black-preferred candidate in those races. Sumter County flippantly asserts Wright "cannot seriously contend that Barack Obama and Sanford Bishop are not the preferred candidates of the Sumter County black community ...." (Doc. 170 at 18 (emphasis in original).) Yet in the March 2014 District 6 election, an African American faced a white candidate and there was no black-preferred candidate. The Court will not merely assume black voters in Sumter County support every black candidate. Moreover, these elections took place at a different time of year than the current school board elections, included voters from outside of Sumter County, and were for positions other than school board. Accordingly, they are of diminished relevance here because they do they not allow the Court to make inferences about voter patterns in the challenged districts. See Cofield v. City of LaGrange, Ga. , 969 F.Supp. 749, 760 (N.D. Ga. 1997).

Reviewing the elections analyzed by McBride, there can be no doubt black and white voters consistently prefer different candidates. Moreover, white voters are usually able to the defeat the candidate preferred by African Americans. There was only one true "success" in the elections analyzed where an African American candidate preferred by African Americans was able to defeat a white-preferred candidate when the electorate was not predominantly black. The third Gingles factor is satisfied.

Sumter County argues Wright cannot satisfy factor three because African Americans in Sumter County are not a "minority," but rather a majority of the population and a plurality of the voting-age population. (Doc. 170 at 5.) Other courts have found that although a majority group claiming a need for protection under Section 2 "faces an obvious, difficult burden in proving that their inability to elect results from white bloc voting, they are not precluded, as a matter of law, from seeking to prove such a claim."

*1318Salas v. Sw. Texas Jr. Coll. Dist. , 964 F.2d 1542, 1555 (5th Cir. 1992). Like any other group, they must show "less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." Gingles , 478 U.S. at 63, 106 S.Ct. 2752 (citation omitted).

The County hypothesizes ways in which it believes Wright could theoretically satisfy that burden-past discrimination could result in lower voter registration rates, felon disenfranchisement could disproportionately impact African Americas, the voting rolls might be inaccurate, other minority groups may band together with whites, there may be racially gerrymandered districts, or other practical impediments to African Americans voting may exist. (Doc. 170 at 3-7.) But, the County concludes, Wright has made no such showing. (Id. )

While African Americans do outnumber whites on the voter rolls, the voting booth is another story. In the school board elections since the new plan was implemented, white voters have outnumbered black voters in seven of nine races. See Tables 1; 2.⁸ The only exceptions are the elections in District 1 and District 5 where African Americans make up over 60% of the voting-age population. Sumter County cites Missouri State Conference of the Nat'l Ass'n for the Advancement of Colored People v. Ferguson-Florissant Sch. Dist. , 201 F.Supp.3d 1006, 1071 (E.D. Mo. 2016), for the proposition that low voter registration rates can form the basis for a Section 2 claim by a group that is a near-majority in population or voting-age population. (Doc. 170 at 6.) The Court finds no meaningful difference between a failure to register to vote and a failure to cast a vote. It is, of course, true that were more African Americans to register (as in Missouri State ) or turn out to vote (as here), they would likely be able to elect their preferred candidate. But our circuit has roundly rejected any effort to blame African Americans' lack of electoral success on "a failure of blacks to turn out their votes." United States v. Marengo Cty. Comm'n , 731 F.2d 1546, 1568 (11th Cir. 1984) (quoting district court decision) (punctuation corrected). As outlined in the factual findings, Sumter County and the State of Georgia have a long history of discrimination. The effects of that discrimination still linger today in the form of disproportionate educational achievement, employment, income levels and living conditions. Sumter County cites an out-of-circuit case requiring evidence linking past discrimination to low turnout today. (Doc. 170 at 8 (citing Salas v. Sw. Texas Jr. Coll. Dist. , 964 F.2d 1542, 1556 (5th Cir. 1992) ).) Our circuit has no such stringent requirement. "[W]hen there is clear evidence of present socioeconomic or political disadvantage resulting from past discrimination, as there [is] in this case, the burden is not on the plaintiffs to prove that this disadvantage is causing reduced political participation, but rather is on those who deny the causal nexus to show that the cause is something else." Marengo , 731 F.2d at 1569. Sumter County has produced no evidence or argument showing what the low African American voting rate is attributable to. (See generally Doc. 170.) The Court, therefore, must assume a causal connection to the past discrimination.

*1319Having found that all three Gingles factors are satisfied, the Court moves on to the Senate factors.

D. Senate Factor One

The first Senate factor is the history of voting-related discrimination in the State or political subdivision. "[P]ast discrimination can severely impair the present-day ability of minorities to participate on an equal footing in the political process. Past discrimination may cause blacks to register or vote in lower numbers than whites. Past discrimination may also lead to present socioeconomic disadvantages, which in turn can reduce participation and influence in political affairs." United States v. Marengo Cty. Comm'n , 731 F.2d 1546, 1567 (11th Cir. 1984).

The Parties have stipulated that "Georgia and Sumter County have a long and extensive history of voting discrimination against African Americans." This factor weighs heavily in Wright's favor.

E. Senate Factor Two

The second factor is the extent to which voting in the elections of the State or political subdivision is racially polarized. "[T]his factor will ordinarily be the keystone of a dilution case." United States v. Marengo Cty. Comm'n , 731 F.2d 1546, 1566 (11th Cir. 1984). The Court finds the Sumter County's voters to be highly polarized. In ten of the twelve elections analyzed, over 85% of African American voters voted for the same candidate. Less than a quarter of white voters supported the black-preferred candidate in any of those races. The average level of white support in those races was under 10%. In one of the two races which were not polarized, there was no African American candidate. The election results, therefore, would surely "have been different depending upon whether it had been held among only the white voters or only the black voters." Thornburg v. Gingles , 478 U.S. 30, 54, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986) (citation omitted). This factor also weighs heavily in Wright's favor.

F. Senate Factor Three

Factor three is the extent to which the State or political subdivision has used voting practices or procedures that tend to enhance the opportunity for discrimination against the minority group, such as unusually large election districts, majority vote requirements, and prohibitions against bullet voting. The current plan employs three parts relevant to this Senate factor.

First, Sumter County uses staggered terms for the at-large seats, with one at-large seat filled at each regular election. (MUF ¶ 6; PX 26 (House Bill 836).) Were the County to instead seat the top two vote-getters for at-large seats every four years, African Americans would have an enhanced opportunity for election in those seats. An illustrative example is the May 20, 2014, election for at-large seat number 1. There, an African American candidate received 36.7% of the vote and a white candidate received 36.4% of the vote. The white candidate won the subsequent run-off. Had the two candidates receiving the most votes instead been elected, the African American candidate-Michael Coley-would be a school board member.

Second, Sumter County uses a majority-vote requirement in elections for the at-large seats. (MUF ¶ 3.) The impact on African American candidates is apparent in the same race. Had Sumter County employed a plurality-win system, Michael Coley would have won the May 20, 2014, election. Because of the majority-win system, he was defeated. Majority-vote requirements have long been recognized as enhancing an opportunity for discrimination. See League of United Latin Am. Citizens, Council No. 4434 v. Clements , 986 F.2d 728, 749 (5th Cir.), on reh'g , *1320999 F.2d 831 (5th Cir. 1993) ; City of Rome v. United States , 446 U.S. 156, 183, 100 S.Ct. 1548, 64 L.Ed.2d 119 (1980), abrogated on other grounds by Shelby Cty., Ala. v. Holder , 570 U.S. 529, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013).

Third, the addition of at-large districts enhanced the opportunity for discrimination. Several witnesses with experience in local politics testified that running at large in Sumter County is more expensive than running in a district and therefore presents a particular barrier for African-American candidates. (Doc. 158 at 52:24-54:23, 77:20-78:4, 135:23-138:2.) Although Sumter County itself is not unusually large, the larger area nonetheless requires greater costs. One white candidate, Sylvia Roland, received unsolicited money to assist with those added costs. (Doc. 159 at 54:21-23.) There is no testimony of any African American candidate receiving a similarly unsolicited donation.

The third factor weighs in Wright's favor.

G. Senate Factor Four

The fourth factor is the exclusion of members of the minority group from candidate slating processes. "The term 'slating' is generally used to refer to a process in which some influential non-governmental organization selects and endorses a group or 'slate' of candidates, rendering the election little more than a stamp of approval for the candidates selected." Westwego Citizens for Better Gov't v. City of Westwego , 946 F.2d 1109, 1116 n.5 (5th Cir. 1991). There is no evidence in the record of any slating process in Sumter County. Accordingly, the Court cannot find whether a slating process would or would not exclude African Americans. This factor carries no weight.

H. Senate Factor Five

The fifth factor is the extent to which minority group members bear the effects of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process. As recognized in the original Senate Report, "disproportionate educational, employment, income level and living conditions arising from past discrimination tend to depress minority political participation. Where these conditions are shown and where the level of black participation in politics is depressed, plaintiffs need not prove any further causal nexus between their disparate socio-economic status and the depressed level of political participation." Johnson v. Mortham , 926 F.Supp. 1460, 1519 (N.D. Fla. 1996) (quoting S. Rep. No. 97-417 at 29 n.114 (1982) ).

As detailed in the factual findings, only 13.6% of white residents lack a high school diploma. The rate is over double-29.9%-for African Americans. (Doc. 164 at 4.) White residents are over three times more likely to have a bachelor's degree or higher-30.9% versus 8.8% of African Americans. (Id. ) The educational differences are reflected in employment numbers as well. Among those in the workforce aged sixteen years or over, the unemployment rate is 7.1% for white residents and 18.2% for African Americans. (Id. ) Only 15.3% of white residents live in poverty compared to an astonishing 46.2% of African Americans. (Id. at 5.) Three in four African American households receive Supplemental Nutrition Assistance Program benefits. (Id. ) The number is reversed for white residents: only one in five households receive the same benefits. (Id. ) The median African American household earns $22,736, less than half of the median $48,672 for white households. (Id. ) There can be no doubt that African Americans in Sumter *1321County face "disproportionate educational, employment, income level and living conditions arising from past discrimination ...."

There can also be no doubt that the level of black participation in Sumter County politics is depressed. In the elections analyzed in the case, African Americans were on average over 60% less likely than their white counterparts to cast a vote. See Table 2.

Having shown a disparate socio-economic status between white and black residents of Sumter County and a depressed level of political participation by African Americans, this factor weighs heavily in Wright's favor. See S. Rep. No. 97-417 at 29 n.114 (1982).

I. Senate Factor Six

The sixth factor is the use of overt or subtle racial appeals in political campaigns. Wright points to alleged incidents of African American candidates facing hostile and racist constituents while on the campaign trail. (Doc. 171 at 45.) The sixth factor, however, concerns racist messages being communicated to constituents, not constituents communicating racists messages to the candidates. See, e.g., Meek v. Metro. Dade Cty., Fla. , 805 F.Supp. 967, 982 (S.D. Fla. 1992), aff'd in part, rev'd in part , 985 F.2d 1471 (11th Cir. 1993) (voters were told that "Black candidates share common goals with Jesse Jackson or Nelson Mandela"). There is no evidence, and Wright does not allege, that any political campaign employed overt or subtle racist appeals. Accordingly, this factor weighs in favor of Sumter County. However, the Court recognizes that "overtly bigoted behavior has become more unfashionable." Marengo , 731 F.2d at 1571 (quoting Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights , 558 F.2d 1283, 1290 (7th Cir. 1977) ). While this factor can weigh heavily in favor of a plaintiff when present, "its absence should not weigh heavily against a plaintiff proceeding under the results test of section 2." Id.

J. Senate Factor Seven

The seventh factor is the extent to which members of the minority group have been elected to public office in the jurisdiction. Here, it is undisputed that no African American has ever been elected to an at-large seat on the School Board under the challenged plan. (Doc. 125 at 20.) No African American has been elected in a School Board district except in districts where African Americans make up a majority of the voting-age population. (Id. ) The only evidence of an African American being elected to county-wide office was in 2016 when Cortisa Barthell became Clerk of Superior Court. Barthell won the Democratic nomination where African Americans made up 81.9% of the electorate and did not face a general election opponent. There is no evidence in the record of an African American in Sumter County winning a contested race for county-wide office.

In sum, African Americans have lacked success in Sumter County elections. This factor weighs heavily in Wright's favor.

K. Additional Senate Factors and Considerations

The Senate Report and courts applying Section 2 have recognized several other factors that may be relevant in determining the totality of the circumstances. The Court reviews those relevant to this case.

First is 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. S. Rep. No. 97-417 at 29. Wright asserts the School Board's policy assertion for adding at-large districts was tenuous. (Doc. 171 at 48-51.) The Court *1322disagrees. At nine members, the Sumter County School Board was one of the largest in the state. At the recommendation of its accreditation agency, the School Board reduced the number of members and realigned the districts to mirror those of the County Board of Commissioners. Those decisions were entirely reasonable. But the Board of Commissioners only has five members. Wright asserts there was no reason to move to a seven-member Board with two at-large seats rather than a five-member Board with no at-large seats. The only testimony on this issue was Busman's testimony that it was "easier" to go from nine to seven members rather than down to five. (Doc. 159 at 16:10-15.) As noted in the factual findings, the Court infers that the smaller shift was "easier" because fewer incumbent seats were put at risk. There is nothing tenuous about minimizing changes to make the districts more politically palatable.

Another reasonable interpretation is that nine was the status quo, and the further one strays from the status quo, the more difficult the transition can be. Again, this justification would not be tenuous. It can be challenging to predict the problems which will arise when shifting to a new district alignment. The more the new system resembles the old, the more familiar it will be to election officials, candidates, and voters.

Wright argues the asserted rationale is belied by the timing of the changes. While it is true that the final Board vote approving the plan did not occur until a lame-duck session immediately before the newly-elected black board member would give African Americans a majority on the Board, the bill was introduced before that election, and the final resolution passed without any opposition. The timing does not undermine the asserted purpose.

Further, the General Assembly's re-enactment of a plan not precleared by the Department of Justice following the Shelby County decision is not evidence of an improper motive. (See Doc. 171 at 51.) When the Department requested additional information to decide if the plan should be given preclearance, it was the African American majority on the Board which refused to provide that information. The plan's lack of preclearance, therefore, is not evidence of discrimination toward African Americans.

The lack of a tenuous policy justification thus weighs toward Sumter County. However, a showing that the policy justification is race-neutral does not negate "a plaintiff's showing through other factors that the challenged practice denies minorities fair access to the process." S. Rep. No. 97-417 at 29 n.117.

Second is the proportionality inquiry. " 'Proportionality' as the term is used here links the number of majority-minority voting districts to minority members' share of the relevant population." Johnson v. De Grandy , 512 U.S. 997, 1014 n.11, 114 S.Ct. 2647, 129 L.Ed.2d 775 (1994). Proportional districts help to assure that "minority voters have an equal opportunity, in spite of racial polarization, 'to participate in the political process and to elect representatives of their choice.' " Id. at 1020, 114 S.Ct. 2647 (quoting 42 U.S.C. § 1973(b) (1994) ). Wright argues that "African Americans constitute [49.5%] of Sumter's County's voting-age population, but they constitute a majority of the voting-age population in only two (28.6%) of the board's seven seats."⁹ (Doc. 171 at 52 (citations *1323omitted).) The Court agrees that the districts are not proportional. While African Americans hold a majority or plurality in four of the seven districts, it is abundantly clear plurality districts do not provide an equal opportunity for African Americans to elect representatives of their choice given the history of discrimination in the county. Accordingly, because African Americans hold a majority in only two districts, this factor weighs toward Wright.

Wright asserts a third relevant factor: racial separation. (Doc. 171 at 52.) The Court finds no support for racial separation being a consideration in a Senate Factors analysis. Wright cites three cases arguing otherwise. Two, United States v. City of Euclid , 580 F.Supp.2d 584, 592-93 (N.D. Ohio 2008) and United States v. Charleston Cty. , 316 F.Supp.2d 268, 291 (D.S.C. 2003), aff'd sub nom. , 365 F.3d 341 (4th Cir. 2004), are out-of-circuit district court cases with no precedential value in this Court. The third, McMillan v. Escambia Cty., Fla. , 688 F.2d 960, 967-68 (5th Cir. 1982), vacated , 466 U.S. 48, 104 S.Ct. 1577, 80 L.Ed.2d 36 (1984), only noted the district court's observation of racial separation in its Fourteenth Amendment-not Voting Rights Act-analysis, and in any event the judgment was vacated on appeal and never reinstated. See Tallahassee Branch of NAACP v. Leon Cty., Fla. , 827 F.2d 1436, 1440 (11th Cir. 1987) (" McMillan has no binding precedential effect."). In the absence of any authority recognizing this factor, the Court declines to consider it.

II. Totality of the Circumstances

The Court must "consider the 'totality of circumstances' to determine whether members of a racial group have less opportunity than do other members of the electorate." League of United Latin Am. Citizens v. Perry , 548 U.S. 399, 425-26, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006). "[I]t will be only the very unusual case in which the plaintiffs can establish the existence of the three Gingles factors but still have failed to establish a violation of § 2 under the totality of circumstances." N.A.A.C.P., Inc. v. City of Niagara Falls, N.Y. , 65 F.3d 1002, 1019 n.21 (2d Cir. 1995) (citation omitted); see Thompson v. Glades Cty. Bd. of Cty. Comm'rs , 493 F.3d 1253, 1261 (11th Cir.), reh'g en banc granted, opinion vacated , 508 F.3d 975 (11th Cir. 2007), and on reh'g en banc , 532 F.3d 1179 (11th Cir. 2008) (noting the Niagara Falls standard, though the opinion was later vacated and the district court affirmed by an evenly divided en banc panel). "In such cases, the district court must explain with particularity why it has concluded, under the particular facts of that case, that an electoral system that routinely results in white voters voting as a bloc to defeat the candidate of choice of a politically cohesive minority group is not violative of § 2 of the Voting Rights Act." Niagara Falls , 65 F.3d at 1019 n.21 (citation omitted).

The Court finds, based on the totality of the circumstance, that African Americans in Sumter County have less opportunity to elect candidates of their choice than do white citizens. Under the totality standard, the Court finds the following facts particularly compelling: (1) the incredibly high rates of polarized voting in races that pit an African American candidate against a white candidate; (2) the glaring lack of success for African American candidates running for county-wide office, both historically and recently, despite their plurality in voting-age population; (3) the undisputed history of discrimination in Sumter County and throughout Georgia; (4) the lingering effects of that discrimination today, including the comparatively low income and education levels and high rates of poverty for African Americans in Sumter County; and (5) the low rate of African American turnout *1324in these elections which-in the absence of evidence to the contrary-the Court attributes to the history of discrimination and the socioeconomic disparities. Because of these factors, the elections for at-large seats do not give African Americans in Sumter County a meaningful opportunity to elect the candidates of their choice.

III. Illustrative Plan

"In a § 2 vote dilution suit, along with determining whether the Gingles preconditions are met and whether the totality of the circumstances supports a finding of liability, a court must find a reasonable alternative practice as a benchmark against which to measure the existing voting practice." Holder v. Hall , 512 U.S. 874, 880, 114 S.Ct. 2581, 2585, 129 L.Ed.2d 687 (1994) (footnote omitted).

The Court finds that, by the bare minimum, Wright has shown his illustrative plan "achieve[s] a more proportional representation of minorities than did the previous multi-member system." Meek v. Metro. Dade Cty. , 908 F.2d 1540, 1548 (11th Cir. 1990) (quoting Solomon v. Liberty County , 865 F.2d 1566, 1572 n.5 (11th Cir. 1988) ). African Americans in Sumter County are currently able to elect two of seven candidates of their choice. The Court accepts, based on the evidence presented, that they are unable to elect the candidates of their choice in the at-large districts where they account for 49.5% of the voting-age population. In the illustrative plan, neither party contests that African Americans would be able to elect the candidates of their choice in District 1 and District 5. The question for the Court, then, is whether they would have an opportunity to elect the candidate of their choice in illustrative District 6, a single-member district where they represent 54.5% of the voting-age population. As McBride readily concedes, the answer is "guesswork." Based on the cohesion and crossover voting patterns, that percentage would be sufficient in some of the current single-member districts, but not others. McBride asserts that a 49.5% district is not a black-majority district, so it would behave like districts with far less black voters. Meanwhile, a 54.5% district is a black-majority district, so it would behave differently from the at-large districts. (Doc. 158 at 13:10-20:3.) The Court finds no support for the idea that a five percentage point shift would have such a drastic impact on voting behaviors.

That said, the testimony before the Court is that illustrative District 6 has a greater percentage of African American voters than has been needed in other districts to elect the candidate of their choice. (Doc. 153-87 at 24.) The five percentage point increase in African American voters over the current at-large district, combined with the corresponding eight percentage point drop in white voters from the at-large district to the illustrative District 6 could create a potentially sizable shift in the election results. The only testimony before the Court is that "black voters would have a meaningful opportunity to elect candidates ... of their choice in [illustrative District 6]." (Id. ) Sumter County makes the argument that this is not the case, (Doc. 170 at 24-29), but it did not ask its expert to conduct any analysis of Wright's illustrative plan in this stage of the case. (Doc. 159 at 221:16-21.)

Sumter County suggests that a minimum of 60% of the voting-age population is needed to give African Americans an opportunity to elect candidates of their choice, citing cases in which have adopted a similar number. (Doc. 170 at 25); Ketchum v. Byrne , 740 F.2d 1398, 1415 (7th Cir. 1984). But in those cases, the courts recognized a higher number was necessary based on the evidence in the case. See, e.g., *1325Ketchum , 740 F.2d at 1415 ("During the trial, witnesses for both sides testified that 65% of total population is a widely recognized and accepted criterion in redistricting formulations.") This case has no such evidence. Again, the only testimony is that 54.5% would likely be sufficient.

Sumter County next argues that the illustrative plan would be a step backward because it trades two 49% African American districts for a 54% district and a 41% district. The Court has already found that African Americans do not have a meaningful opportunity to elect candidates of their choice in the at-large districts. They are stuck at two representatives of seven. In the illustrative plan, they would at least have an opportunity to win a third seat.

This is not to say the Court believes Wright's illustrative plan is the one which should ultimately be put into place. Africans Americans currently constitute a majority of the population in Sumter County. Their numbers, by percentage of the population, continue to grow each year. If these trends continue, African Americans will soon make up a majority of the voting age population in Sumter County, as well. At some point under the current plan-if the trends continue-one would expect black-preferred candidates to win at-large seats and constitute a majority of the School Board. Under the illustrative plan, African Americans would need to win a district where they represent roughly forty percent of the voting-age population to pick up a fourth seat. Even with a growing share of the population, the Court finds it unlikely they will be able to do so in the foreseeable future so long as voting in the country remains racially polarized. The Parties agree that Sumter County and Georgia's elected officials must be given the first opportunity to craft a remedial plan. (Doc. 140 at 3; 141 at 3.) The Court encourages the Parties and elected officials to be creative in exploring possible remedies. Redrawn district lines are but one tool available for remedying a Section 2 violation. For example, a discriminatory anti-single-shot voting rule can be fixed by removing the rule. See Holder v. Hall , 512 U.S. 874, 880, 114 S.Ct. 2581, 129 L.Ed.2d 687 (1994). The problem for African Americans in Sumter County is not the number of voters, but how often they turn out to cast votes. The Parties and the General Assembly may consider whether any tools at their disposal could meaningfully improve turnout such that African Americans have an equal opportunity to elect candidates of their choice.

Finally, Sumter County argues that the illustrative plan "inflicts a constitutional injury." (Doc. 170 at 29 (capitalization altered).) The Equal Protection Clause of the Fourteenth Amendment "prevents a State, in the absence of 'sufficient justification,' from 'separating its citizens into different voting districts on the basis of race.' " Cooper v. Harris , --- U.S. ----, 137 S.Ct. 1455, 1463, 197 L.Ed.2d 837 (2017) (quoting Bethune-Hill v. Virginia State Bd. of Elections , --- U.S. ----, 137 S.Ct. 788, 797, 197 L.Ed.2d 85 (2017) ). To prove a violation, a plaintiff must show that "race was the predominant factor motivating the legislature's decision to place a significant number of voters within or without a particular district." Miller v. Johnson , 515 U.S. 900, 916, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995). "That entails demonstrating that the legislature subordinated other factors-compactness, respect for political subdivisions, partisan advantage, what have you-to racial considerations." Cooper , 137 S.Ct. at 1463-64 (quotations and citations omitted). Here, there is no evidence that the illustrative plan subordinated any factors for race considerations. McBride testified, and Sumter County does not contest, that the illustrative plan "complie[s] with the one-person, one-vote principle, the Voting Rights Act, and traditional *1326redistricting criteria including compactness, contiguity, respect for communities of interest, [and] respect for political boundaries." (Doc. 153-87 at 5.) The plan does not raise any constitutional concerns.

Accordingly, the Court concludes that the illustrative plan-while far from perfect-is likely to give African Americans a more proportional representation on the Board of Education than does the current plan. The Parties should not take this as an indication of how the Court will view the proposed remedial plans in the next step of this case. The Parties have already begun a much more robust discussion of remedial plans in post-trial briefing. (See Docs. 174; 176; 180.) While that evidence is not before the Court at the liability stage, (see Doc. 189), the Court expects a much more expansive body of evidence to determine the effectiveness of proposed remedial plans following post-trial discovery.

CONCLUSION

The Court finds that Plaintiff Mathis Kearse Wright, Jr. has established all three Gingles factors, that the majority of the Senate Factors weigh toward him, and that he has shown an illustrative plan which is likely to give African Americans a more proportional representation on the Board of Education than does the current plan. Accordingly, the Court finds, based on the totality of the circumstances, that the at-large districts of the Sumter County Board of Education dilute African-American voting strength in violation of Section 2 of the Voting Rights Act of 1965, as amended, 52 U.S.C. § 10301.

The case now moves to a remedial stage. The Court agrees with the Parties that elected officials should have the first opportunity to remedy the unlawful election plan. See Wise v. Lipscomb , 437 U.S. 535, 540, 98 S.Ct. 2493, 57 L.Ed.2d 411 (1978) ; (Docs. 140; 141). The Court notes that the General Assembly will be in session through at least Thursday, March 29, 2018. S.R. 631, 154th Gen. Assemb., Reg. Sess. (Ga. 2018). The Sumter County Board of Elections and Registration is ORDERED to confer with Sumter County's legislative delegation and inform that Court no later than Monday, March 26, 2018 whether the General Assembly is inclined to enact a remedial plan before adjourning sine die or, if not, a timeline for when it believes a remedial plan could be adopted. While the time period is short, the Parties have already put considerable effort into their proposed remedial plans, which will greatly assist the General Assembly in its efforts.

Given the Court's holding, Wright's Motion for Preliminary Injunction (Doc. 190) is DENIED WITHOUT PREJUDICE . Following the status report from the General Assembly, the Court will consider whether the May elections must be enjoined.

SO ORDERED , this 17th day of March 2018.

1 The Parties submitted a set of material undisputed facts as Exhibit A to their proposed pretrial order. The Court has adopted the proposed pretrial order. (Doc. 134.) All references to Exhibit A will be to "MUF."

2 All demographic data is the most up-to-date available to the Court.

3 For the at-large seats, the Court accepts the updated 2012-2016 American Community Survey date rather than that in McBride's report.

4 Defendants list in their Proposed Findings of Fact and Conclusions of Law what they allege to be the race of each candidate in the 2012, 2014, and 2016 elections. (Doc. 172 at ¶¶ 19-42.) However, their citations to the record do not identify the race of any candidate. As far as the Court is aware, the only testimony adduced at trial as to the race of any candidate other than those for school board was that of Robert Edward Brady. The Court relies solely on his testimony in identifying the races of the candidates in these elections.

5 Wright pleaded a packing claim in his pro se complaint. (Doc. 1 at 7.) However, he has since abandoned the claim and offered no support for it at trial. The Court finds no evidence in support of it.

6 The Court cannot refrain from commenting on one other argument advanced by the County. McBride testified that EzI runs on 32-bit operating systems and cannot be run on newer 64-bit systems. (Doc. 160 at 128:1-8.) Sumter County contorts this fact in an attempt to discredit McBride: "Dr. McBride admitted that the program he used is so outdated that he had to adjust his computer settings to run it." (Doc. 170 at 15.) EzI is merely a program which allows a researcher to perform a mathematical analysis. The program's age may cause computer compatibility issues and slow load times, but the math underlying it never changes. The results of Adrien-Marie Legendre's regression models in 1805 would be no different if run again today. Any attempts to impugn the credibility of McBride's analysis based on the age of the program he used to run it is illogical, not credible, and completely irrelevant to the matter at hand.

7 Sumter County also wants to attribute any success by Democrats in these elections to African Americans. (See Doc. 172 at ¶ 7.) The Court declines to do so. First, in many races, the Democrat and Republican are both likely to be white. (See generally Doc. 154-10.) The Court has already explained it would be discounting such races. See Johnson , 196 F.3d at 1221. Second, there is no statistical evidence before the Court of how likely African Americans in Sumter County are to support Democrats. Sumter County relies solely on McBride's testimony that (1) "the candidate of choice in the black community would be the Democrat"-but providing no basis or statistics in support of that position, and (2) "more than possibly 92 percent of African Americans support the Democrat Party" nationally-but providing no evidence of the percentages in Sumter County. (Doc. 157 at 180:25-181:4, 185:3-8.) Third, the Court is unable to determine if there was any minority cohesion or white bloc voting in these races because no EI analysis has been run on them. Any finding by the Court that wins by Democrats are wins by the black-preferred candidate over the white-preferred candidate would be pure speculation.

8 The Court finds these numbers by multiplying the percentage of the white and black voting-age population data contained in Table 1 by the respective white and black turnout data in Table 2. While the numbers will be slightly off because the demographic data has changed since the 2014 elections took place, any error is too small to impact the Court's conclusions. McBride provided black voting-age populations for each election at the time of that election (or near to), but it does not include corresponding white voting-age population numbers. (Doc. 153-87 at 11.)

9 Wright actually claims African Americans constitute 48.1% of the voting-age population. The Court refers only the most recent demographic data before it, which puts that number at a slightly higher 49.5%.