Christian Ministerial Alliance v. Arkansas, State of

Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 1 of 48 IN THE UNITED STATES DISTRICT COURT EASTERN DISTRICT OF ARKANSAS CENTRAL DIVISION THE CHRISTIAN MINISTERIAL ALLIANCE, et al. PLAINTIFFS v. Case No. 4:19-CV-00402-JM SARAH HUCKABEE SANDERS 1, in her official capacity as Governor of the State of Arkansas, et al. DEFENDANTS ORDER Plaintiffs challenge Arkansas’s use of at-large voting for the Arkansas Supreme Court and the current district configuration for the Arkansas Court of Appeals as being violative of Section 2 of the Voting Rights Act of 1965. The Court held a bench trial on these claims from April 25-28, 2022 and on May 2, 2022. The parties submitted post-trial briefs. After consideration of all the evidence and the arguments of counsel, the Court finds as follows: I. Findings of Fact 1. Since 1874, when Arkansas adopted its current constitution, all Arkansas Supreme Court justices have been elected in statewide elections. 2. The Arkansas Supreme Court consists of seven justices. 3. Of the twenty-two states that elect their supreme courts, Arkansas and seventeen other states elect the supreme court statewide, while only four states opt for districting. 4. The Arkansas Court of Appeals, an intermediate appellate court with statewide jurisdiction, was created in 1979. 5. In 1999, the General Assembly assigned appointees to the Court of Appeals, 1 Current Arkansas Governor Sarah Huckabee Sanders and Attorney General Tim Griffen, in their official capacities, are substituted as defendants in place of Asa Hutchinson and Leslie Rutledge. Fed. R. Civ. P. 25(d). — 2 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 2 of 48 Judges Neal, Roaf, and Griffen, who are Black 2, to interim districts and made them eligible to run for reelection in the November 2000 general election. 6. They were assigned, respectively, to Districts 1, 5, and 6. 7. In 2000, those districts had non-Hispanic white voting-age population of 84%, 70%, and 73%, respectively. 8. Judges Neal, Roaf, and Griffen were re-elected in 2000 without drawing an opponent. 9. The current seven districts for the Arkansas Court of Appeals were established in 2003. 10. Five of the seven districts elect two judges to numbered positions, while the other two districts elect a single judge. 11. Each district consists of a number of whole counties. 12. Arkansas has never made any attempt to perfectly equalize its Court of Appeals districts. 13. According to the 2000 Census, the population of Arkansas Court of Appeals District 7 was 48.6% Black, and its voting-age population was 44.4% Black. 14. Because of the General Assembly’s plan for a gradual transition into the Court of Appeals districts adopted in 2003, the first election was not held in District 7 until 2008. 15. By the 2010 Census, District 7’s citizen voting-age population was 49.2% Black. 16. Today, its citizen voting-age population is roughly 50.9% Black. 17. Since its inception, District 7 has elected just one judge, Judge Waymond Brown, who is Black. 2 The Court will follow Plaintiffs’ lead in this opinion and capitalize Black as a race designation. — 3 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 3 of 48 18. Candidates cannot win judgeships in Arkansas unless they receive a majority of the votes. 19. Accordingly, Arkansas holds two types of judicial elections: general elections, and if necessary, runoffs. 20. General elections for nonpartisan offices, such as judgeships, are held on the same date as primary elections for partisan elections, called the “preferential primary elections” in the Arkansas election code. 21. Nonpartisan runoff elections, by contrast, take place “on the same date and at the same times and places as the November general election.” 22. If any judicial candidate receives a majority of the votes during the nonpartisan general election (in May or March, depending on the election cycle), that candidate wins the election for the judicial office outright, and there will be no nonpartisan runoff election for that office in November. 23. If no candidate receives a majority of the votes, the two candidates who received the greatest shares of votes are certified to a runoff election, which takes place on the same date as the November general election for partisan offices. 24. Since 2001, Arkansas’s judicial elections have been nonpartisan. See Ark. Const. amend. 80, secs. 17(A), 18(A). 25. With partisanship removed from consideration, diverse coalitions of voters have frequently come together to support the same candidates, and Black voters have elected judicial candidates of their choice at rates far in excess of their share of the population. 26. In 2010, 16.07% of Arkansans were “Any Part Black,” a demographic that includes voters who identify as only “some part Black, including Hispanic Black.” — 4 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 4 of 48 27. By 2019, that number had risen slightly to 16.60%, or almost exactly one out of six. 28. A majority of the judges and justices currently serving on Arkansas’s appellate courts were preferred by Black voters when they were elected to their current position—even excluding the Court of Appeals’ District 7, where Black voters make up a majority of the electorate. 28. Plaintiffs’ expert, William S. Cooper, developed Illustrative Plans to assess whether the Black population in Arkansas is sufficiently large and geographically compact to allow for: (1) two single-member majority-Black districts for the Arkansas Court of Appeals and (2) one single-member majority-Black district for the Arkansas Supreme Court, which currently elects seven justices at large. Mr. Cooper also reviewed current and historical demographics of Arkansas, including the socio-economic, employment, education, and health characteristics of the Black and non-Hispanic white populations. 29. Mr. Cooper is qualified to serve as an expert witness in redistricting and demographics. Since 1986, Mr. Cooper has prepared redistricting maps for hundreds of jurisdictions in Section 2 cases, comment letters under Section 5 of the Voting Rights Act, and other efforts to comply with the Voting Rights Act (“VRA”). Since the release of the 2010 Census, Mr. Cooper has developed statewide redistricting plans in nine states, as well as over 150 local redistricting plans in approximately thirty states. Mr. Cooper has qualified as an expert witness on redistricting and demographics in federal courts in approximately forty-five voting rights cases in eighteen states, having been retained by both civil rights plaintiffs and government entities. 30. To develop the Illustrative Plans, Mr. Cooper used (1) population and geographic — 5 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 5 of 48 data from the 1980 to 2020 Censuses, (2) the 2019 U.S. Census Bureau population estimates, and (3) geographic boundary files created from the U.S. Census and 1990, 2000, 2010, and 2020 Topologically Integrated Geographic Encoding and Referencing (TIGER) files. He used Maptitude for Redistricting, a geographic information system (“GIS”) software that many local and state bodies employ for redistricting. 31. Mr. Cooper used population data from the U.S. Census 1980, 1990, 2000, 2010, and 2020 PL 94-171 data files. The PL 94-171 dataset is the complete count population file designed by the Census Bureau for use in legislative redistricting. It is published in electronic format. 32. As is standard redistricting practice, Mr. Cooper disaggregated block group citizen voting age estimates to the block level by race and ethnicity based on the distribution of block group level voting age population (“VAP”) by race and ethnicity from the 2015-19 American Community Survey (“ACS”). 33. To develop the Illustrative Plans, Mr. Cooper also obtained (1) a PDF map depicting the 1979 Court of Appeals Plan online, (2) a geographic shapefile for the 2003 Court of Appeals Plan adopted by the Arkansas Legislature online, (3) 2019 population estimates from the U.S. Census Bureau online, and (4) data from the ACS – specifically, the 5-year 2008-2012 and 2015-2019 ACS Special Tabulation of citizen population and voting age population by race and ethnicity. 34. It is clear that race was the predominate factor motivating all of Mr. Cooper’s proposed districts. This fact does not invalidate his findings. See Allen v. Milligan, 143 S. Ct. 1487 (2023). 35. Mr. Cooper developed the Illustrative Plans mostly in accordance with traditional redistricting principles, including compactness; contiguity; one person, one vote; communities of — 6 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 6 of 48 interest; traditional boundaries; and non-dilution of minority voting strength. While Mr. Cooper’s illustrative districts stretch the bounds of geographic compactness, the Court will accept his plans as compact. 36. The traditional redistricting principles Mr. Cooper applied and the manner in which he applied them were consistent with the established approach to redistricting that has been used to draw both legislative and judicial districts in Arkansas. For example: a. Compactness: Compare Trial Tr. vol. 1, 106:4-14 (Cooper Direct), and PTX 073 at ¶¶ 15, 42, 60 (Cooper Decl.), with PTX 077 at 3 (Board of Apportionment, “Redistricting Standards and Requirements”), and PTX 467 at 280:14-15. b. Contiguity: Compare Trial Tr. vol. 1, 106:4-14, and PTX 073 at ¶¶ 15, 42, 60, with PTX 077 at 3. c. Population equality: Compare Trial Tr. vol.1, 120:17–19, and PTX 073 at ¶¶ 15, 42, 60, with PTX 077 at 2, and PTX 467 at 280:2-4. Nondilution of Minority Voting Strength: Compare Trial Tr. vol. 1, 128:18–129:3, and PTX 073 at ¶¶ 15, 42, 60, with PTX 077 at 2, and PTX 467 at 280:12-13. d. Communities of Interest: Compare Trial Tr. vol. 1, 107:2-3, 125:12-15, and PTX 073 at ¶¶ 15, 42, 60, with PTX 077 at 3, and PTX 467 at 280:10–11. e. Maintaining Traditional Boundaries: Compare Trial Tr. vol. 1, 106:10-24, with PTX 077 at 3, and PTX 467 at 280:5-9. 37. While some of this evidence about redistricting principles refers to legislative districting, Defendants’ own prior redistricting expert, former Assistant Attorney General Tim Humphries, testified that Arkansas applies the same principles to both legislative and judicial redistricting. Defendants have not disputed that testimony. 38. Defendants have not contested any of Plaintiffs’ evidence concerning the nature of — 7 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 7 of 48 traditional redistricting principles as they have been applied in Arkansas. 39. District 7 in Illustrative Appellate Court (“AC”) Plan 3 has an Any-Part Black VAP of 50.30%. District 8 in Illustrative AC Plan 3 has an Any-Part Black VAP of 50.19%. 40. The Black populations in Illustrative AC Plan 3, as measured by the non-Hispanic Black citizen VAP and Any-Part Black VAP, are sufficiently numerous to constitute the majority of the voting age population in two single-member, majority-Black districts for the Arkansas Court of Appeals. 41. District 7 in Illustrative Supreme Court (“SC”) Plan 3 has an Any-Part Black VAP of 50.75%. 42. The Black populations in Illustrative SC Plan 3, as measured by the non-Hispanic Black citizen VAP and Any-Part Black VAP, are sufficiently numerous to constitute the majority of the voting age population in a single-member, majority-Black district for the Arkansas Supreme Court. 43. A visual estimation and comparison to current Arkansas Court of Appeals districts shows that the current configuration of District 7 is not ideal and the illustrative District 7 is worse. This however does not prevent this Court from finding that the majority Black districts in Plaintiffs’ Illustrative Plans are geographically compact enough to pass muster. 44. As noted above, they also comport with traditional redistricting principles as they have been applied in Arkansas. In fact, the Illustrative Plans perform as well as—or in some cases, better than—districting plans of the State’s own design when measured against Arkansas’s own traditional redistricting criteria. 45. The districts in the Illustrative Plans are contiguous. 46. The districts in the Illustrative Plans have reasonably equal populations, within — 8 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 8 of 48 plus or minus five percent. 47. The districts largely follow traditional boundaries, and the boundaries in the Illustrative Plans correspond to existing voting tabulation districts. Plaintiffs’ Illustrative Plans also adhere to Arkansas’s traditional principle of keeping counties and other political subdivisions intact “[w]hen possible.” For both the Supreme Court and Court of Appeals, Plaintiffs’ Illustrative Plans maintained most county boundaries and split just a small handful of counties. In the few instances where Plaintiffs’ Illustrative Plans do split counties, Mr. Cooper had multiple reasons for doing so as part of a broader “balancing” act; he split counties to better equalize population, preserve minority voting power, or both. Mr. Cooper did not split precincts, which is a best practice that simplifies administration for election officials. This is consistent with traditional redistricting principles as applied in Arkansas, as evidenced by the testimony of Defendants’ former redistricting expert, Tim Humphries, who explained that keeping counties intact “was not a priority at all.” Mr. Humphries also testified that county splitting “wouldn’t be a problem” “as long as you split them on precinct lines,” as Mr. Cooper did in this case. The few county splits in the Illustrative Plans mirror the legislature’s own recently enacted legislative and congressional maps, which confirms the general permissibility and consistency of the county splits in the Illustrative Plans with Arkansas redistricting principles. 48. The Court of Appeals Apportionment Commission historically considered plans that split as many or more counties than the Illustrative Plans. And other pronouncements by Defendants show an understanding that county-splitting is commonplace. The Secretary of State’s “Redistricting 101” guide shows that “changes within” a county are commonplace and familiar to deal with as a technical matter during redistricting. For all these reasons, Mr. Cooper is correct that county splits are “very common,” both in districting generally and in Arkansas — 9 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 9 of 48 specifically. 49. The districts in the Illustrative Plans respect communities of interest. The Black residents of the districts share some but few demographic commonalities that span counties in the Delta and Lower Arkansas, as well as Jefferson and Pulaski Counties, including socio-economic status, employment status, educational attainment, and health outcomes. Black residents of the districts do have substantially overlapping familial, civic, and religious interests. 50. Retired Judge Marion Humphrey testified about his experience moving from the Delta to Little Rock and about various clubs and churches with membership that extend across both areas. Veter Executive Director of the Arkansas Community Institute Neal Sealy testified about her family ties across the Delta and Pulaski County. Executive Director of the Arkansas Community Institute Neal Sealy testified ACI’s work and membership spanning Little Rock and the Delta. Reverend Maxine Allen testified about how patterns of pushout, redlining, and gentrification have determined housing patterns among congregants of CMA membership spanning from Little Rock to the Delta. Senator Joyce Elliott described shared cultural and historic experiences of Black Arkansans in Little Rock and the Delta. Although the proposed districts in the Illustrative Plans combine rural and urban voters to some degree, this is permissible and indeed common practice in Arkansas districting and does not undermine the conclusion that the Illustrative Plans respect communities of interest. 51. Mr. Cooper’s proposed map for the Supreme Court combines regions of the Delta, Jefferson County, and parts of Pulaski County within his illustrative majority-minority district. His illustrative map for the Court of Appeals creates two districts that, one that encompasses the Delta region as well as parts of Central Arkansas, including parts of Pine Bluff, and a second that encompasses parts of Pulaski County (including Little Rock) and Jefferson County, as well as — 10 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 10 of 48 Dallas, Cleveland, Ouachita, and Calhoun Counties. 52. District 7 in Illustrative AC Plan 3 encompasses the Delta counties that border the Mississippi River, part of Mississippi County, and part of Jefferson County. District 8 in Illustrative AC Plan 3 encompasses the remainder of Jefferson County, part of Pulaski County, and all of Dallas, Cleveland, Ouachita, and Calhoun counties. 53. District 7 in Illustrative SC Plan 3 includes the six Delta counties that border the Mississippi River, all of Arkansas, Jefferson, Lincoln, Cleveland, Dallas, and Monroe, and St. Francis Counties, and part of Pulaski County. 54. Census and American Community Survey data show that Black residents in rural counties in river-adjacent Delta and Lower Arkansas, as well as in Jefferson and Pulaski Counties, share similar socio-economic characteristics, educational attainment rates, unemployment rates, home ownership rates, levels of access to health insurance, and levels of access to transportation. In general, the socio-economic, educational, employment, and health disparities between Black and white Arkansans in Pulaski County—the county with the highest Black population in the state—are substantially similar to the relative disparities in these metrics between Black and white Arkansans statewide. 55. While it is a very close call as to geographic compactness, the Black population is sufficiently numerous and geographically compact to comprise a majority of the voting age population in two Court of Appeals districts and one Supreme Court district in Plaintiffs’ Illustrative Plans. 56. Defendants’ expert Dr. John Alford analyzed every contested election for an appellate judgeship in Arkansas over the last twenty years, a total of thirty elections. 57. Plaintiff’s expert Dr. Baodong Liu did not analyze twenty-six of those elections — 11 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 11 of 48 because a Black candidate did not run. 58. Dr. Alford’s analysis is unrebutted by Dr. Liu’s analysis and proves that Black voters’ preferred candidates in elections for appellate judgeships had a robust success rate, which has steadily risen in recent years. 59. In contrast to Dr. Alford’s data-rich analysis, Dr. Liu analyzed only four judicial contests—all of which occurred between 2004 and 2010, and three of which involved the same candidate, Judge Wendell Griffen. 60. A representative sample here requires considering more than the four elections for appellate judgeships that Dr. Liu considered. 61. Dr. Liu’s limited analysis of endogenous elections provides little information directly relevant to the analysis of racially polarized voting in Arkansas judicial elections today. 62. The results of Judge Griffen’s elections in 2004, 2006, and 2008 are not particularly probative evidence that Black voters are politically cohesive in elections for appellate judgeships, but only that over a decade ago, Black voters cohesively supported Judge Griffen. 63. Dr. Liu’s analysis ignores almost 90% of the elections relevant to this case (twenty-six out of thirty). 64. Little Rock Mayor Frank Scott Jr., who is Black, defeated white candidates in his last two elections, 2018 and 2022. Pulaski County Sheriff Eric Higgins, who is Black, defeated white candidates in his last two elections, 2018 and 2022. Pulaski County Circuit Clerk Theresa Hollingsworth, who is Black, defeated white candidates in her first election in 1018 and ran unopposed in 2022. While these three elections do not involve judicial races, they are illustrative of recent Black candidate success in central Arkansas elections. — 12 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 12 of 48 65. Since Arkansas judicial elections became nonpartisan, there have been fifteen contested elections for the Arkansas Supreme Court, including both nonpartisan general elections and nonpartisan runoff elections, the first of which occurred in 2004. 66. Of these fifteen races, twelve were two-way races, whether nonpartisan general elections with only two candidates, or nonpartisan runoff elections. 67. The other three elections featured three candidates, none of whom won a majority. In these three-way races, success was advancing to the runoff. 68. Table 1 attached as Exhibit 1 to this order lists the results of these fifteen elections for the Arkansas Supreme Court. (Note: Where a name appears in bold type, this candidate is Black.) 69. Across the contested supreme-court races over the last twenty years, Black voters’ preferred candidates were successful a majority of the time. 70. There have been twelve two-way races for the Arkansas Supreme Court in the last twenty years. 71. In seven of the twelve, or 58%, the candidate who received a majority of Black voters’ support won the election. 72. In two-way races prior to 2010, Black voters’ preferred candidates had a 50% success rate (two out of four). 73. Since 2010, their preferred candidates have won 63% (five out of eight) of two-way races. 74. In the five two-way races since 2014, Black voters’ preferred candidates have won four, which is an 80% success rate. 75. Black voters in Arkansas have had increasing success in recent years at electing — 13 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 13 of 48 their preferred supreme-court candidates in two-way races. 76. Black voters’ preferred candidates have been even more successful in three-way races for the Arkansas Supreme Court over the last twenty years. 77. There have been three races for the Arkansas Supreme Court since 2001 that included three candidates, and in each of these races the candidate who received the greatest support from Black voters advanced to the runoff. 78. Black voters’ second-choice candidates have also been successful. 79. In two of the races (the 2010 race for Place 6, and the 2004 race for Place 4), Black voters had a clear second-choice candidate. 80. In each race, Black voters supported two different candidates at essentially the same level (Justice Baker and Judge Fox, in 2010, and former Justice Gunter and former Judge Kilgore, in 2004), with only about 2–3% difference in Black voters’ support for each candidate, which is a margin that is so close that there is a chance the second-choice candidate was, in fact, Black voters’ preferred candidate. 81. In these two races, the second choice also advanced to the nonpartisan runoff election. 82. Put differently, in these two three-way races for the Arkansas Supreme Court, Black voters’ least favorite candidate was defeated, and their first and second choices advanced to the nonpartisan runoff election. 83. In the third three-way race since 2001, which was the 2018 election for Place 3, Black voters’ most-preferred candidate (Justice Hudson) received 10% more support from Black voters than Black voters’ second- and third-choice candidates (Sterling and Hixson), who were statistically tied. — 14 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 14 of 48 84. Black voters’ preferred candidates have been usually successful in winning nonpartisan runoff elections. 85. In two of the three nonpartisan runoff elections (the 2018 race for Place 3, and the 2004 race for Place 4), Black voters’ preferred candidate in the nonpartisan general election (Justice Hudson in 2018, and former Justice Gunter in 2004) remained their preferred candidate in the nonpartisan runoff, and that candidate won the seat. 86. In the third election (the 2010 race for Place 6), Black voters’ close second choice in the general election, now-Justice Baker, became their overwhelmingly preferred candidate in the runoff. Justice Baker prevailed in the runoff. 87. Over the last twenty years, combining two-way races and three-way races, Black voters’ first-choice candidates for the Arkansas Supreme Court were successful ten out of fifteen times, or 67%. 88. Adding close second-choice candidates to the mix, Black voters’ preferred candidates were successful twelve out of seventeen times, or 71%. 89. Since 2010, Black voters’ first-choice candidates for the Arkansas Supreme Court have been successful in seven out of ten races, or 70%. 90. Including close second-choice candidates since 2010, Black voters’ preferred candidates for the Arkansas Supreme Court have been successful in eight out of eleven races, or 73%. 91. Since 2014, Black voters’ preferred candidates for the Arkansas Supreme Court have been successful in five out of six races, or 83%. 92. Therefore, Black voters’ preferred candidates for the Arkansas Supreme Court are not usually defeated by bloc voting by white voters. — 15 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 15 of 48 93. Of the seven justices on the Arkansas Supreme Court, five were Black voters’ strongly preferred candidates, winning majorities of Black voters’ support between an estimated 57% and 67%: Justices Baker, Hudson, Webb, Womack, and Wynne. 94. Justice Wood ran unopposed in 2016. 95. Only one of the court’s members, Chief Justice Kemp, defeated a candidate preferred by Black voters. 96. There have also been fifteen contested elections, including both nonpartisan general elections or nonpartisan runoff elections, for the Arkansas Court of Appeals since judicial elections became nonpartisan, the first of which occurred in 2008. Thirteen of these were two-way elections, whether general elections with only two candidates or runoffs. 97. Two elections featured three candidates, none of whom won a majority; in those races, success was advancing to the runoff. 98. Table 2 attached as Exhibit 2 to this order lists the results of these fifteen elections for the Arkansas Court of Appeals. (Note: Where a name appears in bold type, this candidate is Black.) 99. Since Arkansas judicial elections became nonpartisan, Black voters’ preferred candidates for the Arkansas Court of Appeals were successful nearly a majority of the time, a pattern that has become more pronounced in recent years. 100. Beginning with the thirteen two-way races, in six out of thirteen, or 46%, the candidate who received a majority of the support of Black voters won the election. 101. Before 2010 Black voters’ preferred candidates won one out of three races. 102. Since 2010 Black voters’ preferred candidates have won five out of ten two-way races. — 16 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 16 of 48 103. Since 2014, Black voters’ preferred candidates have won four out of six two-way contests, or 67%. 104. Since Arkansas’s judicial elections became nonpartisan, Black voters’ preferred candidates for the Arkansas Court of Appeals have usually prevailed in two-way contests. 105. Black voters’ preferred candidates have also usually prevailed in three-way races. 106. There have been two three-way races for the Arkansas Court of Appeals since 2001 (in 2016 for District 5, and in 2012 for District 1, Place 2). 107. In both these races, the candidate who received the greatest support from Black voters (McMenis in 2016, and Robertson in 2012) advanced to the runoff. 108. Also in both races, Black voters had a clear second-choice candidate (Judge Klappenbach in 2016, and Lusby in 2012). 109. And in one out of two, the second-choice candidate advanced to the runoff (Judge Klappenbach in 2016). 110. In the two ensuing runoffs, Black voters’ preferred candidates went one for two, with Judge Klappenbach prevailing in 2016. 111. Over the last twenty years, combining contested two- and three-way races for the Arkansas Court of Appeals, Black voters’ first choices were successful eight out of fifteen times, or 53%. 112. Including second-choice candidates with that group, Black voters’ preferred candidates were successful nine out of seventeen times, or 53%. 113. Since 2010, Black voters’ first-choice candidates for the Arkansas Court of Appeals have been successful in seven out of twelve races, or 58%. 114. Including second-choice candidates since 2010, Black voters’ preferred candidates — 17 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 17 of 48 for the Arkansas Court of Appeals have been successful in eight out of fourteen races, or 57%. 115. Since 2014, Black voters’ first-choice candidates for the Arkansas Court of Appeals have been successful in five out of seven races, or 71%. 116. Including second-choice candidates since 2014, Black voters’ preferred candidates for the Arkansas Court of Appeals have been successful in six out of eight races, or 75%. 117. Since judicial elections became nonpartisan in Arkansas, Black voters’ preferred candidates for the Arkansas Court of Appeals have consistently won over half of the contested elections. 118. In the most recent elections, the success rate of Black voters’ preferred candidates for the Arkansas Court of Appeals has been greater than two-thirds of contested elections, and by some metrics has even approached three-quarters of those elections. 119. Therefore, Black voters’ preferred candidates for the Arkansas Court of Appeals have not usually been defeated by bloc voting by white voters. 120. In the thirty contested elections for appellate judgeships in Arkansas since these elections became partisan, eighteen out of thirty of Black voters’ top-choice candidates were successful, and Black voters gave about 56% of their support, on average, to their top choices. 121. Over the thirty contested elections for the Arkansas Supreme Court and Arkansas Court of Appeals since those courts became nonpartisan, Black voters supported their top choice by an average of 56.2%. 3 122. Black voters supported their top choice for the Arkansas Supreme Court by an average of 55.6%. 3 The percentages in this and the following paragraphs are determined by averaging the column entitled “Estimated Black Vote” for the candidate in each election who received the highest support from Black voters. — 18 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 18 of 48 123. Black voters supported their top choice for the Arkansas Court of Appeals by an average of 56.7%. 124. In races since 2010, Black voters have supported their top choice for both appellate courts by an average of 54.8%. 125. In races since 2010, Black voters have supported their top choice for the Arkansas Supreme Court by an average of 54.9%. 126. In races since 2010, Black voters supported their top choice for the Arkansas Court of Appeals by an average of 54.6%. 127. Limiting the analysis to only two-way races, Black voters have given, on average, 59.6% of their vote to their first-choice candidate, and 58.6% since 2010. 128. Plaintiffs’ illustrative remedial districts for the Arkansas Court of Appeals districts would have a Black voting-age population of only 50.2% to 51.1%. 129. In the last decade, Black voters have given, on average, only 58.2% of their support to their preferred Court of Appeals candidates in two-way races. 130. In a 51% Black district, assuming equal turnout between Black and white voters, a candidate with 58.2% of the vote from Black voters would need 41.4% of the vote from white voters to win the election. 4 131. Similarly, Plaintiffs’ illustrative remedial Supreme Court districts would have a Black voting-age population of only 51.6% to 53.6%. 132. In the last decade, Black voters have only given, on average, 59.1% of their 4 At these rates, Black voters would only provide such a candidate about 29.7% of the total vote (58.2% * 0.51 = 29.7%). This candidate would thus need to look to white voters for the remaining 20.3% of the total vote necessary to be elected (50% – 29.7% = 20.3%). Obtaining that would require this candidate to get 41.4% of the votes from white voters (41.4% * 0.49 = 20.3%). (Note that 49% is the percentage of white voters in this hypothetical district.) — 19 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 19 of 48 support to their preferred Supreme Court candidates in two-way races. 133. Assuming a 53.6% Black district, with equal levels of turnout among Black and white voters, a candidate with support from 59.1% of Black voters would need support from 39.4% of white voters to win the election. 5 134. In elections for the Arkansas Supreme Court and the Arkansas Court of Appeals, 40–50% of white voters usually support Black voters’ preferred candidates. 135. Black voters regularly express distinct—if on average less than cohesive— preferences between white candidates, in several instances supporting one white candidate over another by landslide margins, for example, in 2020, supporting Justice Webb by a margin of two-to-one. 136. Two years before the 2004 election, Judge Griffen was admonished by the Judicial Discipline and Disability Commission for urging the Arkansas General Assembly to defund the University of Arkansas on the ground that it had, in his view, too few Black students and faculty members. Jake Bleed & Linda Satter, Decision on judges hits close to home, Ark. Democrat-Gazette, Aug. 5, 2005, Defs.’ Ex. 6. 137. One year before the election, the Arkansas Supreme Court vacated the admonishment in a high-profile 4–3 decision. Bleed & Satter, supra, Decision on judges hits close to home; see Griffen v. Ark. Judicial Discipline & Disability Comm’n, 130 S.W.3d 524, 526 (Ark. 2004) (declaring disciplinary action invalid, but describing Judge Griffen’s comments as a “call[] upon the legislators to engage in economic retaliation during the legislative session” 5 Similar arithmetic applies here. Black voters would provide this candidate 31.7% of the total vote (59.1% * 0.536 = 31.7%). Thus, this candidate would need to obtain 18.3% of the total vote from white voters (50% – 31.7% = 18.3%). And obtaining that would require achieving support from 39.4% of white voters (39.4% * 0.464 = 18.3%). (Note that 46.4% is the percentage of white voters in this hypothetical district.). — 20 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 20 of 48 against the University of Arkansas). 138. Another judicial-discipline investigation, not dismissed until after the 2004 election, considered charges that Judge Griffen had, supposedly in his capacity as pastor, advocated for the release of a convicted robber before the parole board. James Jefferson, Commission tosses complaint against appeals judge, Ark. Democrat-Gazette, Nov. 20, 2004, Defs.’ Ex. 5. 139. Around the time of these elections, Judge Griffen openly identified as a prochoice candidate. 140. During the 2004 campaign itself, Judge Griffen went on a radio show and, in response to a caller’s asking about his views on abortion, declared himself “one of those prochoice ministers” who did not “like the idea of government telling women how to handle their pregnancies.” Michael R. Wickline, High court hopeful sees no problem in voicing his opinion, Ark. Democrat-Gazette, May 15, 2004, Defs.’ Ex. 4. 141. His opponent, Jim Hannah, by contrast, refused to publicly air his views, saying that doing so would violate judicial ethics rules. Wickline, supra, High court hopeful sees no problem in voicing his opinion. 142. Judge Griffen proceeded to lose by a landslide margin to Hannah, receiving only an estimated 58–61% of Black voters’ support and 32–33% of white voters’ support. 143. During the 2006 election, Judge Griffen was the subject of three separate judicial misconduct investigations. Jake Bleed, Griffen at risk of impeachment, Ark. Democrat-Gazette, May 3, 2006, Defs.’ Ex. 8. 144. One of the investigations grew out of a 2005 speech in which Judge Griffen called prominent evangelical leaders “pimps of piety” who called themselves “the Christian right” but — 21 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 21 of 48 were “not Christian or right”; criticized then-President George W. Bush, then-Vice President Dick Cheney, and Associate Justice of the U.S. Supreme Court Clarence Thomas; and claimed that people who came to assist with relief efforts after Hurricane Katrina were “hypocrites” who “had nothing to do with poor people before Katrina.” Debra Hale-Shelton, In NAACP speech, judge blasts storm response, Ark. Democrat-Gazette, Sept. 11, 2005, Defs.’ Ex. 7. 145. Judge Griffen received 38% of white voters’ support that election year. 146. In the 2008 election for the Arkansas Court of Appeals, Judge Griffen’s support from Black voters rose to 85–87%, but his share of the vote from white voters dropped to 22– 23%. 147. In 2008, Judge Gruber “offer[ed] no criticism of Griffen and g[ave] no reason why voters shouldn’t re-elect him,” focusing on her own qualifications instead. Seth Blomeley, Outspoken judge faces challenger, Ark. Democrat-Gazette, May 4, 2008, Defs.’ Ex. 9. 148. When he lost to Judge Rita Gruber, Judge Griffen remarked to the press, “I am sure there were people who didn’t like what I said and people who didn’t mind what I said but didn’t like the idea of a judge talking.” Seth Blomeley, Gruber unseats Griffen on court, Ark. Democrat-Gazette, May 21, 2008, Defs.’ Ex. 10. 149. Judge Gruber, for her part, told reporters that though she had not run on Judge Griffen’s judicial-discipline investigations, voters had asked her about them anyway. Blomeley, supra, Gruber unseats Griffen on court. 150. Judge Griffen is not the typical judicial candidate, Black or white. Unlike most judicial candidates, he is a well- known pastor who publicly denounces what he perceives as “racism and classism in society.” Hale-Shelton, supra, In NAACP speech, judge blasts storm response, at 5B. — 22 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 22 of 48 151. Immediately after his last defeat, a Black candidate, Evelyn Moorehead, ran for the Arkansas Supreme Court and did not polarize the electorate. 152. In 2010, the election cycle immediately following Judge Griffen’s defeat, Moorehead ran for a position on the Supreme Court and, in a field of three candidates, only received between 30% and 36% of the vote from Black voters. 153. Both Black and white voters preferred Justice Baker and Judge Fox over Moorehead. 154. The results of races for partisan office in Arkansas reveal partisan polarization. 155. Table 3 attached as Exhibit 3 to this order lists the results of contested partisan elections over the last 20 years. (Note: Where a name appears in bold, this candidate is Black.) 156. These results show partisan polarization. 157. With contests in the same years between white Democrats and white Republicans, the voting patterns are identical to contests between Black Democrats and white Republicans. 158. For example, in 2018, Black voters supported the Black Democratic candidate (Anthony Bland) for Lieutenant Governor by a margin of 88% to 7%, while white voters opposed him by a margin of 72% to 25%. 159. But Black voters also supported the white Democratic candidate (Jared Henderson) for Governor in 2018 by almost the exact same margin—87% to 10%—and white voters also opposed that white candidate by almost the same margin—73% to 24%. 160. When Black voters were offered a choice between a Black Republican and a white Democrat, in the 2006 State Treasurer race, they overwhelmingly chose the white Democrat, by a margin of 88% to 12%—essentially the same number as the 2018 contests for Governor and Lieutenant Governor. — 23 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 23 of 48 161. Black and white voters both supported President Obama in 2012 by a similar margin to their preference for President Biden in 2020. 162. One Black judge has been repeatedly elected to the Court of Appeals: Judge Waymond Brown in District 7. 163. When District 7 was drawn in 2003, its voting-age population was only 44% Black, a figure that rose to 49%, still a minority, by 2010, two years after Judge Brown’s first election. 164. Today, the voting-age population in Judge Brown’s district is 51% Black. 165. Judge Brown won election in 2008 and has continued to win since. He has only been opposed by Eugene Hunt, a Black candidate. 166. Three Black judges have been appointed to the Court of Appeals and then reelected unopposed in majority-white districts: Judge Olly Neal, Judge Griffen, and Judge Andree Roaf, at one point making up a quarter of the Court of Appeals. 167. Of the members of the Arkansas House of Representative, 15% are Black, a number roughly equal to the Black population in Arkansas. The evidence establishes that Black voters’ preferred candidates usually won in elections for the Arkansas Supreme Court and Court of Appeals. 168. There have been no racial appeals in Arkansas judicial elections whatsoever. II. Conclusions of Law A. The Governor and the Attorney General 169. The Governor and the Attorney General are immune from suit pursuant to the doctrine of sovereign immunity reflected in the Eleventh Amendment unless the narrow exception announced in Ex parte Young, 209 U.S. 123 (1908), applies. Under this exception, “a suit to — 24 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 24 of 48 enjoin a state official's enforcement of state legislation on the ground that the official's action would violate the Constitution is not a suit against the State, and is thus not barred by the Eleventh Amendment, so long as the official has “some connection with the enforcement of the act.” Digital Recognition Network, Inc. v. Hutchinson, 803 F.3d 952, 956–57 (8th Cir. 2015) (quoting Young at 155-60). 170. Ex parte Young’s exception does not apply to the Governor and the Attorney General here, because they have no connection to enforcement of the policies and practices that Plaintiffs challenge. Id. at 960; see also Calzone v. Hawley, 866 F.3d 866, 869 (8th Cir. 2017). 171. The Governor’s general power to execute Arkansas law does not give him any enforcement method over the system for electing Arkansas’s appellate judges. See Ark. Const. art. VI, secs. 2, 7. 172. A “governor’s general-enforcement authority is not ‘some connection’ to enforcement” of a particular law, unless “that authority gives the governor methods of enforcement” for that law. Church v. Missouri, 913 F.3d 736, 749 (8th Cir. 2019). 173. The Eighth Circuit has specifically held that “the executive authority of the [Arkansas] governor” under the Arkansas Constitution is not the sort of connection to challenged state law or policy that Ex parte Young requires. Dig. Recognition Network, 803 F.3d at 960 (citing Ark. Const. art VI, secs. 2, 7). 174. The Governor’s general-enforcement authority does not make him a proper defendant to this lawsuit. 175. The Governor’s power to sign legislation into law does not give him any enforcement method over the system for electing Arkansas’s appellate judges. See Ark. Const. art. VI, sec. 15. — 25 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 25 of 48 176. State officials cannot be sued under Ex parte Young “for their legislative acts.” Church, 913 at 751 (quoting Sup. Ct. of Va. v. Consumers Union of U.S., Inc., 446 U.S. 719, 732 (1980)). 177. The Eighth Circuit has also held the Arkansas Attorney General was not a proper defendant in a challenge to an Arkansas statute, because the Attorney General lacked authority to enforce it. Dig. Recognition Network, 803 F.3d at 962. 178. The Attorney General’s role as “the attorney for all state officials, departments, institutions, and agencies” does not confer any enforcement method over the system for electing Arkansas’s appellate judges. Ark. Code Ann. § 25-16-702(a). 179. The Attorney General’s duty to “maintain and defend the interests of the state in matters before the United States Supreme Court and all other federal courts” does not confer any enforcement method over the system for electing Arkansas’s appellate judges. Ark. Code Ann. § 25-16-703(a). 180. The Attorney General’s oath to support the U.S. Constitution does not confer any enforcement method over the system for electing Arkansas’s appellate judges. See 4 U.S.C. § 101. 181. As an initial matter, the Constitution and the Voting Rights Act are not synonymous. Thornburg v. Gingles, 478 U.S. 30, 35 (1986) (explaining important differences between constitutional and VRA analyses). 182. Even if the oath to support the U.S. Constitution were broader to include other federal laws like the Voting Rights Act, such an oath would not confer on the Attorney General any state-law power that the office otherwise lacks but would only require the Attorney General to exercise the office’s state-law power in accordance with federal law. — 26 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 26 of 48 183. The Attorney General’s duty to “provide legal assistance to the State Board of Election Commissioners in answering questions regarding election laws” does not confer any enforcement method over the system for electing Arkansas’s appellate judges. Ark. Code Ann. § 7-4-101(g). 184. The Eighth Circuit has held that “[t]he Arkansas attorney general’s authority to advise state officials on the constitutionality of [a state statute], by itself, does not suffice to establish ‘some connection with the enforcement’ of [that state statute].” Dig. Recognition Network, 803 F.3d at 962. 185. Because the Governor and the Attorney General lack “methods of enforcement” for the laws Plaintiffs challenge, they are not proper Ex parte Young defendants. Church, 913 F.3d at 749. 186. For largely the same reasons, there is no “causal connection” between Plaintiffs’ alleged injury and the conduct of either the Governor or the Attorney General, therefore, Plaintiffs lack standing to sue these two Defendants. See Dig. Recognition Network, 803 F.3d at 957-59; Tex. Democratic Party v. Abbott, 961 F.3d 389, 401 (5th Cir. 2020); see also Lujan v. Defs. of Wildlife, 504 U.S. 555, 560-61 (1992). B. Section 2 187. Section 2 of the Voting Rights Act prohibits voting practices that “result[] in a denial or abridgment of the right . . . to vote on account of race or color,” which “is established if ” the members “of a class of citizens . . . have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 52 U.S.C. 10301. — 27 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 27 of 48 188. In order to establish a Section 2 violation, plaintiffs must first prove by a preponderance of the evidence the three preconditions established in Gingles: “(1) the racial group is sufficiently large and geographically compact to constitute a majority in a single-member district; (2) the racial group is politically cohesive; and (3) the majority votes sufficiently as a bloc to enable it usually to defeat the minority's preferred candidate.” Bone Shirt v. Hazeltine, 461 F.3d 1011, 1018 (8th Cir.2006). Failure to prove even one of these three preconditions is fatal to Plaintiffs’ claims, although proof of all three does not, standing alone, entitle Plaintiffs to relief. See Cooper v. Harris, 137 S. Ct. 1455, 1472 (2017); Bartlett v. Strickland, 556 U.S. 1, 11 (2009); Johnson v. De Grandy, 512 U.S. 997, 1012 (1994). 189. If plaintiffs satisfy each of the three Gingles preconditions, they still must go on to show “under the totality of circumstances that the political process is not equally open to minority voters” to prevail on their Section 2 claim. Allen v. Milligan, 143 S. Ct. 1487, 1503 (2023) (cleaned up). C. Gingles I 190. The first precondition, Gingles I, requires a plaintiff to show that the minority group is sufficiently large and geographically compact to constitute a majority in a single-member district. Gingles, 478 U.S. at 50. This is “needed to establish that the minority has the potential to elect a representative of its own choice in some single-member district.” Allen at 1503 (quoting Growe v. Emison, 507 U.S. 25, 40 (1993)). The “ultimate end of the first Gingles precondition is to prove that a solution is possible, and not necessarily to present the final solution to the problem.” Bone Shirt , 461 F.3d at 1019 (quotation omitted); see also Clark v. Calhoun Cnty., 21 F.3d 92, 95 (5th Cir. 1994). Gingles I sets a minimum threshold requirement to distinguish between colorable Section 2 claims and those that have no chance of success. See — 28 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 28 of 48 Bartlett, 556 U.S. at 21. Plaintiffs typically satisfy Gingles I by drawing hypothetical majority-minority districts. See Bone Shirt, 461 F.3d at 1018; Clark v. Calhoun Cnty., 88 F.3d 1393, 1406 (5th Cir. 1996). 191. The Supreme Court has held that a bright-line 50% plus one rule applies to determine numerosity; it is an objective test. See Bartlett, 556 U.S. at 18; see also Cottier v. City of Martin, 604 F.3d 553, 571 (8th Cir. 2010) (Cottier III) (“[A] minority need only make up more than 50 percent of the voting-age population in the relevant geographic area [to satisfy the first Gingles factors].” (cleaned up) (emphasis in original). Further, when Black voters are the only minority group whose exercise of the franchise is at issue, “it is proper to look at all individuals who identify themselves as [B]lack.” Georgia v. Ashcroft, 539 U.S. 461, 473 n.1 (2003) (emphasis in original). Accordingly, the proper demographic measure to use in assessing numerosity is “Any-Part Black VAP.” 192. Defendants did not challenge Plaintiffs’ satisfaction of the numerosity requirement or the use of the Any-Part Black VAP at trial or in their post-trial briefing. The Court concludes Plaintiffs have established that numerosity requirement. See Giron v. City of Alexander, 693 F. Supp. 2d 904, 936 (E.D. Ark. 2010) (noting waiver of arguments not raised by parties at trial or in post-trial briefing). 193. To analyze the second element of the first Gingles precondition, compactness, the Court must consider whether a proposed plans is consistent with traditional districting criteria such as “respect for political and administrative boundaries, geographic considerations, and communities of interest” (Bone Shirt, 461 F.3d at 1019); contiguity (Miller v. Johnson, 515 U.S. 900, 916 (1995); and population equality (Shaw v. Reno, 509 U.S. 630, 651–52 (1993)). 194. The Court finds that Plaintiffs’ Illustrative Plans sufficiently maintain traditional — 29 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 29 of 48 boundaries by minimizing split counties and avoiding split precincts entirely. The plans split a single county for the Arkansas Supreme Court districts, and just two or three counties for the Court of Appeals districts. While Plaintiffs need not prioritize maintaining county lines to satisfy Gingles I, they have certainly minimized splitting counties here. Further, where Plaintiffs did split counties, they did not split any precincts. Mr. Cooper’s testimony supports the conclusion that it is common practice in Arkansas redistricting to split counties along precinct lines in a redistricting plan. 195. The Court concludes that Plaintiffs’ Illustrative Plans, while spread out, are sufficiently geographically compact. 196. In analyzing whether a proposed plan maintains communities of interest, “[a] State is free to recognize communities that have a particular racial makeup” so long as there is “some common thread of relevant interests.” Miller, 515 U.S. at 920. These shared characteristics may include social and economic needs of the communities. For example, in Theriot v. Parish of Jefferson, the Fifth Circuit found that a majority-Black district for the Jefferson Parish Council included “low-income residents who are less-educated, more often unemployed, and more poorly-housed” and thus shared “common social and economic needs.” 185 F.3d 477, 486 (5th Cir. 1999). The Court held that, “[g]iven the common thread which binds the [B]lack voters within [that district], they are entitled to an effective voice in the electoral process and to an influence over the outcome of elections.” Id. at 487 (internal quotation and citation omitted). Considering the testimony of Senator Joyce Elliott, retired Judge Marion Humphrey, Executive Director of the Arkansas Community Institute Neal Sealy, and Reverend Maxine Allen, the Court has no trouble finding that there is a common thread that binds Black voters in Arkansas, including those who reside in the Delta and Lower Arkansas as well as those — 30 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 30 of 48 in Jefferson and Pulaski Counties (including Little Rock). 197. Contiguity is a traditional redistricting principle that requires districts to be contiguous, meaning that all parts of a district are connected to one another. See Harris v. Ariz. Indep. Redistricting Comm’n, 578 U.S. 253, 258 (2016) (citation omitted). The Court concludes that the Illustrative Plans comport with the traditional redistricting principle of contiguity. 198. Regarding population equality, the Court concludes that the plus or minus five percent deviation that Mr. Cooper applied in creating the Illustrative Plans is an appropriate measure of approximate population equality to apply for purposes of assessing the Gingles I. 6. See Fusilier v. Landry, 963 F.3d 447, 458 (5th Cir. 2020) (holding there was no clear error in the district court’s finding that plaintiffs satisfied Gingles I and noting plaintiffs’ remedial plan for judicial districts considered, inter alia, respect for the principle of one-person, one-vote). Therefore, the Illustrative Plans comply with the traditional redistricting principle of population equality. 199. The Court concludes that the Illustrative Plans comport with the traditional redistricting principle of geographical compactness and, therefore, that Plaintiffs have established the first Gingles precondition. 200. Defendants argue that even if Plaintiffs’ Illustrative Plans are reasonably compact and comply with all traditional redistricting criteria, they nevertheless cannot satisfy Gingles I because they are racially gerrymandered. The Court disagrees. 6 Even though this Court is not constitutionally required to consider the principle of one-person, one-vote in the judicial redistricting context, see Chisom, 501 U.S. at 402–03 (population equality not required for judicial elections), the Supreme Court’s decisions for legislative districts “have established, as a general matter, that an apportionment plan with a maximum population deviation under 10%” is consistent with the principle of one-person, one-vote, Brown v. Thomson, 462 U.S. 835, 842 (1983). — 31 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 31 of 48 201. This distinction between the Gingles I inquiry and equal protection “racial gerrymandering” claims is well established. The Supreme Court has emphasized that the Gingles I inquiry “embraces different considerations” than equal protection racial gerrymandering case law. LULAC, 548 U.S. at 433–34. In a recent decision, the Supreme Court “made clear that there is a difference between being aware of racial considerations and being motivated by them, 7” holding that “[t]he former is permissible; the latter is usually not.” Allen v. Milligan, 143 S. Ct. 1487, 1510–11 (2023) (quoting Miller, 515 U.S. at 916) (cleaned up). The Court further maintained its position that “Section 2 itself ‘demands consideration of race.’” Id. (quoting Abbott v. Perez, 581 U.S. ___, 138 S. Ct. 2305, 2315 (2018)). 202. The Eighth Circuit has expressly foreclosed Defendants’ position that racial gerrymandering doctrine should apply to Plaintiffs’ Section 2 claim. See Harvell v. Blytheville Sch. Dist. No. 5, 71 F.3d 1382, 1391 (8th Cir. 1995) (“Miller analyzed the equal protection problems involved in drawing voting districts along race-based lines, but did not purport to alter our inquiry into the vote-dilution claim” under Section 2). Other courts of appeals have reached the same conclusion and explained at greater length why racial gerrymandering concepts are inapposite to the Gingles framework. See Clark, 88 F.3d at 1406– 07 (holding that Miller and its progeny do not change the Gingles I inquiry); Davis v. Chiles, 139 F.3d 1414, 1425 (11th Cir. 1998) (holding that the “Miller and Gingles lines [of cases] address very different contexts”). D. Gingles II and III 203. The second and third Gingles preconditions are often analyzed together. 204. Gingles II requires plaintiffs to establish that minority is politically cohesively; this 7 While the Court did clearly state the distinction, it also acknowledged that “line between racial predominance and racial consciousness can be difficult to discern.” Id. at 1510-111. — 32 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 32 of 48 “typically requires a statistical and non-statistical evaluation of the relevant elections.” Bone Shirt, 461 F.3d at 1020. 205. Gingles III requires proof that the white majority typically votes as a bloc to defeat minority voters’ preferred candidate. To establish this precondition, the Court must determine who the minority-preferred candidate is, if the white majority voted as a bloc and to defeat that candidate, and whether there were special circumstances involved in each particular election. Ibid. 206. The parties agree that “interracial elections are the best indicators of whether the white majority usually defeats the minority candidate.” Id. at 1020-1021. The Eighth Circuit has also held that “[t]he more recent an election, the higher its probative value.” Id. at 1021. 207. The Eighth Circuit has “specifically rejected the presumption . . . that only African-American candidates can be preferred by African-American voters.” Clay v. Bd. of Ed. of City of St. Louis, 90 F.3d 1357, 1361 (8th Cir. 1996) (citing Harvell, 71 F.3d at 1386). It would be inconsistent with Gingles to apply a rule limiting this Court’s consideration to only elections involving minority candidates. 208. As the Fourth Circuit has put it, “[b]y focusing exclusively on elections in which a minority candidate appeared on the ballot, it can only be determined whether the ‘minority candidate’ is usually defeated, not, as required by Section 2, whether the ‘minority-preferred candidate’ is usually defeated.” Lewis v. Alamance Cnty., 99 F.3d 600, 608-09 (4th Cir. 1996). 209. The assumption that Black voters have little interest in white candidates and, indeed, care about candidate race to the exclusion of all other characteristics is nothing more than a racial stereotype, and “[s]uch stereotyping runs afoul of the principles embodied in the Equal Protection Clause.” Harvell, 71 F.3d at 1386. Rather, “[t]he preferences of the minority voters — 33 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 33 of 48 must be established on an election-specific basis, viewing all the relevant circumstances.” Id. 210. Courts should “decline to adopt an approach precluding the possibility that a white candidate can be the actual and legitimate choice of minority voters.” NAACP, Inc. v. City of Niagara Falls, 65 F.3d 1002, 1015-16 (2d Cir. 1995). “Such an approach”— advocated by Plaintiffs here— “would project a bleak, if not hopeless, view of our society” that is “inconsistent with our people’s aspirations for a multiracial and integrated constitutional democracy.” Id. at 1016. 211. If courts cannot presume that minority voters never prefer white candidates, but instead must presume that in each election minority voters state a preference and that their preferred candidate is simply whomever receives the highest share of minority votes, it logically follows that courts must review all elections, whether minority candidates run or not. 212. The four biracial judicial elections considered by Plaintiffs’ expert Dr. Liu are insufficient as a matter of law for the Court to make a determination as to the second and third Gingles preconditions. The Court cannot determine whether the minority’s preferred candidate is typically defeated if it ignores—as Plaintiffs ask this Court to do—almost 90% of the relevant elections. 213. Whether Black voters invariably prefer Black candidates over white candidates additionally has no bearing on the legal analysis, because Section 2 simply affords minorities a right to equal opportunity to elect “representatives of their choice,” not minority representatives of their choice. 52 U.S.C. 10301(b). If Section 2 conferred on minorities the right to elect their ideal candidates, it would not grant minority voters merely equal opportunity, but a right that no one in the political system enjoys. See De Grandy, 512 U.S. at 1020 (holding that “minority voters are not immune from the obligation to pull, haul, and trade to find common political — 34 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 34 of 48 ground”). 214. In Clay, the Eighth Circuit rejected the plaintiffs’ implication that “minority-preferred candidate” should be defined based solely on the candidate’s race, clarifying that the definition instead is the candidate receiving the highest number of minority votes. Clay, 90 F.3d at 1361-62. “Absent a showing that minority preferred candidates are, for some reason, excluded from the ballot,” the Eighth Circuit directed courts to presume that all elections feature some minority-preferred candidate and that that candidate is the one who receives the highest number of minority votes. Id. at 1362. 215. Elsewhere, the Eighth Circuit has rejected the argument that a series of candidates who had received minority support were not minority-preferred candidates, stating that it “cannot accept” the proposition that “the black voters of Blytheville have gone five consecutive years without stating a preference.” Harvell, 71 F.3d at 1387. A minority-preferred candidate “should generally be one able to receive [minority] votes, not some idealized figure whose absence from the ballot keeps a disappointed electorate at home.” Id. 216. In a recent Section 2 case, the Eighth Circuit reviewed “all 12 of the contested elections between 2000 and 2015,” irrespective of candidate race. Mo. State Conf. of the NAACP v. Ferguson-Florissant Sch. Dist., 894 F.3d 924, 935 (8th Cir. 2018). 217. The Eighth Circuit is clear: Courts must consider all the relevant elections and not limit themselves to elections involving minority candidates. Cottier, 604 F.3d 553.8 8 “The number of interracial elections presented to the district court was very small. There were only four head-to-head interracial countywide races; one race was non-polarized, and the non-Indian-preferred candidate won the other three in the city. There were only three statewide interracial head-to-head races, and the non-Indian-preferred candidate won those in the city. But there were twenty-five state and federal races with white-only candidates, and the Indian-preferred candidate won fifteen of those contests in the city. There were three white-only countywide races; one contest was non-polarized, and the Indian- preferred candidate was victorious in one of the other two in the city. These results taken as a whole show almost equal numbers of victories for Indian-preferred candidates and non-Indian-preferred candidates. — 35 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 35 of 48 218. Unless a minority votes cohesively, it cannot “establish that the minority has the potential to elect a representative of its own choice in some single-member district”—at least, not without substantial support for its preferred candidates from white voters. Growe, 507 U.S. 40. 219. A minority whose preferences are not cohesive enough to elect candidates without substantial white support has no Section 2 claim. See Bartlett, 556 U.S. at 14-15). 220. The Eighth Circuit and other courts have recognized that 60% or more of minority voters must support the same candidates to establish sufficient cohesion, although they treat this as a guideline rather than “an absolute threshold” for finding cohesion. Cottier v. City of Martin, 445 F.3d 1113 (8th Cir. 2006) (quotation marks omitted), overruled on other grounds, Cottier III, 604 F.3d 553; see also Bone Shirt v. Hazeltine, 336 F. Supp. 2d 976, 999 (D.S.D. 2004) (holding that “cohesion exists at levels above 60 percent and may exist, albeit more weakly, at lower levels”); Marylanders for Fair Representation, Inc. v. Schaefer, 849 F. Supp. 1022, 1056 (D. Md. 1994) (three-judge court) (treating 60% as a presumptive minimum). 221. Additionally, the Supreme Court’s rationale in Bartlett leads to the conclusion that 60% is likely not a demanding enough guideline but that courts should instead require Section 2 plaintiffs to show even higher levels of cohesion. See 556 U.S. at 14-15, 26 (plurality op.); id. at 26 (Thomas, J., concurring in the judgment). 222. The lack of political cohesion among Black voters in Arkansas appellate court elections establishes that they cannot elect their preferred candidates without significant support from white voters. 223. Section 2 does not guarantee the right for minority voters to combine with white They do not compel a finding that a white majority in Martin votes sufficiently as a bloc usually to defeat the Indian-preferred candidate.” Id. at 560 (emphasis added). — 36 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 36 of 48 voters to elect a minority-preferred candidate. Bartlett, 556 U.S. at 15 (plurality op.) (“Nothing in § 2 grants special protection to a minority group’s right to form political coalitions.”). 224. Section 2 is only violated when an electoral scheme deprives minorities of the ability to elect their preferred candidates on their own. Id. at 20 225. Additionally, Bartlett was “skeptical” that the “majority-bloc-voting requirement” could be satisfied when 20% of white voters supported minority-preferred candidates. Id. at 16. 226. In Arkansas’s appellate-court elections, for minority-preferred candidates to prevail, white voters would have to support minority-preferred candidates at rates doubling Bartlett’s 20% figure. 227. Therefore, Plaintiffs have failed to prove that Black voters’ preferences in Arkansas’s appellate-judge elections are cohesive enough to satisfy the second Gingles precondition. 228. White Arkansas voters do not usually vote as bloc to defeat Black voters’ preferred candidates for the Arkansas Supreme Court and Court of Appeals; instead, Black voters’ preferred candidates usually won these elections: 229. Since 2014, Black voters’ preferred candidates have been successful in 71% of contested elections for the Arkansas Court of Appeals and in 83% of contested elections for the Arkansas Supreme Court. 230. Even limiting the analysis to two-way contests since 2014, Black voters’ preferred candidates have won 67% of elections to the Arkansas Court of Appeals and 80% to the Arkansas Supreme Court. 231. Considering all contested elections over the last 20 years, Black voters’ preferred candidates have been successful in 53% of the elections for the Arkansas Court of Appeals (46% — 37 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 37 of 48 of two-way races), and 67% of the elections to the Arkansas Supreme Court (58% of two-way races). 232. The Eighth Circuit was presented with a stronger fact pattern of polarized voting in Cottier III. There, the minority-preferred candidates in over half-a-dozen biracial elections were almost invariably defeated, while minority-preferred candidates (including those who were white) had an almost a 50% success rate for overall. The Court, sitting en banc, held that the plaintiffs had not proved racially polarized voting. See 604 F.3d at 560. 233. Considering every contested election over the last two decades, minority-preferred candidates succeed at rates comparable to the rates in Cottier III. 234. Plaintiffs’ evidence of three judicial elections all involving the same failed candidate is insufficient proof of racially polarized voting. 235. The evidence that Black voters supported Judge Griffen with a fair degree of cohesion in three elections over a decade ago (2004, 2006, and 2008) is not strong evidence that Black voters are politically cohesive in all elections for appellate judgeships. 236. Special circumstances explain the three failed elections involving Judge Wendell Griffen. See Mo. State Conf. of the NAACP, 894 F.3d at 936 (“The third Gingles precondition . . . requires that the district court analyze the ‘special circumstances’ that attend elections to make sure that there are no non-racial factors at play that would appear to either defeat or demonstrate a section 2 violation.” (emphasis added) (quoting Gingles, 478 U.S. at 51, 54, 57)). 237. Judge Griffen’s history of investigations by the Judicial Discipline and Disability Commission and his views about controversial topics like abortion and the death penalty are special circumstances that explain his defeats. 238. Because of the lack of evidence of racially polarized voting here, that minority- — 38 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 38 of 48 preferred candidates succeed far more often than they lose, Plaintiffs have failed to satisfy the third Gingles precondition. E. Totality of the Circumstances 239. Notwithstanding Plaintiffs’ failure to prove two of the three Gingles preconditions—each of which is necessary for them to succeed—they have additionally failed to make the necessary “show[ing] that the ‘totality of the circumstances’ demonstrates a section 2 violation.” Mo. State Conf. of the NAACP, 894 F.3d at 937-38. 240. Plaintiffs have failed to “prove that the totality of the circumstances indicates minority voters ha[ve] less opportunity than other members of the electorate to participate in the political process and elect representatives of their choice.” Bone Shirt, 461 F.3d at 1021 (internal quotations omitted). 241. This totality-of-the-circumstances analysis requires “a searching practical evaluation of the past and present reality”—typically leading courts to review a number of nonexhaustive factors derived from the Senate Report accompanying the passage of the 1982 amendments to Section 2. Gingles, 478 U.S. at 65-66; see id. at 44-45; Mo. State Conf. of the NAACP, 894 F.3d at 931. 242. The Eighth Circuit has long held that two of those factors predominate the analysis: the extent to which voting in the jurisdiction is racially polarized (Factor Two) and the extent to which members of the minority group have been elected to office in the jurisdiction (Factor 7). See Mo. State Conf. of the NAACP, 894 F.3d at 938 and Bone Shirt, 461 F.3d at 1022. Absent a strong showing on these factors, Plaintiffs cannot succeed. See Gingles, 478 U.S. at 48-49 n.15 (“recognizing the primacy” of these factors and deeming them “essential to” a Section 2 claim). — 39 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 39 of 48 243. As already discussed, voting in Arkansas’s appellate judicial elections, which are nonpartisan, is not racially polarized. Plaintiffs have not established Factor 2. 244. Additionally, without evidence that white voters are unreceptive to Black candidates, Plaintiffs have also failed to establish Factor 7. 245. Of the five Black candidates for the Court of Appeals the parties have identified— Judge Waymond Brown, Judge Olly Neal, Judge Andree Roaf, Judge Griffen, and Judge Eugene Hunt—four were elected to the Court of Appeals. 246. Judge Hunt did not lose to a white candidate, but to Judge Brown. 247. In early 1999, after being appointed to unnumbered seats by the Governor, Judges Neal, Roaf, and Griffen were assigned to districts from which challengers could oppose them at the November 2000 general election. 248. Although their districts were overwhelmingly white, each of them was re-elected un-opposed. 249. Thus, Judges Neal and Roaf did not even draw opponents in majority-white districts; and Judge Griffen too was reelected unopposed and only lost his second bid for reelection after years of conflict with the judicial discipline commission. 250. With the exception of Judge Griffen’s Supreme Court campaigns and of Evelyn Moorehead’s campaign for the Supreme Court in 2010 (in which she lacked the support of even Black voters), Black candidates have been elected with ease. 251. Little Rock Mayor Frank Scott Jr., who is Black, defeated white candidates in his last two elections, 2018 and 2022. Pulaski County Sheriff Eric Higgins, who is Black, defeated white candidates in his last two elections, 2018 and 2022. Pulaski County Circuit Clerk Theresa Hollingsworth, who is Black, defeated white candidates in her first election in 1018 and ran — 40 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 40 of 48 unopposed in 2022. While these three elections do not involve judicial races, they are illustrative of recent Black candidate success in central Arkansas elections. 252. While the relatively small number of those candidates may not preclude Section 2 relief, it does not support it, when those candidates who have run have generally prevailed—and when Griffen did not it was because he was an unorthodox candidate. 253. Another Senate factor, sometimes numbered as the ninth factor, asks “whether the policy underlying the state or political subdivision’s use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.” Gingles, 478 U.S. at 37. 254. Regarding Plaintiffs’ request that this Court create districts for the Arkansas Supreme Court, every court of appeals and district court to consider the question has declined to fashion districts for courts that States choose to elect at-large. See Cousin v. Sundquist, 145 F.3d 818, 827-28 (6th Cir. 1998); Davis v. Chiles, 139 F.3d 1414, 1421-22 (11th Cir. 1998); Milwaukee Branch of the NAACP v. Thompson, 116 F.3d 1194, 1200-01 (7th Cir. 1997); So. Christian Leadership Conference of Ala. v. Sessions, 56 F.3d 1281, 1296-97 (11th Cir. 1995); Nipper v. Smith, 39 F.3d 1494, 1543-45 (11th Cir. 1994) (en banc) (plurality opinion); id. at 1547 (Edmondson, J., concurring in the opinion in part and concurring in the result); LULAC v. Clements, 999 F.2d 831, 868-74 (5th Cir. 1993); Lopez v. Abbott, 339 F. Supp. 3d 589, 617-19 (S.D. Tex. 2018); Alabama State Conf. of Nat'l Ass'n for Advancement of Colored People v. Alabama, 612 F. Supp. 3d 1232 (M.D. Ala. 2020). 255. That is because, as those courts have consistently explained, states indisputably have a “powerful, indeed dispositive” interest in “matching the boundaries of a court’s jurisdiction to the boundaries of the judges’ electoral base”—an interest often called linkage. Thompson, 116 F.3d at 1200. — 41 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 41 of 48 256. In other words, states have an interest in linking the territory from which judges are elected to the territory over which they have jurisdiction, and at-large electoral systems, like Arkansas’s system for electing supreme court justices, provide for such linkage. 257. Courts find that linking judges’ jurisdiction to the voters who elect them simultaneously promotes the goals of accountability and judicial independence. Linkage makes judges accountable to everyone over whom they have jurisdiction. See, e.g., Cousin, 145 F.3d at 827 (“[L]inkage ensures that . . . the entire electorate which will be subject to [a] judge’s jurisdiction has the opportunity to hold him or her accountable at the polls.”). 258. And while promoting accountability, linkage also promotes judicial independence. “Larger jurisdictions . . . free the judge to follow the law dispassionately” by “diluting the reaction to individual decisions.” Thompson, 116 F.3d at 1201. 259. Electing judges from smaller districts, by contrast, “increase[s] the potential for ‘home cooking’” in cases that particularly affect a judge’s local constituency. Nipper, 39 F.3d at 1544. 260. The particular remedy that Plaintiffs propose that this Court devise—a supreme-court district designed to elect a justice of a particular race—would uniquely harm the accountability and judicial independence that linkage serves, because the “announced purpose” of this system “would be to assist a predominantly black section of the [State] in electing black judges.” Id. 261. The Supreme Court has described the effect that systems like the one Plaintiffs propose would have on elected officials: “When a district obviously is created solely to effectuate the perceived common interests of one racial group, elected officials are more likely to believe that their primary obligation is to represent only the members of that group, rather than — 42 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 42 of 48 their constituency as a whole.” Shaw v. Reno, 509 U.S. 630, 648 (1993). 262. With judicial elections in particular, such a system would “send[] the message . . . that race matters in the administration of justice.” So. Christian Leadership Conf., 56 F.3d at 1297. 263. Additionally, Arkansas has the right to choose linkage over districting because it has the right to define what sort of office its supreme court justices hold. See Cousin, 145 F.3d at 827 (“The decision to make jurisdiction and electoral bases coterminous is more than a decision about how to elect state judges. It is a decision of what constitutes a state court judge.” (quoting Clements, 999 F.2d at 872)). 264. District courts in Texas and Alabama have rejected challenges to their states’ at-large systems for electing state supreme courts on the basis of linkage. See Lopez 339 F. Supp. 3d 589; Alabama 612 F. Supp. 3d 1232. 265. A district court’s rejection of a challenge to Wisconsin’s at-large method of electing its intermediate appellate court on the basis of linkage was affirmed by the Seventh Circuit. See Thompson, 116 F.3d 1194 (7th Cir. 1997). 266. Arkansas “has a weighty interest in preserving all of its citizens’ right to have a say in who the state’s most powerful judges are.” Alabama, 612 F. Supp. at 1259. 267. The State’s linkage argument in favor of the Arkansas Supreme Court redistricting seems to undercut the State’s argument regarding the Arkansas Court of Appeals but a Circuit Court has ruled otherwise. It does not affect Arkansas’s linkage interest in the Arkansas Supreme Court that it does not require linkage for all courts in the State. See, e.g., Thompson, 116 F.3d at 1200-01 (holding that even though “Wisconsin already ha[d] weakened the link between jurisdiction and electoral base” for some courts, Section 2 still did “not compel [it] to — 43 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 43 of 48 disregard a belief that larger jurisdictions promote impartial administration of justice”). 268. As an alternative remedy to creating districts for the Arkansas Supreme Court, Plaintiffs briefly suggest a system of cumulative voting. 269. This Court lacks the power to order that radical change in how Arkansas conducts its judicial elections. 270. No court has ever ordered cumulative voting in judicial elections under Section 2; indeed, one of the few courts to ever have ordered cumulative voting as a remedy in a Section 2 case about nonjudicial-elections did so only because it was the defendant’s preferred remedy. See United States v. Vill. of Port Chester, 704 F. Supp. 2d 411, 447-48, 451-52 (S.D.N.Y. 2010) (acknowledging that cumulative voting was so unknown to American voters that the court had to order defendant to adopt an educational program to teach voters how to cumulatively vote, but nevertheless ordering cumulative voting out of “deference” to the defendant jurisdiction’s preference). 271. Beside the remedy’s novelty, Arkansas has powerful interests in not using cumulative voting, because its “practical effect” would be that “judicial colleagues who previously ran for designated positions” would have to “run against each other,” which would in turn “undermin[e] the treasured institution of judicial collegiality.” Cousin, 145 F.3d at 830. 272. In rejecting cumulative-voting remedies for judicial elections, courts have also found that because a cumulative-voting system would require judges seeking reelection to “compete against every [other] judge up for reelection,” cumulative voting would “adversely affect the independence of the judiciary” and “dampen lawyer interest in a judicial career” by making reelection much tougher. Nipper, 39 F.3d at 1546. 273. A system where judges seek reelection in a single seat against a single opponent, like Arkansas’s current system, gives a strong advantage to incumbents, strengthening judicial independence, making judicial offices more attractive, and most importantly, ensuring some — 44 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 44 of 48 measure of institutional continuity. See id. 274. Plaintiffs’ proposed cumulative-voting system, by contrast, would open up the possibility that the entire Arkansas Supreme Court could turn over every eight years, leaving Arkansans with no continuity from election to election. 275. For these reasons, Defendants have a powerful interest in adhering to the State’s current system for electing the Arkansas Supreme Court and not adopting Plaintiffs’ proposed remedies. 276. Even if the Court had found for the Plaintiff’s on all three Gingles factors, Defendants would still be entitled to judgment on both of the two essential Senate factors just discussed. III. CONCLUSION Based on these findings of fact and conclusions of law, judgment should be entered for Defendants. IT IS SO ORDERED this 25th this day of July, 2023. _________________________________ UNITED STATES DISTRICT JUDGE Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 45 of 48 EXHIBIT 1 — 46 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 46 of 48 EXHIBIT 2 — 47 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 47 of 48 EXHIBIT 3 — 48 — Case 4:19-cv-00402-JM Document 193 Filed 07/25/23 Page 48 of 48