Ala. State Conference of N.A. for Advancement of Colored People v. Merrill

Full title: ALABAMA STATE CONFERENCE OF THE NATIONAL ASSOCIATION FOR THE ADVANCEMENT…

Court: UNITED STATES DISTRICT COURT FOR THE MIDDLE DISTRICT OF ALABAMA NORTHERN DIVISION

Date published: Feb 5, 2020

CASE NO. 2:16-CV-731-WKW [WO] (M.D. Ala. Feb. 5, 2020)

Opinion

CASE NO. 2:16-CV-731-WKW [WO]

02-05-2020

ALABAMA STATE CONFERENCE OF THE NATIONAL ASSOCIATION FOR THE ADVANCEMENT OF COLORED PEOPLE, SHERMAN NORFLEET, CLARENCE MUHAMMAD, CURTIS TRAVIS, and JOHN HARRIS, Plaintiffs, v. STATE OF ALABAMA and JOHN H. MERRILL, in his official capacity as Alabama Secretary of State, Defendants.

W. Keith Watkins UNITED STATES DISTRICT JUDGE

MEMORANDUM OPINION AND ORDER

TABLE OF CONTENTS

I. INTRODUCTION — 4 II. JURISDICTION AND VENUE — 10 III. BACKGROUND — 11 IV. STANDARD OF REVIEW FOR BENCH TRIALS — 19 V. DISCUSSION — 20

A. Section 2 Vote Dilution — 20

1. Section 2: The Statute — 20

2. Burden of Proof — 21
3. The Meaning of Section 2 — 23

a. Legislative History — 23
b. Gingles Preconditions and Totality-of-Circumstances Test — 25

4. Section 2 and At-Large Judicial Elections — 29

a. Nipper v. Smith and Later Eleventh Circuit Caselaw Developments — 35

i. The Importance of a State's Interests — 36
ii. Nipper's Applicability to Appellate Judicial Elections — 40
iii. The Role of Causation in the § 2 Vote Dilution Analysis — 41

b. Summary — 44

5. Analysis of the Gingles Preconditions — 45

a. Introduction — 45
b. The First Gingles Precondition — 46

i. The Inextricably Intertwined Nature of Liability and Remedy in the Eleventh Circuit — 47
ii. Factors Governing the First Gingles Precondition — 48
iii. Plaintiffs' Illustrative Plans for Alabama's Appellate Courts — 54
iv. Whether the Illustrative Plans for the Alabama Supreme Court Satisfy the First Gingles Precondition — 55
v. Whether the Illustrative Plans for Alabama's Intermediate Courts Satisfy the First Gingles Precondition — 57

c. The Second and Third Gingles Preconditions — 68

i. Proof of the Second and Third Gingles Preconditions: Generally — 71
ii. Plaintiffs' Statistical Evidence of Racially Polarized Voting — 75
iii. Findings of Fact and Conclusions of Law: Racially Polarized Voting — 88

6. The Totality of Circumstances — 99

a. Senate Factors 1 and 5: Alabama's History of Discrimination and Its Lingering Effects — 102

i. Senate Factor 1: History of Discrimination in Voting — 102
ii. Senate Factor 5: Discrimination in Education, Employment, and Health — 106
iii. Evaluation and Weighing of Senate Factors 1 and 5 — 109

b. Senate Factor 2: The Extent of Racially Polarized Voting (Causation) — 111

i. Party Alignment in Alabama — 112
ii. The Relative Strength of the Two Parties — 116
iii. Partisan Judicial Elections — 121
iv. Straight-Ticket Voting — 123
v. The Competitiveness of Black-Preferred Candidates in Appellate Judicial Races — 125
vi. Whether Party Is a Proxy for Race — 127
vii. Weighing of Senate Factor 2 — 145

c. Senate Factor 3: Election Devices that Enhance Potential for Discrimination — 146
d. Senate Factor 4: Slating — 150
e. Senate Factor 6: Racial Appeals — 152

f. Senate Factor 7: Extent of Minority Success — 158
g. Senate Factor 8: Responsiveness to Minority Needs — 167
h. Senate Factor 9: Whether the State's Policy Is Tenuous — 168
i. Conclusion: Totality of Circumstances — 168

7. Remedy: Balancing the State's Interests Against Plaintiffs' Interests in Remedial Districting - 171

a. Interference with Alabama's Constitutional Model — 171
b. Linkage Between Appellate Judges' Jurisdictions and Electoral Base — 176

i. Impartial Administration of Justice in Alabama's Appellate Courts — 177
ii. Judicial Accountability and Independence — 179
iii. A Matter of Law, Not Fact — 190

c. The Pool of Candidates — 192
d. Weighing of the Interests — 194

B. Constitutional Claims — 197 VI. CONCLUSION — 206

I. INTRODUCTION

In 1868, Alabama adopted an at-large, statewide method of electing its Supreme Court. That method has been in place for 152 years. "At-large" means that justices of the Alabama Supreme Court run statewide, not in districts. There presently are nine seats on the Alabama Supreme Court. Justices are elected in partisan, staggered elections on a rotating basis of one, three, and five justices, every two years (to fill six-year terms). Eight of the justices run in numbered places to avoid having to run against each other, with the ninth being the position of Chief Justice.

In 1911, the Alabama Legislature created an intermediate Court of Appeals, and in 1969 it divided the Court of Appeals into the Court of Criminal Appeals and the Court of Civil Appeals. Since 1911, a partisan at-large, statewide election structure has been in place for Alabama's intermediate appellate courts. There presently are five seats on each court of appeals, and the judges are elected to numbered places for six-year, staggered terms. Judges of the Court of Criminal Appeals elect the presiding judge, while the most senior judge on the Court of Civil Appeals serves as the presiding judge. Mid-term vacancies in the appellate courts are filled by gubernatorial appointment.

Plaintiffs — the Alabama State Conference of the National Association for the Advancement of Colored People and registered Alabama voters (Sherman Norfleet, Clarence Muhammad, Curtis Travis, and John Harris) — seek an injunction against the enforcement of the State's at-large method of electing its appellate judges on grounds that the method contravenes § 2 of the Voting Rights Act of 1965 ("VRA"), as amended, and the Fourteenth and Fifteenth Amendments' prohibition against intentional racial discrimination. See U.S Const. amend. XIV, § 1, XV; Voting Rights Act of 1965, § 2, 52 U.S.C. § 10301 (2018). Plaintiffs also seek a federal court order directing the State to create a new method of election for Alabama's appellate courts that complies with § 2 of the VRA and the Constitution; their proposed remedy is single-member districts for partisan elections of appellate judges. Defendants are the State of Alabama and John H. Merrill, in his official capacity as Alabama Secretary of State.¹

1 On February 3, 2020, the Eleventh Circuit Court of Appeals affirmed on interlocutory appeal this court's order denying the State of Alabama's motion to dismiss on grounds of sovereign immunity. (Doc. # 179.) In this opinion, for simplicity, Defendants are referred to collectively as "the State."

Plaintiffs allege that the at-large method "dilutes" the votes of African-American Alabamians "on account of race." The phrase "on account of race" is taken directly from § 2 of the VRA. The political history of Alabama confirms that the 1965 VRA was passed with the impetus of the famous Selma-to-Montgomery March in the spring of that year, led by Dr. Martin Luther King, and managed by Judge Frank M. Johnson, Jr. of this court (for whom this court complex is named). And the momentum behind the 1982 VRA amendment arose out of an Alabama federal court case. See Bolden v. City of Mobile, 423 F. Supp. 384 (S.D. Ala. 1976), aff'd, 571 F.2d 238 (5th Cir. 1978), rev'd, 446 U.S. 55 (1980).

In fact, the political history of Alabama intersects with the modern Civil Rights Movement at almost every major turn. But Alabama history did not begin and end with the Civil Rights Movement. Since the early 1900s, that history has intersected in interesting and unusual ways with the federal court system and with Alabama's own state courts. Courts at both the state and federal level have dominated the civic life of Alabamians, often as a refuge of justice and fairness and sometimes to the point of exasperation (both of the courts and Alabama politicians). More of that later.

"Dilution" of a vote "on account of race" does not intrinsically help the reader understand the concept. Vote dilution, in the § 2 "sense, . . . refers to the impermissible discriminatory effect that a multimember or other districting plan has when it operates 'to cancel out or minimize the voting strength of racial groups.'" Thornburg v. Gingles, 478 U.S. 30, 87 (1986) (O'Connor, J., concurring in the judgment) (quoting White v. Regester, 412 U.S. 755, 765 (1973)). In its simplistic form, if a vote is diluted, it has less strength — "weight" in judge-speak — than other votes.²

2 For example, if all African Americans in Alabama were crammed into one state senate district out of thirty-five districts, the votes of those African Americans would be watered down on a statewide basis — worth less, or less effective — than those of other races. Thus, African Americans would have less political influence. In fact, there are eight minority senate districts in Alabama. The composition of the state legislature today is roughly proportional to the percentage of African-American citizens in Alabama.

Vote dilution matters politically when Alabama is approximately 69% white and 26% black and when candidates supported by a majority of white voters have won all of the races for appellate court positions lately. But analyzing the political losses of black-preferred candidates under § 2 is anything but simple, especially when causation "on account of race" is the measure by which an illegal condition is said to exist in state politics.³ Dilution can occur on account of intentional racial discrimination by the State, but intentional discrimination is not required for Plaintiffs to prevail under § 2. On the other hand, mere correlation of race and voting is not enough; Plaintiffs must prove that the electoral losses were on account of race, that is, caused by racial considerations. In § 2 jargon, Plaintiffs must establish that the at-large, statewide system of elections of Alabama's appellate judges "results in" the denial of their right to vote "on account of race," which is proven if the "totality of circumstances" shows that the State's method of electing its appellate judges is not "equally open" to African Americans. § 10301. This is the sweet spot of Plaintiffs' § 2 burden here. Absent proof of race as a cause of electoral losses, such losses "may plausibly be attributed to the inescapable fact that, in a majoritarian system, numerical minorities lose elections," irrespective of the race of the minority. Holder v. Hall, 512 U.S. 874, 901 (1994) (Thomas, J., concurring in the judgment).

3 Courts interpreting § 2 vote dilution claims use the phrase "black-preferred candidate." The black-preferred candidate is the candidate — of any race — whom the African-American electorate prefers. See Gingles, 478 U.S. at 68 (plurality opinion); Askew v. City of Rome, 127 F.3d 1355, 1377 n.8 (11th Cir. 1997); see also Trial Tr. IV, at 156 (Jerome Gray) (testifying that the black-preferred candidate does not have to be black). Preferred candidates of the African-American electorate (or, as phrased in § 2, "representatives of their choice," § 10301(b)) was a hot-button phrase at trial. Plaintiffs objected to the idea that white Democratic candidates would always be viewed as the preferred candidates of African-American voters when, in fact, they had no other choice.

The State says that the political losses for appellate judgeships can be explained by other factors generally under the heading of partisanship: that the losses of candidates favored by African-American voters are the result of several partisan divides that have nothing to do with race. Additionally, the State argues that Alabama's interests outweigh Plaintiffs' interests in adopting a remedial districting scheme for the election of appellate judges. First, Alabama's interest in maintaining its chosen format for appellate judicial elections established under its state Constitution is primary under our federal system of shared sovereignty, and a federal court has no business upsetting the State's choice of 152-year-old and 109-year-old at-large voting systems for statewide offices. Second, Plaintiffs' proposed districting remedy contravenes Alabama's important interest of maintaining "linkage" between a judge's territorial jurisdiction and its electoral base to promote judicial independence and accountability and ensures that state appellate judges are held accountable at the ballot box by the entire electorate over which they preside.

In tandem with their § 2 vote dilution claim, Plaintiffs also bring constitutional claims against Alabama's at-large, statewide system for electing appellate judges. Unlike under § 2, proof of intentional discrimination is required for Plaintiffs to prevail on their Fourteenth and Fifteenth Amendment claims.

The case was tried without a jury over the course of six days in November 2018. Twenty-one witnesses testified, and hundreds of pages have been generated in the trial transcript. More than 200 exhibits were admitted into evidence, and the parties have submitted extensive post-trial briefing and stipulations. As befits a challenge to a state's chosen form of electing a branch of its government, the court has reviewed and considered all the materials in reaching a reasoned decision. Adhering to the requirements of Federal Rule of Civil Procedure 52(a), this opinion represents the court's best effort to sort and analyze the case in plain language by applying the § 2 framework established in Gingles for evaluating vote dilution claims, and by considering whether there is evidence of intentional racial discrimination.

For the reasons that follow, Plaintiffs have not proven a § 2 vote dilution claim. And they have not carried their burden to show that intentional racial discrimination was a motivating factor in the adoption or maintenance of Alabama's system for electing its appellate judges; hence, their Fourteenth Amendment and Fifteenth Amendment claims also must fail. Judgment is due to be entered in favor of the State.

II. JURISDICTION AND VENUE

Subject-matter jurisdiction is exercised pursuant to 28 U.S.C. §§ 1331 and 1343. The court also has jurisdiction to grant declaratory and injunctive relief pursuant to 28 U.S.C. §§ 2201 and 2202. Plaintiffs have demonstrated that they have Article III standing to bring this action. See Post-Trial Stipulations, at 149-50, ¶¶ 3-8 (Doc. # 176); Order, at 16-18 (Doc. # 45); Rumsfeld v. Forum for Acad. & Institutional Rights, Inc., 547 U.S. 47, 52 n.2 (2006) ("[T]he presence of one party with standing is sufficient to satisfy Article III's case-or-controversy requirement."). Personal jurisdiction and venue are uncontested.

III. BACKGROUND

The raw facts are these: With the exception of two Supreme Court justices in the 1980s and 1990s, no African-American candidate has ever been elected to statewide office in Alabama. These two justices — the late Justice Oscar W. Adams, Jr. and Justice Ralph D. Cook — came to the bench initially by gubernatorial appointment in 1980 and 1993, respectively, to fill unexpired terms and, thus, had the benefit of incumbency to win their elections. Justice Adams won two statewide elections, in 1982 and 1988. Justice Cook won in 1994. Between 2000 and 2008, six African Americans ran for eight appellate court seats and lost (Justice Cook in 2000, Vicky Underwood Toles in 2002, Judge John England in 2000 and 2006, Aubrey Ford in 2000 and 2006, Gwendolyn Kennedy in 2006, and Clyde Jones in 2008). Since 2008, no African American has run for an appellate court seat in Alabama, but black-preferred white candidates have: Justice Sue Bell Cobb; Judge Deborah Bell Paseur; and Judge Robert Smith Vance, Jr. Then-Judge Sue Bell-Cobb ran for Chief Justice as a Democrat in 2006 and won. Judge Paseur ran for the Supreme Court as a Democrat in 2008, and lost by only 12,892 votes out of 2,032,395 cast. Judge Vance ran twice as a Democrat, in 2012 and 2018, and he was highly competitive in 2012 and competitive in 2018, garnering 48.11% and 42.54% of the vote, respectively. See Bonneau Expert Report, at 11, ¶ 26 (Defs.' Trial Ex. 24; Doc. # 85-1) (describing "races won by 55% of the vote or less to be highly competitive"). Additionally, in a non-judicial statewide race, Doug Jones, who was the black-preferred candidate, won a United States Senate seat as a Democrat in a 2017 special election, narrowly defeating his Republican opponent, Roy Moore.

African Americans in Alabama prefer the Democratic Party by overwhelming margins — in excess of 90% in some recent races. This is not unique to Alabama; African Americans heavily prefer the Democratic Party nationally. White Alabamians on the other hand identify with the Republican Party at a 50% rate, but vote Republican at a much higher rate, close to 75% in some recent races. So Alabama voting is racially polarized; that is, there is a significant correlation between race and voting behavior in Alabama. The question of this case is, why? Is it on account of race, as condemned by § 2 of the VRA, or on account of some other cause or causes, such as partisan politics?

Alabama's twentieth-century political history reveals two major, even historic, partisan political shifts that are relevant here. The first occurred around the time of the VRA's passage in 1965 when African-American Alabamians followed the national trend and practically overnight ditched the Republican Party ("the party of Mr. Lincoln") for the Democratic Party. The bloom was off the Republican rose; Republican Senator Barry Goldwater voted against the VRA, and the rest is history, almost instant history, with black voters ditching 100 years of Republican party allegiance. No Republican had been elected in a statewide race in Alabama since before the Civil War until war hero Jeremiah Denton won the 1980 Senate race in a close one. Until Senator Denton's election, there had been over 110 years of absolute dominance by the Democratic Party. But that was about to change, albeit at a snail's pace.

The second partisan shift took its sweet time and is much more complicated. A majority of white Alabamians were under the thrall of George Wallace at the time of the VRA passage, and Wallace strongly encouraged whites to stay unreconstructed, to fight, to obfuscate and irritate, to pontificate, to stand in the way and obstruct anything that resembled racial progress. And he did it as a Democrat. Democrats elected Wallace in 1962, and his wife Lurleen in 1966, as governor. The Legislature remained over 90% Democrat, and all statewide offices were held by Democrats. In 1970, Wallace was in the gubernatorial race of his life in the Democratic primary (the only primary that mattered for nearly 100 years) against Governor Albert Brewer, who had succeeded Governor Lurleen Wallace upon her untimely death in 1968. Wallace interjected race in the runoff and barely won election in that 1970 race.⁴

4 Shorty Price, a colorful, serial candidate for Governor, referred to his 1970 opponents as "King George" and "Prince Albert."

In 1972, Wallace survived an assassination attempt but was paralyzed and in severe pain for the rest of his life. He was reelected governor in 1974, and he quietly reverted to the populist progressive he had been in college, law school, and in the state legislature in the early 1950s, repenting of his racist past. The long and short of it is he asked African Americans to forgive him, and they did, reelecting Wallace to the governorship (with white help) in 1982.

Wallace's conversion — more likely restoration, though cynics dispute it — left Alabama African Americans and whites in a shotgun marriage for a decade or more under the banner of the segregationist Democratic Party a la Alabama. But the Republican Party began to grow and gain votes in greater numbers in presidential elections, carrying the state for Goldwater in 1964, Nixon in 1968 and 1972, and Reagan but barely in 1980. (President Jimmy Carter, a southern Democrat from neighboring Georgia, carried Alabama in 1976 and made a strong showing here in 1980.) After Jeremiah Denton broke the Republican ice in 1980, the Democrats began to fumble the political ball in state races, and, in the 1986 governor's race, they suffered the partisan equivalent of Yale beating Alabama in football. It was alleged that Republicans, who supported their very weak ticket in the Republican primary (remember, they had not won an Alabama gubernatorial race since 1872), sauntered over to the Democratic table for the Democratic primary runoff to vote for the more conservative Democrat, since the hapless Republican candidate, Guy Hunt, was entirely unelectable by any stretch. The conservative Democrat, then-Alabama Attorney General Charles A. Graddick, apparently won the primary runoff. A hot mess of litigation followed in state and federal courts, with the Democratic state committee ultimately disqualifying Graddick, the apparent winner, and installing the second-place finisher, former Alabama Attorney General and then-Lt. Governor William J. Baxley, as the Democratic general election candidate.

Alabama voters were paying attention, and Yale won. In the 1986 gubernatorial race, the unelectable Republican Guy Hunt, a little-known probate judge from northern Alabama, defeated the well-known Democratic candidate Baxley in a stunner. Hunt became the second Republican ever elected in a statewide race in post-Civil War Alabama, and the first Republican governor since David P. Lewis, a Republican reconstructionist, governed from 1872 to 1874.

The second partisan conversion had finally found a voice in the person of Governor Hunt. Wallace had faded; the Democrats could not elect their choice for Governor, Mr. Baxley. Contemporaneously, the business community, now led by a Republican governor, was motivated to engage the public in a debate over tort reform, alleging that Alabama had become "Tort Hell." Eventually, an undisclosed touch-up paint job on a new BMW sold by an Alabama dealer — for which the BMW's buyer secured a $2 million punitive damages award blessed by the Alabama Supreme Court — had a lot to do with that. See BMW of N. Am., Inc. v. Gore, 517 U.S. 559 (1996). After much legislating and litigating, a business-friendly tort reform package passed by the state legislature in the late 1980s was wiped out by the Alabama Supreme Court, which was all Democrat. Subsequently, Republicans began to compete in Supreme Court races for the first time, with tort reform the issue de jure. In the early 1990s, while the Supreme Court was still 9-0 Democrat, Republican business leaders hired Karl Rove from Texas to try to replicate in Alabama the recent business victory in Texas Supreme Court races. That effort was aimed squarely at the Alabama Supreme Court. Beginning in 1994 and concluding in 2004, Alabama voters flipped the Supreme Court from 9-0 Democrat to 9-0 Republican. A similar trend occurred in the intermediate appellate courts.

The party change started with the first Republican elected to the Alabama Supreme Court after the Civil War, Judge Perry Hooper in 1994. It was not easy, as will be detailed later. After eleven months of highly publicized and bitter litigation, Hooper was certified the winner as the Chief Justice, and the partisan shift of the Alabama Supreme Court and courts of appeals from Democrat to Republican began to bloom. In 1996, Republicans won all five appellate judicial races on the ballot and won another one in 1998.

With the exception of Democrat Sue Bell Cobb, who in 2000 was reelected to the Court of Criminal Appeals and in 2006 was elected the first female Chief Justice of the Alabama Supreme Court, no Democrat has been elected to an Alabama appellate court since 2000. In fact, in 2000, there were eleven contested appellate court races in Alabama, and Republicans won ten of them. In 2002 and 2004, Republicans won all eight combined appellate court races, and the Supreme Court was 9-0 Republican from 2004 until 2006, and again from Justice Cobb's resignation in 2011 until the present.

This partisan shift of voters to the Republican Party, though led by appellate court races, included a complete transformation of the 140-member state legislature from heavily Democrat in 1986 to a veto-proof Republican majority in both houses in 2010, continuing until today. The last Democrat Governor was elected in 1998 and was defeated by a Republican in 2002. The shift affected United States Senate races as well. Senator Richard Shelby, who had beaten Republican Jeremiah Denton in the 1986 Senate race and had easily won reelection as a Democrat in 1992, switched to the Republican Party in 1994. Democrat Howell Heflin's Senate seat turned Republican in 1996 when Jeff Sessions won by over 100,000 votes, and it remained Republican until the special election won by Democrat Doug Jones in 2017.

Finally, how did African-American appellate judicial candidates (not the broader category of black-preferred candidates) fare during the shift of voters to the Republican Party? In 1980, Governor Fob James appointed the first African-American Alabama Supreme Court justice (and consequently, the first African-American statewide office holder) to the Alabama Supreme Court: Justice Oscar Adams. In 1982, Justice Adams won the Democratic primary and primary election runoff, beating his white Democratic opponents, and was elected into office in an unopposed general election. In 1988, he defeated a white Republican in his reelection race. In 1993, Justice Adams retired, and Governor Jim Folsom, Jr. appointed Judge Ralph Cook, an African-American circuit judge from Jefferson County, to fill the unexpired term of Justice Adams. Justice Cook, running as a Democrat, won his election to a full term in 1994, defeating a white Republican candidate.

In 1998, Governor Siegelman appointed another African-American circuit judge, Judge John England of Tuscaloosa County, to fill an unexpired term on the Supreme Court.⁵ Both Justice Cook and Justice England were on the Democratic ballot in the 2000 general election for Supreme Court justice, and both lost to white Republicans: Cook with 47.3% of the total vote, and England with 45.7%. No African-American justice has served since Justice Cook's and Justice England's terms ended in mid-January 2001. No African American has ever served on one of Alabama's intermediate appellate courts.

5 After Judge England served as an associate justice, he won re-election to his circuit judgeship, a position he holds to this day. Therefore, this opinion will refer to him as Judge England unless referencing his time as an Alabama Supreme Court Justice.

IV. STANDARD OF REVIEW FOR BENCH TRIALS

"In an action tried on the facts without a jury . . . , the court must find the facts specially and state its conclusions of law separately." Fed. R. Civ. P. 52(a)(1). In the context of § 2, the Eleventh Circuit has explained:

Because the resolution of a voting dilution claim requires close analysis of unusually complex factual patterns, and because the decision of such a case has the potential for serious interference with state functions, we have strictly adhered to the rule 52(a) requirements in voting dilution cases and have required district courts to explain with particularity their reasoning and the subsidiary factual conclusions underlying their reasoning.

Johnson v. Hamrick, 196 F.3d 1216, 1223 (11th Cir. 1999) (quoting Cross v. Baxter, 604 F.2d 875, 879 (5th Cir. 1979)). Because the issue of vote dilution in § 2 cases presents intertwined questions of fact and law, to spare the reader repetition, the findings of fact and conclusions of law are not set out in separate sections. They are, though, set out with particularity. To the extent that the witness testimony, exhibits, or stipulations conflict with the facts and law as stated in this opinion, they are rejected.

In a bench trial, "it is the exclusive province of the judge . . . to assess the credibility of witnesses and to assign weight to their testimony." Childrey v. Bennett, 997 F.2d 830, 834 (11th Cir. 1993). Additionally, "[a] trial judge sitting without jury is entitled to great latitude concerning the admission or exclusion of evidence." Wright v. Southwest Bank, 554 F.2d 661, 663 (5th Cir. 1977).⁶

6 In Bonner v. City of Prichard, the Eleventh Circuit adopted as binding precedent all decisions of the former Fifth Circuit issued prior to October 1, 1981. See 661 F.2d 1206, 1209 (11th Cir. 1981) (en banc).

V. DISCUSSION

The discussion is separated into two parts. Part V.A. addresses Plaintiffs' § 2 claim of vote dilution. Part V.B. addresses Plaintiffs' constitutional claims based on intentional racial discrimination. A. Section 2 Vote Dilution

1. Section 2: The Statute

It is appropriate to begin the § 2 discussion with the language of the statute, which is set out here in full:

§ 10301. Denial or abridgement of right to vote

on account of race or color through voting qualifications or

prerequisites; establishment of violation

(a) No voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color, or in contravention of the guarantees set forth in section 10303(f)(2) of this title, as provided in subsection (b).

(b) A violation of subsection (a) is established if, based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of a class of citizens protected by subsection (a) in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, That nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.

§ 10301. This statute is commonly known as § 2.

2. Burden of Proof

To prevail on a § 2 claim, a plaintiff has a two-step burden. First, he or she must prove each of the three preconditions in Gingles, 478 U.S. at 30, by a preponderance of the evidence. See City of Carrollton Branch of NAACP v. Stallings, 829 F.2d 1547, 1558 (11th Cir. 1987). These preconditions "present mixed questions of law and fact." Solomon v. Liberty Cty., 899 F.2d 1012, 1017 n.6 (11th Cir. 1990) (en banc) (Kravitch, J., concurring) (per curiam). "[T]he polity's demographics and actual voting patterns in particular elections" are fact-finding endeavors for the district court, while "[t]he subsequent determination of the legal inferences to be drawn from those facts . . . involve questions of law and the application of legal standards." Id.

Second, when a plaintiff has established the Gingles preconditions, the court must undertake a "searching practical evaluation of the 'past and present realities'" of a challenged system to determine whether, under the totality of circumstances, "the political process is equally open to minority voters." Gingles, 478 U.S. at 79 (plurality opinion) (quoting S. Rep. No. 97-417, at 30 (1982), as reprinted in 1982 U.S.C.C.A.N. 177, 179, 200).

The ultimate finding that minorities do or do not possess equal opportunities to participate in the political process is a question of fact. See id. at 78 ("[W]e have treated the ultimate finding of vote dilution as a question of fact . . . ."); S. Christian Leadership Conference of Alabama v. Sessions, 56 F.3d 1281, 1294 (11th Cir. 1995) [hereinafter SCLC] (en banc) (stating that the determination of vote dilution is an "ultimate issue of fact"). "'This determination is peculiarly dependent upon the facts of each case' . . . and requires 'an intensely local appraisal of the design and impact' of the contested electoral mechanisms." Gingles, 478 U.S. at 79 (quoting Rogers v. Lodge, 458 U.S. 613, 621-622 (1982)); see also Johnson v. De Grandy, 512 U.S. 997, 1011 (1994) (holding that the finding of vote dilution must rest "on comprehensive, not limited, canvassing of relevant facts"). Thus, in evaluating the totality of circumstances, courts should consider "all of the circumstantial evidence — both statistical and anecdotal — that [is] offered."⁷ SCLC, 56 F.3d at 1292. The "ultimate burden" to prove that, under the totality of circumstances, the political process is not equally open to the minority group remains with the plaintiff. Askew v. City of Rome, 127 F.3d 1355, 1375 (11th Cir. 1997).

7 This legal requirement is entirely fair, but it results in long opinions such as this one.

3. The Meaning of Section 2

The task here is not an easy one. "Because [§ 2] reflects a compromise between two very different views in the Congress that passed the 1982 amendment to section 2, much of that section's language seems inherently inconsistent and, at times, virtually meaningless." Solomon, 899 F.2d at 1022 (Tjoflat, J., concurring). Regardless, the court must ascertain the meaning of § 2 as it relates to judicial elections and apply it to the facts at hand. In doing so, it is necessary to first review both the legislative history of the 1982 amendments and the Eleventh Circuit precedents applying § 2 to judicial elections.

a. Legislative History

The Voting Rights Act of 1965 was enacted to secure the Fifteenth Amendment's promise of equal access to the political process to all Americans. See Pub. L. No. 89-110, 79 Stat. 437 (1965) (codified as amended at 52 U.S.C. § 10301 (2018)); South Carolina v. Katzenbach, 383 U.S. 301, 315 (1966) ("The Voting Rights Act of 1965 reflects Congress' firm intention to rid the country of racial discrimination in voting."). As originally enacted, the Supreme Court held that § 2 offered no more protection than did the Fifteenth Amendment. City of Mobile v. Bolden, 446 U.S. 55, 60-61 (1980) (plurality opinion). Thus, to invalidate a voting practice, a plaintiff had to show that the challenged law is "motivated by a discriminatory purpose." Id. at 62. The standard articulated in Bolden is known as the "intent test."

Congress disagreed with the Supreme Court's intent-based reading of § 2. So it amended the VRA in 1982 to make it easier for plaintiffs to challenge official practices that have the effect of weakening minority voting strength on account of race, regardless of the actual motivations of the decisionmakers. Under amended § 2, the plaintiff need not show discriminatory intent in a challenged law. He or she need only prove that a "voting qualification or prerequisite to voting or standard, practice, or procedure . . . results in a denial or abridgement of the right . . . to vote on account of race or color." § 10301(a) (emphasis added). Section 2(b) provides that the test for determining a section 2(a) violation is whether, under the "totality of circumstances," the political process is "not equally open to participation" by members of a minority group in that they "have less opportunity" than majority voters "to participate in the political process and to elect representatives of their choice" under the current system. Id. § 10301(b). Congress thus changed the "intent test" into a "results test" that had been articulated in two pre-Bolden cases, Whitcomb v. Chavis, 403 U.S. 124 (1971), and White v. Regester, 412 U.S. 755 (1973). See S. Rep. No. 97-417, at 2, 23 (1982), as reprinted in 1982 U.S.C.C.A.N. 177, 179, 200).

Congress was careful, however, to deny that minorities are entitled to proportional representation. See § 10301(b) ("[N]othing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population."). The court must therefore "make every effort to be faithful to the balance Congress struck." Gingles, 478 U.S. at 84 (O'Connor, J., concurring) (noting the "inherent tension" between the results test and Congress's disclaimer of a right to proportional representation "because any theory of vote dilution must necessarily rely to some extent on a measure of minority voting strength that makes some reference to the proportion between the minority group and the electorate at large").

b. Gingles Preconditions and Totality-of-Circumstances Test

The 1982 amendments to the VRA clarified that a plaintiff could make out a § 2 claim by showing that a facially neutral system operates to deny minorities equal access to the political process on account of their race. See S. Rep. No. 97-417, at 27-30. The Supreme Court established the standards governing § 2's "results test" in Gingles. See 478 U.S. at 30. To succeed on a § 2 vote dilution claim, a plaintiff initially must prove three preconditions: (1) that the minority group is "sufficiently large and geographically compact to constitute a majority in a single-member district"; (2) that the minority group is "politically cohesive" (i.e., that members of the group generally vote the same way); and (3) that "the white majority votes sufficiently as a bloc to enable it . . . usually to defeat the minority's preferred candidate."⁸ Id. at 50-51. The first prong speaks to the feasibility of a plaintiff's proposed remedy, see Nipper v. Smith, 39 F.3d 1494, 1530-31 (11th Cir. 1994) (en banc), while the second and third prongs are statistical measures of minority voting power. The Court collectively shorthanded the three Gingles preconditions as proof of "racially polarized voting" or "racial bloc voting," using the terms interchangeably. Gingles, 478 U.S. at 52 n.18.

8 Gingles refers to the "minority group"; section 2 refers to the "protected class." The phrases are used in this opinion interchangeably.

Proof of the three preconditions is necessary, but not sufficient, to succeed on a § 2 claim. Johnson, 512 U.S. at 1011. "This is because it is entirely possible that bloc voting . . . could exist, but that such bloc voting would not result in a diminution of minority opportunity to participate in the political process and elect representatives of the minority group's choice." Solomon v. Liberty Cty. Comm'rs, 221 F.3d 1218, 1225 (11th Cir. 2000) (en banc); SCLC, 56 F.3d at 1293-94 (affirming the district court's judgment in favor of the State on a § 2 claim when "factors other than race, such as party politics . . . , were driving the election results"); see also Askew, 127 F.3d at 1379 (citing the latter holding from SCLC with approval). A more vernacular way to say it is that bloc voting may exist, but defeat of preferred candidates of the protected class may occur on account of causes other than race.

Thus, after a plaintiff has satisfied the three Gingles preconditions, the court must determine, as § 2(b) dictates, whether the plaintiff has proven that, "based on the totality of circumstances, . . . the political processes leading to . . . election in the State . . . are not equally open to participation by members" of the protected class. § 10301(b). "Gingles provided some structure to [§ 2's] 'totality of circumstances' test in a case challenging multimember legislative districts." Johnson, 512 U.S. at 1010 (citing Gingles, 478 U.S. at 46-51). Gingles held that, depending on the nature of the claim, some or all of the factors in the Senate Report accompanying the 1982 amendments may be relevant to a claim of vote dilution:

1. the extent of any history of official discrimination in the state or political subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process;

2. the extent to which voting in the elections of the state or political subdivision is racially polarized;

3. the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority group;

4. if there is a candidate slating process, whether the members of the minority group have been denied access to that process;

5. the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as

education, employment and health, which hinder their ability to participate effectively in the political process;

6. whether political campaigns have been characterized by overt or subtle racial appeals;

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

Additional factors that in some cases have had probative value as part of plaintiffs' evidence to establish a violation are:

[8.] whether there is a significant lack of responsiveness on the part of elected officials to the particularized needs of the members of the minority group.

[9.] whether the policy underlying the state or political subdivision's use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.

478 U.S. at 36-37 (quoting S. Rep. No. 97-417, at 28-29) (alterations added). The two most important factors "are the 'extent to which minority group members have been elected to public office in the jurisdiction' and the 'extent to which voting . . . is racially polarized.'" Id. at 48 n.15 (quoting S. Rep. No. 97-417, at 28-29).

The nine "objective factors" are indicia of whether the minority group has equal access to the political process, id. at 44 (quoting S. Rep. No. 97-417) (internal quotation marks omitted), but are "neither comprehensive nor exclusive," id. at 45 (citing S. Rep. No. 97-417, at 29-30). And while the Senate factors are relevant to some types of § 2 cases, they may not be relevant to others. See id. at 45 (citing S. Rep. No. 97-417, at 29-30); see also Nipper, 39 F.3d at 1529 ("[S]erious problems [may] lie ahead in applying the 'totality of circumstances' factors developed in section 2 cases involving legislative elections to the election of judges." (internal quotation marks omitted) (quoting Chisom v. Roemer, 501 U.S. 380, 403 (1991))). Moreover, "there is no requirement that any particular number of factors be proved, or that a majority of them point one way or another." Gingles, 478 U.S. at 45 (quoting S. Rep. No. 97-417, at 29).

At bottom, the totality-of-circumstances inquiry asks whether a neutral electoral standard, practice, or procedure, when "interacting with social and historical conditions," works to deny a protected class the ability "to elect their candidate of choice on an equal basis with other voters." Voinovich v. Quilter, 507 U.S. 146, 153 (1993) (quoting Gingles, 478 U.S. at 47).

4. Section 2 and At-Large Judicial Elections

In 1991, the U.S. Supreme Court interpreted the word "representative" in § 2 to cover Louisiana's state supreme court justices. See Chisom v. Roemer, 501 U.S. 380, 404 (1991). Writing for the majority, Justice Stevens reasoned:

We think . . . the word "representatives" describes the winners of representative, popular elections. If executive officers, such as prosecutors, sheriffs, state attorneys general, and state treasurers, can be considered "representatives" simply because they are chosen by popular election, then the same reasoning should apply to elected judges. . . . The fundamental tension between the ideal character of the judicial office and the real world of electoral politics cannot be resolved by crediting judges with total indifference to the popular will while simultaneously requiring them to run for elected office.

Id. at 399, 400-01. But see id. at 405 (Scalia, J., dissenting) ("[J]udges are not representatives.").

Louisiana appellate courts are unique in their electoral structure. Louisiana is one of only four states that use districts to elect appellate judges. The Louisiana Supreme Court has seven justices. At the time Chisom was decided, five of the seven justices were elected from five single-member districts, and the remaining two justices were elected at-large from a single district.⁹ See id. at 384. The Court held that African-American registered voters could bring a § 2 vote dilution challenge to the at-large multimember district. See id. at 384-85. During the same term, the Supreme Court also held, applying the reasoning in Chisom, that § 2 applied to a vote dilution challenge to Texas's at-large method of electing multiple state district judges from countywide districts (i.e., another challenge to a multimember districting plan). Houston Lawyers' Ass'n v. Attorney Gen. of Texas, 501 U.S. 419, 422-23 (1991). However, the Court left it to the lower courts to figure out how to apply § 2 to judicial elections. See id. at 426 ("We deliberately avoid any evaluation of the merits . . . ."); Chisom, 501 U.S. at 390 (stating that the sole "question involves only the scope of the coverage of § 2 of the Voting Rights Act as amended in 1982. We therefore do not address any question concerning the elements that must be proved to establish a violation of [§ 2] or the remedy that might be appropriate to redress a violation if proved.").

9 Louisiana has gone its own way in music (jazz), cuisine (Cajun and Creole), law (civil) and appellate courts (districts). As one of only four states selecting appellate judges by districts, and as a place where politics may be rougher than football, Louisiana ended up with one of its judicial districts — but only one — with an at-large feature that had to be resolved by the U. S. Supreme Court in Chisom.

The Supreme Court also has held, in contexts outside state judicial elections, that § 2 applies to at-large elections. See Growe v. Emison, 507 U.S. 25, 37-42 (1993); see generally Shelby Cty. v. Holder, 570 U.S. 529, 583 (2013) (discussing Alabama's Dillard § 2 litigation that "ultimately expanded to include 183 cities, counties, and school boards employing discriminatory at-large election systems" (citing Dillard v. Baldwin Cty. Bd. of Educ., 686 F. Supp. 1459, 1461 (M.D. Ala. 1988))). "[A]t-large election schemes, however, are not per se violative of minority voters' rights." Gingles, 478 U.S. at 48 (S. Rep. No. 97-417, at 16).

What the Supreme Court has not held also is significant. Chisom involved a § 2 challenge to Louisiana's districting plan in which two of its seven supreme court justices were elected at-large in a multimember district. Emphasizing that its holding was "limited in character," the Chisom Court took pains to identify "certain matters that [were] not in dispute." 501 U.S. at 390. One of those undisputed matters was that "there [was] no question that the terms 'standard, practice, or procedure' are broad enough to encompass the use of multimember districts to minimize a racial minority's ability to influence the outcome of an election covered by § 2." Id. (emphasis added). "The only matter in dispute [was] whether the test for determining the legality of such a practice [i.e., the use of multimember districts] . . . applies in judicial elections as well as in other elections," id. at 390-91 (alterations added), and the Court held that the § 2 test did so apply.

Chisom's holding makes sense when appellate judges are elected from districts, similar to state trial judges (as in Houston Lawyers'). The Supreme Court did not say — nor has it ever said — that the election of appellate judges in at-large, statewide non-districted elections is a standard, practice, or procedure governed by § 2. See United States v. Jones, 57 F.3d 1020, 1024 n.7 (11th Cir. 1995) ("The Voting Rights Act does not define standard, practice or procedure." (citing Holder, 512 U.S. at 947 (Blackmun, J., dissenting))). This court is not convinced that it would so say.

Alabama's system of at-large, statewide elections for appellate judges makes this case an exceptionally poor conceptual vehicle to travel the vote dilution highway. At its core, this case is about Alabama's choice of the form of its government, specifically the judicial branch. This is not a districting case in which existing districts are subject to analysis on how one vote may be diluted as opposed to another that is, by default, enhanced. Nor is it a legislative districting case about political subdivisions of the state, e.g., municipalities, counties and the like, together with their city councils, county commissions, and other elected bodies. Those subdivisions are not sovereign; they are themselves subdivisions of the sovereign state. When those governmental entities establish multimember bodies for governance, even with the help of the legislature, those decisions are subject to § 2.

Districting is a remedy for dilution of votes in those political subdivisions because the courts have found benchmarks for comparison of remedies. The point is that one must have a "reasonable alternative practice as a benchmark against which to measure the existing voting practice." Holder, 512 U.S. at 880 (plurality opinion). In Holder, the Court held that § 2 could not be used to force a Georgia county to change from a single commissioner form of government to a multimember form of government: A change in the size of the body was beyond the purview of § 2, which is to say, the form of governing that Georgia county was beyond the scope of § 2. There were no districts in that county structure, and a federal court could not require the governing authority to change to a district form of government by calling it an increase in the size of its "commission." See id. at 885.

This form-of-government holding has been overlooked. In Gingles, Justice O'Connor defined vote dilution as "the impermissible discriminatory effect that a multimember or other districting plan has when it operates 'to cancel out or minimize the voting strength of racial groups.'" 478 U.S. at 87 (O'Connor, J., concurring in the judgment) (emphasis added) (quoting White v. Regester, 412 U.S. 755, 765 (1973)). The state at large is decidedly not a districting plan, nor is it a district of anything. Unlike a county, city, state legislature, or school board, the state is itself a sovereign. "[T]he Framers of the Constitution intended the States to keep for themselves, as provided in the Tenth Amendment, the power to regulate elections." Shelby Cty., 570 U.S. at 543 (quoting Gregory v. Ashcroft, 501 U.S. 452, 461-62 (1991)). "[E]ach State has the power to prescribe the qualifications of its officers and the manner in which they shall be chosen." Id. (quoting Boyd v. Nebraska ex rel. Thayer, 143 U.S. 135, 161 (1892)). Plaintiffs here ask this court to condemn Alabama's chosen form of government of the judicial branch by fundamentally changing the way appellate judges are selected.

Plaintiffs assume § 2 applies because subsection (a) addresses a voting "standard, practice, or procedure . . . imposed or applied by any State" § 10301. The court has serious concerns about this assumption. Alabama's at-large, statewide system is a fundamental choice of the form of its judicial branch of government. That it involves procedures is of no moment; all forms of government involve procedures.

Because Plaintiffs' § 2 claim fails on other grounds and because the parties have not briefed this issue, the court merely assumes, without deciding, that the terms "standard, practice, or procedure" encompass the use of a state's at-large, statewide system of electing appellate judges.

a. Nipper v. Smith and Later Eleventh Circuit Caselaw Developments

Taken as a whole, Nipper v. Smith is the Eleventh Circuit's fractured attempt to apply the vagaries of § 2 — as painted over by Gingles — to judicial elections. In Nipper, two judges — of eight judges sitting en banc for a twelve-judge court — delivered the plurality opinion of the court. 39 F.3d at 1496-1547 (Tjoflat, C.J., joined by Anderson) (plurality opinion). In reality, Judge Edmondson's concurrence, joined by three colleagues, in parts of the two-judge opinion is a 6-2 majority holding as to those parts in which Judge Edmondson concurred. See id. at 1547 (Edmondson, J., concurring in part and concurring in the judgment, joined by Cox, J., Birch, J., and Dubina, J.) (concurring "in Parts III(A) and III(B)(1) of Chief Judge Tjoflat's opinion"). This court treats as binding analysis only those portions of Nipper joined in by the four-judge concurrence, which would be a six-judge holding.

Nipper was a § 2 challenge to at-large elections for trial judges in Florida. Two judges of the court concluded that, although there was racially polarized voting under the second and third Gingles preconditions, id. at 1537 (plurality opinion), the State of Florida's linkage interest foreclosed each of the plaintiffs' proposed subdistricting remedies, id. at 1543-47.

i. The Importance of a State's Interests

Judge Edmondson concurred in the result reached by Judge Tjoflat: The State's "legitimate interest in maintaining linkage between jurisdiction and the electoral bases of its trial judges is, as a matter of law, great and outweighs (either at the vote-dilution-finding stage or [as opined by Judge Tjoflat] at the remedy stage) whatever minority vote dilution that may possibly have been shown here."¹⁰ Id. at 1547 (Edmondson, J., concurring in part and concurring in the judgment). Judge Tjoflat agreed that linkage "play[ed] a major role" in the defeat of the plaintiffs' proposed remedies, id. at 1542. Hence, six judges in Nipper concluded that the plaintiffs' § 2 vote dilution claim failed on account of the state's linkage interest.

10 In plain language, "linkage" means citizens get to vote for or against judges who exercise jurisdiction over those citizens. See Part V.A.7.b. infra.

The four concurring judges agreed with those parts of Chief Judge Tjoflat's two-judge opinion that explained that, in § 2 challenges to judicial elections, the court must consider the state's interest in its present system as part of the totality of circumstances, and "there must be a remedy within the confines of the state's judicial model that does not undermine the administration of justice." Id. at 1527-33. The Edmondson approach gives heavy weight to a state's interests as a matter of law in analyzing both the totality of circumstances and the proposed remedy, at least with respect to state trial judges.

That the Edmondson approach more closely approximates the present state of the law is borne out in SCLC, 56 F.3d 1281, and Davis v. Chiles, 139 F.3d 1414 (11th Cir. 1998). In SCLC, the Eleventh Circuit upheld the district court's finding that there was no vote dilution in the at-large elections of Alabama's trial judges. 56 F.3d at 1294. But the court went on to hold that, assuming there was vote dilution, the plaintiffs' districting remedy was precluded as a matter of law based on the State's interests. Id. For starters, districting would limit "the available pool of qualified black potential judicial candidates." Id. at 1296; see also id. ("The State's interest in having a pool of black lawyers large enough to give the voters a reasonable choice from which to select a qualified candidate must be considered when evaluating suggested remedies."). Districting also "would directly implicate the State's linkage interest," id., eviscerating the right of Alabama citizens "to hold a judge accountable for his or her performance" and diluting the voting power of both black and white minorities forced into a subdistrict. Id. at 1296-97. Finally, districting would "increase the specter of 'home cooking'" and favoritism. Id. at 1297.

The reasoning of the en banc court in SCLC seems to preclude any race-based districting in § 2 challenges to at-large elections for state trial judges. In Davis, that is exactly how a later Eleventh Circuit panel interpreted it. Davis also involved a challenge to at-large elections for state trial judges. See 139 F.3d at 1416. Notwithstanding evidence of racially polarized voting, the Eleventh Circuit held that the State of Florida's interest in maintaining its judicial selection model and preserving linkage trumped the plaintiffs' proposed remedy. Id. at 1423. However, the court was "troubled by" circuit precedent that "appear[ed] to conflict" with the Supreme Court's decision in Houston Lawyers' that a state's "'interest in maintaining an at-large, district-wide electoral scheme for single-member [judicial] offices is merely one factor'" in the totality-of-circumstances analysis and "'does not automatically, and in every case, outweigh proof of racial vote dilution.'" Id. at 1423 (quoting Houston Lawyers', 501 U.S. at 427). The Davis court observed that "our circuit in Nipper and SCLC has placed what now seems . . . to be an insurmountable weight on a state's interest in preserving its constitution's judicial selection system and in maintaining linkage between its judges' jurisdictions and electoral bases." Id.; see also id. at 1421 n.16 ("Because of the importance of this linkage interest, our circuit has effectively ruled out the division of at-large judicial election districts into separate subdistricts as a permissible remedy." (first citing Nipper, 39 F.3d at 1543-45 (plurality opinion), 1547 (concurring opinion); and then citing SCLC, 56 F.3d at 1296-97)).

Nipper, SCLC, and Davis addressed § 2 challenges to Alabama's and Florida's method of electing state trial judges. In the only § 2 lawsuit challenging Alabama's at-large method of electing appellate judges (besides the instant case), the Eleventh Circuit held that § 2 prohibited the settlement agreement's remedy of creating a system for nominating and appointing African Americans to the appellate courts to achieve proportional representation.¹¹ See White v. Alabama, 74 F.3d 1058, 1068-73 (11th Cir. 1996). The Eleventh Circuit held that "[a] remedy such as the one fashioned in this case, calling for the appointment of judges to posts which, under state law, are to be filled by election, effectively nullifies voting power and contravenes the stated objectives of section 2." Id. at 1070. Accordingly, the supplementing of elections with appointments, which involved potentially increasing the size of the appeal courts, is not a form of relief within the scope of § 2. Hence, in Davis the Eleventh Circuit observed that, "[t]ogether with Nipper, SCLC, and the additional case of White v. Alabama, we will with this decision have disallowed redistricting, subdistricting, modified subdistricting, cumulative voting, limited voting, special nomination, and any conceivable variant thereof as remedies for racially polarized voting in at-large judicial elections." 139 F.3d at 1423.

11 The settlement agreement's terms were incorporated into the district court's judgment. See id. at 1068. On remand from the circuit, there were no further proceedings in the district court. The plaintiffs in White moved to dismiss their § 2 claims, stating:

Since the filing of this action, a number of significant legal and factual events have occurred, including the decisions of the Eleventh Circuit Court of Appeals in Nipper v. Smith, 39 F.3d 149[4] (11th Cir. 1994) (en banc); Southern Christian Leadership Conference v. Sessions, 56 F.3d 1281 (11th Cir. 1995) (en banc); and White v. State of Alabama, 74 F.3d 1058 (11th Cir. 1996); and the election of Justice Ralph Cook to the Supreme Court of Alabama.

Joint Motion for Dismissal Without Prejudice, White v. Alabama, No. 94-T-94-N (M.D. Ala. Sept. 26, 1996), Doc. # 253; see also White, 74 F.3d at 1072 n.44 (noting that "the district court did not have the benefit of Nipper's holding when it decided this case"). The district court granted the motion. See Order, White v. Alabama, No. 94-T-94-N (M.D. Ala. Oct. 3, 1996), Doc. # 254.

ii. Nipper's Applicability to Appellate Judicial Elections

The foregoing statement would seem to be the end of the matter for this case. But Nipper, SCLC, and Davis dealt with at-large elections for trial judges, not appellate judges. As mentioned in an earlier order in this case (Doc. # 45, at 8), Nipper cautioned that the issue in that case "concern[ed] the election of trial court judges, not the members of a multimember appellate court," and that "there might be more to be said for some form of 'representation' on a collegial court (like a state supreme court) than on a single-judge trial court. The ability to bring diverse perspectives to the court, not the prospect of outright dealmaking, would be relevant in such a context." Nipper, 39 F.3d at 1535 n.78 (plurality opinion). Although the two-judge opinion's reference in this footnote to trial versus appellate judges and lone decisionmakers versus collegial bodies is not binding because the four concurring judges withheld their assent to that part of the opinion, the en banc decision in SCLC and the unanimous panel decision in Davis incorporated the same principle in their holdings that the plaintiffs' proposed districting remedies impermissibly vitiated the states' linkage interest. See SCLC, 56 F.3d at 1297 (explaining that, with districting, the loss of minority influence on judicial decision making "would be more pronounced in [the] context of a lone decisionmaker, a trial judge, who would lack input from the colleagues elected by the rest of the citizenry of the jurisdiction"); see also Davis, 139 F.3d at 1422 (quoting the same passage from SCLC with approval).

Other than these stray statements in Nipper, SCLC, and Davis, the Eleventh Circuit has provided no guidance on how this should be done. At least one district court outside this circuit has considered the issue and found that the State of Texas's then-142-year-old constitutional adoption of linkage (corresponding the jurisdiction of its high courts "with the limits of the State") was strong and prevailed over the plaintiffs' attempts to minimize that interest by distinguishing the work of trial courts from appellate courts. See Lopez v. Abbott, 339 F. Supp. 3d 589, 618-19 (S.D. Tex. 2018); see also id. at 619 ("[T]he collegial nature of a multimember body does not prevent it from, in practice, running roughshod over any given minority of judges. Single-member districts could have the unintended effect of increasing the power of a majority of judges elected from districts with wider polarization levels in favor of white voters."). When evaluating the State's linkage interest, see infra Part V.A.7.b., the court will compare and contrast Alabama's trial and appellate courts.

iii. The Role of Causation in the § 2 Vote Dilution Analysis

The requirement that § 2 plaintiffs must prove that racial bias is driving the election results is not the law. See Nipper, 39 F.3d at 1514-26 (plurality opinion). This requirement, adopted by the Nipper two-judge opinion, has gained no traction in the circuit. See Hamrick, 196 F.3d at 1220 ("Finally, two judges of this Court consider racial bias in the voting community to be a relevant factor.") (citing Nipper, 39 F.3d at 1524 (plurality opinion of Tjoflat, C.J., joined by Anderson, J.); Johnson v. DeSoto Cty. Bd. of Comm'rs, 72 F.3d 1556, 1564 n.8 (11th Cir. 1996) (relegating Nipper's racial-bias requirement to "dictum" that "was joined only by two of the eight members of this Court participating in Nipper" and declaring it "inconsistent with the express contrary holding by the Supreme Court in Voinovich" (citing Voinovich, 507 U.S. at 153)). Obviously, Plaintiffs can present evidence of discriminatory intent, but they are not required to do so to prove a § 2 violation.

In SCLC, the Eleventh Circuit endorsed the district court's view that "any evidence that explained election results was relevant" and held that "there was ample evidence in the record to support the court's conclusion that factors other than race, such as party politics and availability of qualified candidates, were driving the election results." 56 F.3d at 1293-94 (emphasis added); see also Whitcomb, 403 U.S. at 153 (opining that the district court's finding that the "voting power" of the African-American community "may have been 'cancelled out'" seemed to be but a mere "euphemism for political defeat at the polls"); Solomon, 221 F.3d at 1225 (noting that "what appears to be bloc voting on account of race may, instead, be the result of political or personal affiliation of different racial groups with different candidates"). And as Justice O'Connor (joined by three other Justices) explained in Gingles, there may be "an underlying divergence in the interests of minority and white voters" that is benign of racial hostility. 478 U.S. at 100 (O'Connor, J., concurring in the judgment). It is thus appropriate for courts to consider "all evidence concerning voter preferences," not just "statistical evidence of racial voting patterns." Id. at 101.

The Solomon decision indicates that a state's evidence of non-racial causes is to be considered, alongside all other relevant factors bearing on the existence or not of vote dilution, at the totality-of-circumstances stage. See 221 F.3d at 1225. The First, Second, and Seventh Circuits have taken that approach as well. See Goosby v. Town Bd. of Town of Hempstead, 180 F.3d 476, 493 (2d Cir. 1999) ("We . . . ratify the approach taken by the district court to consider the political partisanship argument under the 'totality of circumstances' analysis rather than as part of the third Gingles precondition."); Milwaukee Branch of the NAACP v. Thompson, 116 F.3d 1194, 1199 (7th Cir. 1997) (explaining that the reasons why candidates preferred by black voters lost should be considered in the totality-of-circumstances inquiry); Vecinos De Barrio Uno v. City of Holyoke, 72 F.3d 973, 983 (1st Cir. 1995) (holding that non-racial reasons for divergent voting patterns should be considered under the totality-of-circumstances test).

Based on the foregoing, Plaintiffs are not required to prove the causes of any racially polarized outcomes as part of their burden of proof on the Gingles preconditions. Why black-preferred candidates lost is not part of their burden at this juncture. Rather, the State's evidence that factors other than race are driving election results will be considered at the totality-of-circumstances stage.

b. Summary

Judicial elections are distinctive in § 2 challenges. Although proof of all three Gingles preconditions usually warrants relief in other contexts, the Eleventh Circuit has simply not treated challenges to judicial-election methods like other § 2 cases. Nipper, SCLC, and Davis instruct that federal courts should be highly reluctant to restructure a state's judiciary. These cases make clear that the State has weighty interests in (1) maintaining a judicial selection system that is consistent with its constitutional model, (2) maintaining linkage between a judge's jurisdiction and his or her electoral base, and (3) ensuring the availability of a sufficient pool of qualified African-American potential judicial candidates. These interests bear highly on the feasibility of Plaintiffs' proposed districting remedy, as well as on the totality-of-circumstances analysis, but they may not play out the same way in the contexts of appellate courts as opposed to trial courts.

Evidence that race is or is not driving the election results is relevant in the totality of circumstances. At the totality-of-circumstances stage, the State may introduce evidence that nonracial factors, such as partisan politics or party affiliation, are causing minority electoral defeats.

5. Analysis of the Gingles Preconditions

a. Introduction

Plaintiffs contend that they have satisfied each of the three Gingles preconditions. They assert that they have submitted hypothetical districts for Alabama's appellate courts that contain at least one African-American district that is sufficiently large and geographically compact. These hypothetical districts, according to Plaintiffs, demonstrate that districting is a feasible remedy. They also contend that they have proven the second and third Gingles preconditions principally through statistical evidence of racially polarized voting.

The State admits that Plaintiffs have submitted two hypothetical plans for the Alabama Supreme Court in which one majority-black district satisfies the numerosity and compactness requirements of the first Gingles precondition. However, it argues that Plaintiffs' § 2 claim cannot survive the first Gingles precondition because (1) the African-American population in the majority-black district (District 1) in Plaintiffs' illustrative plans for the intermediate appellate courts is not reasonably compact, (2) District 1 is an indefensible racial gerrymander, and (3) the State's interests in maintaining its at-large, statewide system for electing appellate judges strongly outweigh any interest Plaintiffs have in the adoption of districting as a remedy.

As to the second and third Gingles preconditions, the State says that it "do[es] not dispute that most black Alabama voters prefer a different candidate than most white Alabama voters, in most elections." Post-Trial Stipulations, at 154, ¶ 29 (Doc. # 176). The State says that, if "racially polarized voting' means no more than that, Plaintiffs have met their burden under Gingles 2 and 3."¹² Id.

12 The State contends that Plaintiffs' § 2 vote dilution must fail because partisan politics, not racial bias, is causing black electorate defeat at the polls. Thus, the State's evidence focuses on the causes of racially polarized voting, rather than the fact of racially polarized voting. As stated, the causes of racially polarized voting are addressed in the discussion of the totality of circumstances.

For the reasons that follow, Plaintiffs have satisfied the second and third Gingles preconditions. However, they have not proven the first Gingles precondition in two respects. First, as to the intermediate appellate courts, Plaintiffs have not shown that the African-American population in the hypothetical plans is geographically compact. Second, Plaintiffs have not submitted feasible plans for any of Alabama's appellate courts against which to measure the alleged vote dilution in Alabama's present at-large, statewide system of elections for appellate judges, as required in this circuit to satisfy the first Gingles precondition.

b. The First Gingles Precondition

There is a lot to unpack in the discussion of the first Gingles precondition. Discussed below are: (1) the inextricably intertwined nature of liability and remedy in the Eleventh Circuit; (2) factors governing the first Gingles precondition; (3) Plaintiffs' illustrative plans for Alabama's appellate courts; (4) whether the illustrative plans for the Alabama Supreme Court satisfy the first Gingles precondition; and (5) whether the illustrative plans for Alabama's intermediate appellate courts satisfy the first Gingles precondition.

i. The Inextricably Intertwined Nature of Liability and Remedy in the Eleventh Circuit

The Eleventh Circuit has held that, in § 2 cases, "the inquiries into remedy and liability . . . cannot be separated." Nipper, 39 F.3d at 1530-31. "Implicit in th[e] first Gingles requirement is a limitation on the ability of a federal court to abolish a particular form of government and to use its imagination to fashion a new system." Id. at 1531. Based on this implied constraint, the Eleventh Circuit has added an additional layer to the first Gingles precondition in judicial-elections cases: As part of the first Gingles precondition, Plaintiffs must prove a remedy "within the confines of the state's judicial model that does not undermine the administration of justice." Davis, 139 F.3d at 1421 (emphasis in original) (quoting Nipper, 39 F.3d at 1531); see also Burton v. City of Belle Glade, 178 F.3d 1175, 1199 (11th Cir. 1999) (observing that the Eleventh Circuit "ha[s] repeatedly construed the first Gingles factor as requiring a plaintiff to demonstrate the existence of a proper remedy" and collecting cases, including Nipper, 39 F.3d at 1530-51). To prove liability, Plaintiffs must establish the Gingles preconditions, including a feasible remedy, and then show, under the totality of circumstances, that Alabama's appellate judicial elections "are not equally open to participation" by African Americans. § 10301(b). Thus, Eleventh Circuit law requires the court to decide the propriety of Plaintiffs' remedy before deciding whether there is even vote dilution at all. That is, the court must decide whether there is a remedy before deciding whether there is something to be remedied. While this requirement appears to "put[] the cart before the horse," it is the law of this circuit. See SCLC, 56 F.3d at 1302 (Hatchett, J., dissenting).

Notwithstanding the law of the circuit, the issue of remedy, on which Plaintiffs' § 2 case also fails, is purposefully deferred until Part V.A.7. so as to permit a complete and robust analysis of all facets of Plaintiffs' § 2 vote dilution claim.

ii. Factors Governing the First Gingles Precondition

To establish the first Gingles precondition, a § 2 plaintiff must demonstrate that the members of the minority group are "sufficiently large and geographically compact to constitute a majority in a single-majority district." Gingles, 478 U.S. at 50 & n.17. Gingles explained the purpose of this first requirement: "Unless minority voters possess the potential to elect representatives in the absence of the challenged structure or practice, they cannot claim to have been injured by that structure or practice." Id. at 51 n.17; see also Bush v. Vera, 517 U.S. 952, 979 (1996) (plurality opinion) (noting that, if "because of the dispersion of the minority population, a reasonably compact majority-minority district cannot be created, § 2 does not require a majority-minority district").

A demonstration of numerosity and compactness "requires submitting as evidence hypothetical redistricting schemes in the form of illustrative plans." Gonzalez v. Harris Cty., 601 F. App'x 255, 258 (5th Cir. 2015) (per curiam); see also Gingles, 478 U.S. at 50 n.17. The districts are illustrative only; if Plaintiffs prevail on their § 2 vote dilution claim, the Alabama Legislature would have the first opportunity to create a remedial plan for the election of the State's appellate judges. See, e.g., Clark v. Calhoun Cty., 21 F.3d 92, 95 (5th Cir. 1994) ("[P]laintiffs' proposed district . . . was simply presented to demonstrate that a majority-black district is feasible in Calhoun County. If a § 2 violation is found, the county will be given the first opportunity to develop a remedial plan.").

A minority group is "sufficiently large" if it makes up more than 50% of the voting-age population in the proposed remedial district. Bartlett v. Strickland, 556 U.S. 1, 19-20 (2009) (plurality opinion); see also Growe, 507 U.S. at 38 n.4 ("To establish whether a § 2 violation has occurred . . . [,] other courts have looked to, not the district's total minority population, but the district's minority population of voting age. Gingles itself repeatedly refers to the voting population." (internal citation omitted)). This is an objective 50% plus one rule.

The requirement of compactness, unlike that of numerosity, is subjective. "'[T]he compactness inquiry considers the compactness of the minority population, not the compactness of the contested district." League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 402 (2006) [hereinafter LULAC] (quoting Vera, 517 U.S. at 997 (internal alterations omitted) (quotation marks omitted)). "A district that 'reaches out to grab small and apparently isolated minority communities' is not reasonably compact." Id. (quoting Vera, 517 U.S. at 979). There is, though, "no precise rule" governing § 2 compactness. LULAC provides only generally that, in a § 2 case, compactness should focus on whether the proposed minority district reasonably comports with "'traditional districting principles.'" Id. (quoting Abrams v. Johnson, 521 U.S. 74, 92 (1997)). But the Supreme Court has yet to examine, on the merits, a § 2 vote dilution challenge to a state's method of electing its judges, either at the trial or appellate level. In the congressional apportionment and state legislative districting context, Supreme Court § 2 decisions have recognized six traditional districting principles, summarized here, which the court will apply, but without a lot of confidence.

The first traditional districting principle considers whether there is dispersion of the minority population. See id. at 433 ("[T]here is no basis to believe a district that combines two farflung segments of a racial group with disparate interests provides the opportunity that § 2 requires or that the first Gingles condition contemplates.").

The second traditional districting principle addresses the district's shape. Even in the § 2 compactness inquiry, "[d]istrict shape is not irrelevant." Vera, 517 U.S. at 980 (plurality opinion) (relying primarily on the bizarre shape and noncompactness of district lines to conclude that the district was not compact under § 2 jurisprudence). For instance, a district's distorted shape might expose the dispersion of the district's minority population. See Sensley v. Albritton, 385 F.3d 591, 596 (5th Cir. 2004) ("As the geographical shape of any proposed district necessarily directly relates to the geographical compactness and population dispersal of the minority community in question, it is clear that shape is a significant factor that courts can and must consider in a Gingles compactness inquiry.").

The third traditional districting principle is "maintaining . . . traditional boundaries." LULAC, 548 U.S. at 433 (quoting Abrams v. Johnson, 521 U.S. 74, 92 (1997)). These boundaries include county and precinct lines. See Cooper v. Harris, 137 S. Ct. 1455, 1469 n.3 (2017); Abrams, 521 U.S. at 92 ("[T]he § 2 compactness inquiry should take into account 'traditional districting principles such as . . . traditional boundaries.'" (quoting Vera, 517 U.S. at 977)).

The fourth traditional districting principle is "maintaining communities of interest." Evenwel v. Abbott, 136 S. Ct. 1120, 1124 (2016); see also LULAC, 548 U.S. at 433 (providing that "maintaining communities of interest" is a § 2 traditional districting principle). Those communities of interest might be based on location (rural, urban, coastal, or mountain), occupation (industrial or agricultural), political ties, social similarities, or cultural connections. See Miller v. Johnson, 515 U.S. 900, 908 (1995). "The recognition of nonracial communities of interest reflects the principle that a State may not assume from a group of voters' race that they 'think alike, share the same political interests, and will prefer the same candidates at the polls." LULAC, 548 U.S. at 433 (alterations omitted) (citations and internal quotation marks omitted).

The fifth traditional districting principle is contiguity. See Lawyer v. Dep't of Justice, 521 U.S. 567, 581 n.9 (1997); Davis, 139 F.3d at 1425 (observing that the proposed judicial districts "are contiguous"); see also Terrebonne Par. Branch NAACP v. Jindal, 274 F. Supp. 3d 395, 424 (M.D. La. 2017) ("Contiguity as a traditional redistricting principle does not mean that the concentrations of black voters in the proposed district must be contiguous. Rather, it means that the illustrative district itself must be contiguous — it simply has to be connected in one piece."), appeal dismissed sub nom. Fusilier v. Edwards, No. 17-30756, 2017 WL 8236034 (5th Cir. Nov. 14, 2017), and reconsideration denied, No. CV 14-69-SDD-EWD, 2018 WL 5786215 (M.D. La. Nov. 5, 2018). Contiguity, though, generally is not contentious given modern districting software.

The sixth traditional districting principle is protection of incumbents so as to avoid contests between incumbent office-holders. See Karcher v. Daggett, 462 U.S. 725, 740 (1983). Karcher was decided in the context of congressional reapportionment, where the Court observed that a state's legislative goals, such as "avoiding contests between incumbent Representatives," might warrant population deviations among congressional districts. Id. Karcher's rationale seems only marginally relevant to the § 2 compactness of the minority population in a judicial districting case, where population equality is not a constitutional requirement, see infra Part V.A.5.b. However, a § 2 plaintiff in a judicial districting case might contend that a hypothetical district's shape is justified in order to avoid incumbent face-offs. See, e.g., Gingles v. Edmisten, 590 F. Supp. 345, 379, 382 (E.D.N.C. 1984) (approving the state's remedial districts in a § 2 case over the plaintiffs' proposed remedial districts, which were "significantly more regular in shape," because the state was entitled to serve its "primary concern to protect incumbents"), aff'd in part, rev'd in part sub nom. Thornburg v. Gingles, 478 U.S. 30 (1986). Plaintiffs do not make that argument here.

The map drawer is not bound to only apply these principles. There might be other factors that are relevant in the context of judicial districting that would not be at the forefront in the legislative districting context. See Bartlett, 556 U.S. at 19 ("[T]he Gingles requirements cannot be applied mechanically and without regard to the nature of the claim." (citation and internal quotation marks omitted)). Other potential factors have not been argued in this case in any meaningful way.

iii. Plaintiffs' Illustrative Plans for Alabama's Appellate Courts

Plaintiffs argue that districting is a feasible remedy for their § 2 vote dilution claim and retained William Cooper to draw eight illustrative plans. For the Alabama Supreme Court, Mr. Cooper created two nine-district plans for the election of justices from single-member districts and two eight-district plans for the election of the associate justices from single-member districts. The eight-district plans retain an at-large, statewide election for the chief justice. For Alabama's intermediate appellate courts, Mr. Cooper drew four plans, each containing five single-member districts. As instructed, Mr. Cooper created plans with population deviations of plus-or-minus 5% and also plus-or-minus 1%, and he preserved county and judicial-circuit boundaries to the extent possible. Trial Tr. II, at 149, 165 (Cooper). Applying these guidelines, Mr. Cooper configured two majority-black districts for the Supreme Court plans and one majority-black district for the plans for the intermediate appellate courts using whole 2010 Census voting tabulation districts. Mr. Cooper opines that his eight illustrative plans "comply with traditional redistricting principles, including one-person one-vote, compactness, contiguity, respect for communities of interest, and the non-dilution of minority voting strength." Cooper Expert Report, at 33 (Pls.' Trial Ex. 86; Doc. # 68-1); see Post-Trial Stipulations, at 31 (¶ 155) (Doc. # 176).

Mr. Cooper characterizes himself as "a consultant on geographic information systems software applications for redistricting analysis . . . ." Trial Tr. II, at 146 (Cooper). The lion's share of his expertise is in the creation of state and local legislative districting plans, not judicial districting plans.¹³ See Trial Tr. II, at 146-48 (Cooper); Cooper Expert Report, at 1-4 (Pls.' Trial Ex. 86; Doc. # 68-1). The State did not challenge Mr. Cooper's qualifications as an expert on districting for appellate judicial elections in Alabama, and his report was admitted without objection. See Trial Tr. II, at 150 (interposing "no objection" when Plaintiffs' counsel moved to permit Mr. Cooper "to give expert testimony . . . on judicial election district plans in Alabama"). Although Mr. Cooper's testimony has been admitted, aspects of that testimony have diminished credibility.

13 He has served recently as an expert in a case in which a multimember, at-large voting system for one of Louisiana's forty-two state judicial districts (i.e., trial courts) was subject to a § 2 vote dilution challenge. See Terrebonne Par. Branch NAACP, 274 F. Supp. 3d at 417.

iv. Whether the Illustrative Plans for the Alabama Supreme Court Satisfy the First Gingles Precondition

There is no dispute that, in the SC2 and SC4 plans, District 1 has an African-American, voting-age population greater than 50%. The Census includes data on the voting-age population for three categories that are pertinent here: (1) Non- Hispanic Black; (2) Any Part Black; and (3) Single-Race Black. See Cooper Expert Report, at 5-6 (Pls.' Trial Ex. 86; Doc. # 68-1). Because the percentages in each category for the SC2 and SC4 plans exceed 50%, preference need not be given to one category over another. With that explanation, the percentages for Single-Race Black are cited in this opinion. In this category, the voting-age, African-American population is 53.34% in SC2 and 50.7% in SC4. See Pls.' Trial Exs. 98-99, 101-02 (SC2 & SC4 population reports and maps); Cooper Expert Report, at 20, 24 (Pls.' Trial Ex. 86; Doc. # 68-1). Additionally, as to the compactness of the African-American population, District 1's boundaries in both plans are formed solely by county lines from the east to west in the south-central region of the state and encompass all but one of the Black Belt counties, all of which have substantial African-American populations.¹⁴ District 1 is contiguous and also passes the eyeball test for geographical compactness. See Trial Tr. II, at 152-53, 155 (Cooper). Based on this evidence, the State has stipulated that, for the Alabama Supreme Court, "Plaintiffs have shown that in a nine-district and eight-district plan [the SC2 and SC4 plans], the population of African-Americans is sufficiently large and geographically compact to constitute a majority in one single-member district [District 1]." Post-Trial Stipulations, at 152, ¶ 21 (Doc. # 176) (alterations added).

14 Mr. Cooper defined the Black Belt region as encompassing Sumter, Greene, Hale, Marengo, Perry, Dallas, Wilcox, Lowndes, Montgomery, Bullock, Macon, and Barbour counties. Trial Tr. II, at 158 (Cooper). SC2 and SC4 include all of the Black Belt counties except Barbour County.

Based solely on the State's stipulation, without independently analyzing all of the districting principles, the court concludes that Plaintiffs have shown that African Americans are sufficiently numerous and geographically compact to be a majority in a single-member district in two of the Alabama Supreme Court illustrative plans, SC2 and SC4.¹⁵ Whether single-member districts are a feasible remedy, an additional showing required by the Eleventh Circuit to satisfy the first Gingles precondition, is a separate issue.

15 In light of the stipulation, it is unnecessary to address the State's arguments challenging the second majority-black district in the SC2 and SC4 plans and the other two hypothetical plans for the Alabama Supreme Court. See Defs.' Post-Trial Br., at 16-18 (Doc. # 168). The illustrative eight-district Supreme Court plan, SC4, retains the feature — an at-large election for the chief justice — that Plaintiffs contend dilutes African Americans' voting rights. Plaintiffs provide meager justification for doing so. See Pls.' Post-Trial Reply Br., at 26 n.16 (Doc. # 169) ("Plaintiffs provide 8-district plans to give the Legislature a remedy that most closely resembles the current system of electing a Chief Justice in Alabama. If the Legislature retained the method of electing the Chief Justice while simultaneously drawing districts sufficient to cure the harm to Black voters, Plaintiffs['] injury could theoretically be redressed.") (alterations added) (internal citation omitted); Post-Trial Stipulations, at 8-9, ¶ 46 (Doc. # 176) ("The Chief Justice is elected to a designated position and the remaining Justices are elected to numbered places." (citing Ala. Const. Art. VI, § 154)).

v. Whether the Illustrative Plans for Alabama's Intermediate Courts Satisfy the First Gingles Precondition

There is a partial stipulation by the State as to the illustrative plans for the Alabama Court of Civil Appeals and the Alabama Court of Criminal Appeals. As to the four alternative five-district plans for the courts of appeals, each plan contains one district (District 1) with an African-American, voting-age population that exceeds 50%. The percentages of the African-American, voting-age population in District 1 in each plan are as follows: AC1 (54.32%); AC2 (54.32%); AC3 (54.17%); and AC4 (52.34%). See Pls.' Trial Exs. 107-08, 110-11, 113-14, 116-17 (population reports and maps for AC1, AC2, AC3, AC4); Cooper Expert Report, at 26, 28, 29, 30 (Pls.' Trial Ex. 86; Doc. # 68-1). The State stipulates that District 1 in hypothetical Plan AC3 is sufficiently large because its African-American, voting-age population exceeds 50%, see Defs.' Conclusions of Law, at 181, ¶ 67 (Doc. # 176), and, thus, that District 1 satisfies Gingles's numerosity requirement.

However, the State takes issue with District 1's Jefferson County appendage. The court finds that the expert's articulated justification for carving out over 94% of Jefferson County's African-American population and attaching it to District 1 is not grounded in traditional districting principles. Hence, Plaintiffs have not proven that District 1, which is substantially similar in all the AC plans, has an African-American population that is geographically compact.

The Shape of District 1. Mr. Cooper used two methods to evaluate the shape of District 1 of the AC plans: (1) the eyeball test and (2) the Reock test, a statistical method for measuring a district's compactness that yields a score between 0 and 1, with 1 being the most compact. First, Mr. Cooper said his illustrative plans "pass[ed] the eyeball test." Trial Tr. II, at 152 (Cooper) ("[W]hen drawing plans, one always has to examine a district, so-called eyeball test, to see if it seems to be reasonably shaped and reasonably compact."). Second, according to Mr. Cooper, the plans also fared well under the Reock test. See Trial Tr. II, at 153 (Cooper). But he does not discuss any AC plan individually or elaborate on the bases for his opinions on the shape of District 1; hence, these opinions are not entitled to much weight.

Mr. Cooper's statement that the "plans" pass the eyeball test is cursory and does not address the odd feature in District 1 of the AC plans. District 1 in each AC plan rests largely in the south-central region of the state and mainly follows county boundaries. But each majority-black district in the AC plans contains an irregular feature. Unlike the majority-black districts in the SC plans, each of the four AC plans uses Bibb County — a predominantly white county¹⁶ — as a stepping stone to reach northward to Jefferson County to carve out and incorporate its substantial African-American population. District 1 in each AC plan contains the Jefferson County/Bibb County connector as illustrated by a map. See Ct's Appendix A.

16 In District 1, Bibb County has the lowest African-American population (22%) of all the counties, followed by Monroe County, which has an African-American voting-age population of 41.7%. See, e.g., Plan Component Reports for AC-1 (Defs.' Trial Ex. 84).

Each AC plan depends upon this appendage for more than 48% of the African-American population in District 1. See AC Plan Components Reports (Defs.' Trial Exs. 84-87). And, without splitting Jefferson County to take in its African-American population, Mr. Cooper is unable to configure a district in the AC Plans with an African-American, voting-age population that exceeds 50%. Trial Tr. II, at 186-87 (Cooper). The result in each proposed AC plan is an irregularly crafted District 1 with a distorted northward extension that purposefully excludes Jefferson County's white population and adds its black population in order to create a majority-black district. This appendage, teetering atop the northern tip of Bibb County, is a visually odd extension of District 1.

Plaintiffs have not provided credible evidence that this northward extension of District 1 is shapely, and Mr. Cooper winks at the eyeball test. His cursory statement that the AC plans meet the eyeball test lacks foundation because he did not address this abnormal geographical feature of District 1 in the AC plans or disclose any objective basis for his conclusion. There is no evidence that District 1 in the AC plans is visually comparable to any other existing electoral districts in the state.¹⁷ As discussed later, this appendage also is responsible for the geographical dispersal of the African-American population in District 1.

17 The State Board of Education ("BOE") plan, which Mr. Cooper reviewed, does not contain a similarly shaped district. Trial Tr. II, at 154 (Cooper); 2011 BOE Districts (Pls.' Trial Ex. 94). The entirety of District 4's Jefferson County southwest border abuts District 4's boundaries in Bibb County (to the southwest) and Tuscaloosa County (to the west). Jefferson County's attachment in the BOE plan does not have the teetering effect that it has in District 1 of the AC plans and is substantially more compact than District 1.

Second, Mr. Cooper's testimony about the scores from the Reock test is vague and unhelpful. Mr. Cooper provided only a collective range of Reock scores for both the SC and AC plans — between the "high twenties and mid-thirties" — without specifying where the individual plans fell within that range. Trial Tr. II, at 153 (Cooper). It is reasonable to infer from an eyeball test that the AC plans fell closer to the bottom of the range, and Plaintiffs have not demonstrated otherwise. At least one court has found that a district with a Reock score of .30 was non-compact where it failed the eyeball test. See Comm. for a Fair & Balanced Map v. Ill. State Bd. of Elections, 835 F. Supp. 2d 563, 570 (N.D. Ill. 2011) (three-judge district court).

The Compactness of the Minority Population. The African-American population in District 1's Jefferson County carveout is physically separated from the district's other concentrated areas of African-American population by 1,622 square miles, that being the predominantly white Bibb County.¹⁸ There is slight evidence on how this geographical dispersion of African-American population affects the district's compactness measurements. Mr. Cooper's tests measured the geographical compactness of the district's shape, not the compactness of the African-American population within the district's physical boundaries. And Mr. Cooper admitted that, in his plus or minus 1% plans, "district[] boundaries become less regularly shaped because the line-drawer is forced to combine geographically disparate counties." Cooper Expert Report, at 17 (Pls.' Trial Ex. 86; Doc. # 68-1). For these reasons, Plaintiffs have not shown that the African-American population in District 1 is spatially concentrated.

18 Alabama's African-American population is largely concentrated in Jefferson County and in the Black Belt counties. Post-Trial Stipulations, at 31, ¶ 154 (Doc. # 174).

Adherence to traditional boundaries. As instructed by counsel, Mr. Cooper tried to keep counties whole and to avoid splitting the state's forty-two judicial circuits.¹⁹ Trial Tr. II, at 149, 181 (Cooper). District 1 in each AC plan splits Jefferson County, departing with the AC plans' otherwise adherence to county boundaries. See Political Subdivisions Split Between Districts in AC Plans (Defs.' Trial Exs. 76-79). And the AC plans split either one, seven, eight, or eleven judicial circuits, depending on the plan. Cooper Expert Report, at 25, 27, 28, 30 (Pls.' Trial Ex. 86; Doc. # 68-1).

19 Mr. Cooper did not know the rationale for creating districts that kept judicial circuits together. He thought it "may have [had] something to do with case assignments to the various judges," but he admitted, "I just don't know." Trial Tr. II, at 181 (Cooper). In fact, there is no rationale. Alabama's judicial circuits, like the districts of federal district courts, developed organically. They are the product of political give and take over the last 150 years.

The county and judicial circuit splits were the byproduct of Mr. Cooper's adherence to maximum population deviations of either plus-or-minus 1% or 5% of the ideal population. Trial Tr. II, at 149 (Cooper) (testifying that keeping counties and judicial circuits intact "gets a little problematic" when trying to "stay[] within plus or minus one percent or plus or minus five percent deviation. It's almost inevitable that some counties and particularly some of the circuit courts have to be split."); see also Trial Tr. II, at 154 (Cooper) ("[T]here was no way to meet one-person, one-vote without . . . putting Barbour County [a Black Belt county] into . . . another district that was not majority black."); see also Trial Tr. II, at 151-52, 171 (Cooper); Cooper Expert Report, at 33 (Pls.' Trial Ex. 86; Doc. # 68-1) (providing that the eight illustrative plans "comply with traditional redistricting principles, including one-person one-vote").

The Supreme Court has not extended the one-person, one-vote principle of Reynolds v. Sims, 377 U.S. 533 (1964), to the judicial branch of government. See Chisom, 501 U.S. at 389 (involving a challenge to the districting scheme of the Louisiana Supreme Court and quoting a three-judge district court's holding that "'the concept of one-man, one-vote apportionment does not apply to the judicial branch of the government'" (quoting Wells v. Edwards, 347 F. Supp. 453 (M.D. La. 1972) (three-judge district court), summarily aff'd, 409 U.S. 1095 (1973)))²⁰; see id. at 415 (Scalia, J., dissenting) ("Well before Congress amended § 2, we had held that the principle of 'one person, one vote' does not apply to the election of judges." (citing Wells, 347 F. Supp. at 454)); Wymbs v. Republican State Exec. Comm. of Fla., 719 F.2d 1072, 1087 n.40 (11th Cir. 1983) (observing that "the one person, one vote concept is limited primarily to cases involving legislatures or bodies performing legislative functions" and "does not, for example, apply to the election of state judges" (citing Wells, 347 F. Supp. at 454)); Voter Info. Project, Inc. v. City of Baton Rouge, 612 F.2d 208, 211 (5th Cir. 1980) ("[W]e are bound with respect to this specific issue [that the one-person, one-vote principle does not apply to elected judges] by the Supreme Court's affirmance in Wells."). The Court's continued rejection of equipopulation principles to judicial districts reflects the unique nature of the judicial role as compared to that of other elected office holders. Judicial districting is a square peg that does not fit neatly into the round hole of districting jurisprudence.

20 Notwithstanding the limitations on the precedential value of summary affirmances like Wells, which was decided prior to the 1988 statutory end to the Supreme Court's mandatory appeals jurisdiction, the Chisom Court in 1991 relied upon Wells's holding. See Supreme Court Case Selections Act of 1988, Pub. L. No. 100-352, 102 Stat. 662.

While there might be reasons peculiar to Alabama for configuring appellate judicial districts of roughly equal population, Mr. Cooper does not articulate any. He did not consider whether substantial population equality for the appellate judicial districts is a desirable goal in the formulation of the plans, rather than a required goal.²¹ Trial Tr. II, at 151-52 (Cooper); Pls.' Proposed Findings of Fact, at 56, ¶ 8 (Doc. # 176) ("The plans . . . satisfy the one-person, one-vote criteria." (emphasis added)). Because Mr. Cooper's expertise in districting overwhelmingly is in the areas of state and local legislative districting plans, Trial Tr. II, at 147-48 (Cooper), not judicial districts, the misconception is understandable, and, in his defense, he was likely following client instructions. At bottom, though, Mr. Cooper's adherence to the principle of one-person, one-vote, without explanation for doing so, is problematic for the proposed districts of the courts of appeals.

21 State district judges and probate judges are elected from and generally preside over cases from a single county. State circuit judges are elected from circuits of varying populations; hence, Alabama's forty-one judicial circuits are not equally populous. Trial Tr. II, at 152 (Cooper). But Mr. Cooper testified that, for the appellate courts, he created two SC plans and one AC plan with a population deviation of plus or minus 1% based on the Alabama Legislature's 2011 "plus or minus 1%" guideline for population deviations in congressional, legislative, and state board of education districts. Cooper Expert Report, at 17 (Pls.' Trial Ex. 86; Doc. # 68-1); Trial Tr. II, at 151-52 (Cooper); see also Defs.' Post-Trial Br., at 17 (Doc. # 168) (arguing that the SC and AC plans with plus-or-minus 5% deviations violated the Alabama Legislature's guideline (citing Ala. Legislative Black Caucus v. Alabama, 231 F. Supp. 3d 1026, 1037 (M.D. Ala. 2017) ("The Republican-controlled legislature used the same criteria as previous legislatures, with one exception: they tightened the limit on population deviation to ±1%."))). But that guideline is grounded in the constitutional principle of one-person, one-vote that does not apply to judicial districting. Moreover, the state legislature's guideline did not address or contemplate judicial districting. In any event, the guideline has been criticized. See Ala. Legislative Black Caucus v. Alabama, 135 S. Ct. 1257, 1263 (2015) (The State of Alabama's "goal [of] creating a set of districts in which no district would deviate from the theoretical, precisely equal ideal by more than 1% — i.e., a more rigorous deviation standard than our precedents have found necessary under the Constitution.").

Communities of interest. A shared community of interest cannot be assumed. See Miller, 515 U.S. at 919 ("[M]ere recitation of purported communities of interest" will not suffice.). The portion of Jefferson County in District 1, which incorporates the city of Birmingham, is mostly urban, but the Black Belt region is decidedly rural, with the exception of the city of Montgomery in Montgomery County, which is also urban. Although the cities of Birmingham and Montgomery are urban, they are separated by ninety miles and three counties.

Mr. Cooper did not discuss the regional, cultural, social, economic, or political ties, if any, among the African-American communities in Birmingham, in the Black Belt, or in any other area in District 1 of the AC plans. Nor is he likely familiar with those community characteristics. Mr. Cooper admits that he is "not from Alabama," and the only illustration he gave as to his "effort to keep areas together" was the obvious avoidance of creating a district that extended from the Appalachians in the north to the Alabama-Florida border in the south. Trial Tr. II, at 154 (Cooper). Mr. Cooper testified that he joined the substantial African-American population of Jefferson County for the sheer sake of satisfying a numerical threshold. Trial Tr. II, at 186-87 (Cooper). His explanation reveals that maintaining communities of interest, if considered at all, was subordinated to the necessity of creating an African-American, voting-age population in District 1 of his AC plans with a 50% plus one voting-age population. This feature illustrates the warning of Shaw v. Reno. See 509 U.S. at 647 (incorporating "in one district individuals who belong to the same race, but who are otherwise widely separated by geographical and political boundaries, and who may have little in common with one another but the color of their skin, bears an uncomfortable resemblance to political apartheid. It reinforces the perception that members of the same racial group — regardless of their age, education, economic status, or the community in which they live — think alike, share the same political interests, and will prefer the same candidates at the polls. We have rejected such perceptions elsewhere as impermissible racial stereotypes."). The physical dispersal of the African-American population in District 1 of the AC plans is an indicator of non-compactness.

Contiguity. There is no dispute that the districts in the AC plans, including District 1, are contiguous. Mr. Cooper explained that the software program he uses (Maptitude) has "an automated module that checks for contiguity." Trial Tr. II, at 155 (Cooper). The contiguity principle is easily satisfied.

Incumbency protection. Mr. Cooper did not consider incumbency protection when he drew his plans. He testified that he did not have the mailing addresses for the sitting appellate judges, and, thus, he did not know if his plans would require any incumbents to run against each other. Trial Tr. II, at 188-89 (Cooper); Cooper Expert Report, at 31 (Pls.' Trial Ex. 86; Doc. # 68-1). But Dr. Bonneau confirmed that "moving to district-based elections would throw the courts into turmoil" because "[t]here would almost certainly be multiple incumbents running against each other in the initial elections after a move to district-based elections." Bonneau Expert Report, at 5, ¶ 9, at 19, ¶ 52 (Defs.' Trial Ex. 24; Doc. # 85-1); see also Trial Tr. VI, at 53-54 (Murdock) (testifying about the importance of collegiality on a nine-justice court and the likelihood that sitting supreme court justices would have to run against another in a districting plan).

Findings as to AC Plans. District 1 in the AC plans satisfies the numerosity benchmark of the first Gingles precondition, but Plaintiffs have not proven that the African-American population in District 1 is sufficiently compact. District 1 carves out Jefferson County's substantial African-American population and connects it, just barely, to District 1 by way of Bibb County, so as to create an oddly shaped appendage. There is scant evidence addressing this geographical oddity, other than the testimony that it must be appended to create a majority-black district. Moreover, the plans were formulated based upon a misconception concerning population deviations among districts. There also is no evidence that the AC plans will protect incumbent judges. All factors considered, Plaintiffs failed to show that the AC plans were designed "consistent with traditional districting principles." Davis, 139 F.3d at 1425.²²

22 The State also argues, defensively, that District 1 in the AC plans is "racially gerrymandered" and that, therefore, the State would be unable to defend the district against an equal protection challenge. Defs.' Post-Trial Br., at 19 (Doc. # 168); see also Jefferson County Splits (Defs.' Trial Ex. 89). Defendants cite no authority that, to satisfy the first Gingles precondition, a § 2 plaintiff must prove that the hypothetical plan satisfies the standards in Miller v. Johnson, 515 U.S. 900 (1995) (addressing Fourteenth Amendment racial gerrymandering claims). See Davis, 139 F.3d at 1425 ("The district court's attempt to apply authorities such as Miller to this Section Two case, however, is unpersuasive, because the Miller and Gingles/Nipper/SCLC lines address very different contexts."); Ga. State Conf. of NAACP v. Fayette Cty. Bd., 952 F. Supp. 2d 1360, 1365 (N.D. Ga. 2013) (finding no cases that had incorporated into the first Gingles precondition a requirement that a § 2 plaintiff "show compliance with Miller"). However, in light of the court's finding above, it is unnecessary to further probe the State's racial-gerrymandering argument.

c. The Second and Third Gingles Preconditions

Racial bloc voting, also termed racially polarized voting, is the linchpin of a § 2 vote dilution claim and the embodiment of the second and third Gingles preconditions. Gingles, 478 U.S. at 52 n.18; see also Stallings, 829 F.2d at 1550 ("Gingles's three-part test establishes that racial bloc voting is the hallmark of a vote dilution claim."). "The purpose of inquiring into the existence of racially polarized voting is twofold: to ascertain whether minority group members constitute a politically cohesive unit and to determine whether whites vote sufficiently as a bloc usually to defeat the minority's preferred candidates." Gingles, 478 U.S. at 56.

Under the second Gingles precondition, "[a] showing that a significant number of minority group members usually vote for the same candidates . . . prov[es] the political cohesiveness necessary to a vote dilution claim, and, consequently, establishes minority bloc voting within the context of § 2."²³ Id. (internal citation omitted). There is no bright-line numeric cutoff for a finding of political cohesion, but the second Gingles precondition has not "present[ed] much of a barrier to the assertion of vote dilution claims on behalf of any racial group." Holder, 512 U.S. at 903-04 (Thomas, J., concurring in the judgment). However, if political cohesion does not exist in the voting patterns of the minority group, then the defeat of the minority group's candidate cannot be blamed on the at-large electoral system. Gingles, 478 U.S. at 51.

23 "Political cohesion" and "bloc voting" are different ways of saying the same thing, although the negative connotation is greater as to one.

Under the third Gingles precondition, white bloc voting is "legally significant" if "whites vote sufficiently as a bloc usually to defeat the minority's preferred candidates." Id. at 56. The minority's preferred candidates can be of any race. See supra note 3. In Gingles, Justice Brennan's plurality opinion explains that, although there is often a correlation between the voter's race and the candidate's race, "[u]nder § 2, it is the status of the candidate as the chosen representative of a particular racial group, not the race of the candidate that is important." Id. at 68 (plurality opinion) (emphasis in original). The Eleventh Circuit has cited this part of the Gingles plurality opinion favorably. See Stallings, 829 F.2d at 1557 & n.12 (stating that "[u]nder Section 2, it is the status of the candidate as the chosen representative of a particular racial group, not the race of the candidate that is important," but noting that "[o]nly a plurality of the court agreed with Justice Brennan's analysis" on this point (citing Gingles, 478 U.S. at 69-70)).

Applying varying standards, other federal appeals courts have recognized that the minority-preferred candidate need not match the voter's race. See Jenkins v. Red Clay Consol. Sch. Dist. Bd. of Educ., 4 F.3d 1103, 1125-26 (3d Cir. 1993) (canvassing the federal courts of appeals' different approaches for deciding which candidate the minority group prefers and adopting an approach under which "not all minority candidates will be minority-preferred" and which "look[s] beyond whether a white candidate received a majority of the minority community's vote [to] . . . determine whether that white candidate is truly the minority community's representative of choice"); see also Lopez, 339 F. Supp. 3d at 609 ("The preferred candidate need not be a member of the minority group." (citing E. Jefferson Coal. for Leadership & Dev. v. Par. of Jefferson, 926 F.2d 487, 493 (5th Cir. 1991)).

"[I]n general, a white bloc vote that normally will defeat the combined strength of minority support plus white 'crossover' votes rises to the level of legally significant white bloc voting." Gingles, 478 U.S. at 56. "[I]n the absence of significant white bloc voting it cannot be said that the ability of minority voters to elect their chosen representatives is inferior to that of white voters." Id. at 48 n.15. Discussing the third Gingles precondition, the Eleventh Circuit has elaborated that plaintiffs "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) (internal quotation marks and citation omitted).

i. Proof of the Second and Third Gingles Preconditions: Generally

Both the second and third Gingles preconditions require analyses of bloc voting — one focuses on minority bloc voting and the other on majority bloc voting. These analyses "require[] discrete inquiries into minority and white voting practices." Gingles, 478 U.S. at 56. Bloc voting cannot be presumed: "[P]laintiffs must prove it." Id. at 46. So how do they prove it?

"Ordinarily, direct evidence of minority voting patterns is unattainable, since ballots do not indicate a voter's race." Solomon, 899 F.2d at 1019 (Kravitch, J., concurring). Where that is the case, as here, Trial Tr. II, at 72 (Handley), plaintiffs must rely on expert testimony and statistical evidence to assess the voting patterns of minority and white voters. See Solomon, 899 F.2d at 1019-20 (Kravitch, J., concurring) (discussing established statistical methods for proving bloc voting). Anecdotal evidence also can be helpful, but to a lesser extent than statistical evidence. See Growe, 507 U.S. at 41-42 (holding that the absence of statistical or anecdotal evidence of racial bloc voting under Gingles was fatal to the plaintiffs' § 2 challenge to a state redistricting plan).

For the statistical analyses, both the number and types of elections subject to testing are important considerations. "[T]he number of elections that must be studied in order to determine whether voting is polarized will vary," for example, depending upon the "number of elections in which the minority group has sponsored candidates." Gingles, 478 U.S. at 57 n.25. There must be enough elections, though, from which to discern a voting pattern because the "mere loss of an occasional election" is insufficient proof of vote dilution. Id. at 51 ("[T]he usual predictability of the majority's success distinguishes structural dilution from the mere loss of an occasional election.").

Additionally, the types of elections that are probative of voter behavior are broad but again, depending upon the circumstances, are not necessarily of equal weight. In the hierarchy of probative elections for proving minority cohesion and white bloc voting, the Eleventh Circuit has favored a trial court's focus on endogenous elections — contests for the particular office that is at issue. See Askew, 127 F.3d at 1381 n.13. In one case, which was a § 2 challenge by registered voters to the city of Gainesville, Florida's at-large method of electing city council members, the Eleventh Circuit held that the trial court "did not clearly err by viewing endogenous elections as more probative than exogenous elections" as this view "is the law of this Circuit." Hamrick, 196 F.3d at 1222 (citing Stallings, 829 F.2d at 1560). The endogenous elections were "those which pertain[ed] to Gainesville City Council elections." Johnson v. Hamrick, 155 F. Supp. 2d 1355, 1370 n.11 (N.D. Ga. 2001), aff'd, 296 F.3d 1065 (11th Cir. 2002). And in Stallings, because the electoral jurisdiction was Carroll County, the Eleventh Circuit held that "[t]he district court's reliance on municipal elections in Carroll County as proof of minority electoral success of a county electoral scheme [was] obviously misplaced." 829 F.2d at 1560 (emphases added). At issue in Hamrick and Stallings were elections within a political subdivision of a state, not the state at large, as here.

However, when it comes to the race of the candidates, the Eleventh Circuit has not assigned more weight to elections involving African-American candidates than to elections involving candidates of other races. In Hamrick, the court merely held that, based on circuit precedent that "indicat[ed] elections involving black candidates may be given more weight in determining vote dilution," the district court "did not clearly err by giving more weight to elections involving black candidates." 196 F.3d at 1221 (first citing Davis, 139 F.3d at 1417 n.5; and then citing SCLC, 56 F.3d at 1293); see SCLC, 56 F.3d at 1293, 1294 (holding that there was "no cause to set aside the district court's finding" where the district court "concluded that greater weight should be given to . . . white on black judicial elections, but . . . also consider[ed] political races generally — white on black and white on white — in seeking to determine the question of whether the political process is open to minority voters" (alterations added) (citation and internal quotation marks omitted). At the same time, the court stressed that it was not "imply[ing] [that] district courts should give elections involving black candidates more weight." Hamrick, 196 F.3d at 1221 (emphasis in original). Indeed, crediting only elections that have an African-American candidate would ignore Eleventh Circuit precedent that "the at-large scheme may cause the Black candidate to be someone other than the preferred candidate of Black voters. That a candidate is of the race of the group is not dispositive; the legal standard is whether the candidate is the preferred candidate of the minority group." Meek v. Metro. Dade Cty., 908 F.2d 1540, 1548 (11th Cir. 1990).

Additionally, the Eleventh Circuit's SCLC decision aptly illustrates the broad swath of relevant elections in the § 2 analysis. There, the Eleventh Circuit commended the district court for rejecting the invitation of the experts to "ignore relevant [elections] evidence" and countenanced the court's approach of considering "any evidence that explained election results." 56 F.3d at 1293. SCLC aligns with the well-established § 2 principle that "the resolution of a voting dilution claim requires close analysis of unusually complex factual patterns." Hamrick, 196 F.3d at 1223 (quoting Cross v. Baxter, 604 F.2d 875, 879 (5th Cir. 1979)). Accordingly, the probativeness of an election in the § 2 analysis is highly dependent upon the individual facts and the nature of the local electorate. As summarized in Solomon,

In examining a jurisdiction's electoral history, a court may assign more probative value to elections that include minority candidates, than elections with only white candidates; it may consider endogenous elections more important than exogenous elections; or it may conclude that recent elections are more probative than those which occurred long before the litigation began.

221 F.3d at 1227 (internal quotation marks and citations omitted).

ii. Plaintiffs' Statistical Evidence of Racially Polarized Voting

To show racially polarized voting, Plaintiffs rely principally upon their expert, Lisa Handley, who holds a Ph.D. in political science, has thirty years' experience in the area of voting rights, and works as an election consultant in the United States and internationally. Trial Tr. II, at 66 (Handley). At trial, she was tendered as an expert, without objection, on the topic of racially polarized voting, and her Amended Expert Report was admitted without objection. See Handley Am. Expert Report (Pls.' Trial Ex. 121; (Doc. # 78-1); Trial Tr. II, at 68. However, as will be explained, the parameters for the elections she chose — only statewide elections with a black candidate running against a white candidate — exclude other relevant elections, thereby diminishing the credibility of her conclusions. It is necessary to set out with some detail Dr. Handley's analyses.

Dr. Handley analyzes a subset of elections, occurring between 2000 and 2014, that had an African-American candidate on the ballot. The elections total thirteen: eight general elections for statewide appellate judicial races, occurring between 2000 and 2008; two Democratic primaries for statewide appellate judicial races, one occurring in 2000 and the other in 2008; and three general elections in 2014 for statewide non-judicial races (lieutenant governor, secretary of state, and state auditor). She relies upon data available from the Alabama Secretary of State's website, which consists of data for county-level voter registration by race matched with county election results.²⁴ She also obtained data from the Jefferson County probate court's website and through a public records request. This data consisted of voter registration by race at the precinct-level in Jefferson County matched with election results from that county.²⁵ Handley Am. Expert Report, at 5-6 (Pls.' Trial Ex. 121; Doc. # 78-1)

24 Voter turnout data by race is preferred over registration data by race. Handley Am. Expert Report, at 5 (Pls.' Trial Ex. 121; (Doc. # 78-1). However, according to Dr. Handley, only two states maintain voter turnout data by race, and Alabama is not one of them. Trial Tr. II, at 134 (Handley).

25 Dr. Handley says that she selected the year 2000 as the floor because that was the earliest data available on the Alabama Secretary of State's website and that data, in her opinion, "provided a sufficient number of contests." Trial Tr. II, at 74 (Handley). Her latter opinion as to the magic number of elections needed to decide racially polarized voting goes to the ultimate issue of fact (i.e., whether there is vote dilution). While an expert's "opinion is not objectionable just because it embraces an ultimate issue," Fed. R. Evid. 704, the final determination of vote dilution or not is for the court.

What Dr. Handley excludes from her analysis — entirely — is endogenous evidence of a key victory by a black-preferred candidate in the 2006 race for Chief Justice and three historic victories by black incumbent justices over white challengers in 1986, 1992, and 1994. She also did not analyze the exogenous 2010 Democratic primary for governor in which the Alabama Democratic Conference ("ADC") supported the white candidate, Ron Sparks, over the black candidate, Artur Davis. While these omissions do not affect the conclusion of racially polarized voting in the Alabama elections between black and white candidates for appellate judgeships occurring between 2000 and 2008, they exaggerate the extent of polarization and render the data unreliable for other experts to rely upon for opinions on the cause of defeat of black-preferred candidates. Furthermore, the restricted electoral parameters generate outcomes that do not adequately account for the complexities within the Alabama Democratic Party that factor into whether a particular candidate is the black-preferred candidate. See, e.g., Trial Tr. IV, at 144-46 (Gray) (discussing the Alabama Democratic Conference's endorsements of Hillary Clinton over Barack Obama for the Democratic presidential nominee in 2008, of Walter Mondale over Jesse Jackson for the Democratic presidential nominee in 1984, and of Ron Sparks over Artur Davis for the Democratic governor nominee in 2010).²⁶ As will be demonstrated later, Dr. Handley failed to capture important evidence that explains election results vis-à-vis "unusually complex factual patterns," Hamrick, 196 F.3d at 1223 (citation and internal quotation marks omitted).

26 The Alabama Democratic Conference ("ADC") is the statewide African-American political caucus with significant influence over the affairs of the Alabama Democratic Party. Trial Tr. IV, at 140 (Gray); see also White, 74 F.3d at 1071 n.43 (noting that the ADC is "a statewide black political organization that is an arm of the Alabama Democratic Party").

Using this limited data, Dr. Handley employs three court-accepted statistical techniques: homogeneous precinct analysis; ecological regression ("ER"); and ecological inference ("EI").²⁷ A brief summary of these techniques is helpful.

27 The Supreme Court relied upon the first two techniques in Gingles. See 478 U.S. at 53 n.20 (noting the district court's acceptance of ecological regression and homogeneous precinct analysis as "standard in the literature"). The third technique, created after Gingles, has been recognized as the "gold standard." Wright v. Sumter Cty. Bd. of Elections & Registration, 301 F. Supp. 3d 1297, 1305 (M.D. Ga. 2018) ("EI method is currently the 'gold standard' for use in racial bloc voting analyses.").

The homogeneous precinct analysis requires a geographical unit where 90% or more of the voters are of a single race. Because no county in Alabama met this requirement, Dr. Handley confines her homogeneous precinct analysis to the precincts in Jefferson County, Alabama's largest county by population. About half of the precincts met the 90% criterium. Trial Tr. II, at 79-80 (Handley). In these precincts, Dr. Handley calculates the percentages of African-American votes cast for candidates and then used these percentages as estimates of voting behavior at the county and state levels. She applies the homogeneous precinct analysis to the general and primary elections for appellate judicial contests (i.e., ten of the thirteen elections).

Dr. Handley's ecological regression analysis uses a two-equation method that correlates election results and the race of registered voters to estimate the percentages of black voters and white voters who support particular candidates. The first equation "estimates the percentage of the black and white registered voters who voted for a given candidate[,] and the second estimates the percentage of the black and white registered voters [who] did not vote for the given candidate — either because they did not vote or because they voted for other candidates."²⁸ Handley Am. Expert Report, at 9 (Pls.' Trial Ex. 121; Doc. # 78-1). The results of the two steps are "combined in order to correct for any differential in the voter turnout rates of blacks and whites." Handley Am. Expert Report, at 9 (Pls.' Trial Ex. 121; Doc. # 78-1); Trial Tr. II, at 87-89 (Handley). As part of the ER analysis, in addition to calculating estimated percentages, Dr. Handley examines the strength of the relationship between race and votes cast with a correlation coefficient. The correlation coefficient, which measures the degree of linear association between these variables, ranges from 0 to 1. The closer the correlation is to 1, the stronger the relationship is between race and votes cast. Conversely, the closer the correlation is to 0, the weaker the relationship is between race and votes cast. Handley Am. Expert Report, at 9; Trial Tr. II, at 81-89 (Handley). For the statewide appellate judicial contests, Dr. Handley provides both estimated percentages and correlation coefficients. Handley Am. Expert Report (Tables 1-5; Appendix A).

28 The scatterplot in Dr. Handley's amended expert report illustrates the first equation. Handley Am. Expert Report, at 8 (Figure 1); Trial Tr. II, at 87-88 (Handley).

The third technique, ecological inference, was developed by political scientist Gary King subsequent to the Gingles decision, to improve upon the ecological regression method. Handley Am. Expert Report, at 10 (Pls.' Trial Ex. 121; Doc. # 78-1). Ecological inference is similar to ecological regression but has two notable differences. Ecological inference does not rely on the assumption of linearity underpinning the ecological regression method. Instead, it incorporates the "method of bounds" and "maximum likelihood statistics to produce estimates of voting patterns by race." Handley Am. Expert Report, at 10. The bounds method "provid[es] more information about the voting behavior being estimated" and constrains estimates to real limits by not allowing estimates of the percentage of votes cast for a candidate to exceed 100 or fall below zero. Handley Am. Expert Report, at 10. As an example of how the method of bounds works, Dr. Handley provides an illustration where a precinct has 100 voters, of which 75 are black and 25 are white, and the black candidate receives 80 votes. In this hypothetical, at least 55 of the black voters (80 minus 25) voted for the black candidate and at most all 75 did. These bounds are not used in ecological regression. Handley Am. Expert Report, at 10 & n.16.

Summary of the Results of Dr. Handley's Statistical Analyses. Dr. Handley's opinion is that, while there are some statistical variances among the three techniques, each technique leaves "no question in any of the contests who black voters preferred and who white voters preferred," Trial Tr. II, at 104 (Handley), and reveals "a very clear pattern of racially polarized voting in recent statewide judicial elections that included African American candidates." Handley Am. Expert Report, at 10 (Pls.' Trial Ex. 121; Doc. # 78-1).

In her report, Dr. Handley summarizes her findings on racially polarized voting based principally on the ER and EI percentage estimates derived from statewide data (and not based on the ER, EI, and homogeneous precinct percentage estimates derived from the Jefferson County precinct data).²⁹ Dr. Handley reports the results of the technique — either ER or EI — that yields the lowest percentage of black voters who voted for the black candidate and the lowest percentage of white voters who voted for the white candidate. Considering the lowest percentages, Dr. Handley found a pattern of racially polarized voting. This opinion also refers to the lowest ("at least") percentages.

29 Again, Dr. Handley did not perform the homogeneous precinct analysis on the statewide data obtained from the Alabama Secretary of State's website because she said no county met the benchmark of being 90% homogeneous. She conducted the homogeneous precinct analysis using the Jefferson County precinct data only on the statewide general and primary elections for appellate judges. She did not have Jefferson County precinct data for the three exogenous elections in 2014. Trial Tr. II, at 130 (Handley).

The eight general elections for statewide appellate judicial races. Here is what Dr. Handley found, having analyzed only the eight statewide general elections for appellate judges, occurring between 2000 and 2008, that had an African-American candidate on the ballot. The percentage of black voters who voted for the black candidate ranged from at least 78.3% to at least 89.6%; yet each black candidate lost his or her judicial race to the white candidate, who received between at least 62.6% and at least 73.1% of the white vote. Additionally, Dr. Handley reports that the correlation coefficients between the proportion of African-American registered voters and the proportion of votes cast for the African-American candidate in these eight elections are high, ranging from .778 to .915.

To break it down, in the three 2000 races involving an African-American candidate, at least 83.9% of the African-American voters voted for Ralph Cook for Associate Justice of the Supreme Court, Place 1; at least 84.1% of the African- American voters voted for John England, Jr. for Associate Justice of the Supreme Court, Place 3; and at least 82% of the African-American voters voted for Aubrey Ford for Court of Criminal Appeals, Place 2. However, Cook, England, and Ford (who are black) lost to their white opponents (Lyn Stuart, Tom Woodall, and Kelli Wise, respectively), with Stuart receiving at least 62.6% of the white votes; Woodall, at least 64.8% of the white votes; and Wise, at least 66% of the white votes. Handley Am. Expert Report (Table 1).

In the 2002 race for Court of Civil Appeals,³⁰ at least 78.3% of black voters voted for Vicky Underwood Toles, who is black, while at least 68% of the white voters voted for her white opponent, Bill Thompson.³¹ Handley Am. Expert Report (Table 2).

30 On page 13 of her Amended Expert Report, Dr. Handley erroneously states that Toles "ran for the Court of Criminal Appeals."

31 Because there was only one judicial race in 2002 for a seat on the Court of Civil Appeals, there was no judicial place number used that year. See Notes on Place Numbers for Judicial Candidates, at 1 (Defs.' Trial Ex. 103; Doc. # 151-1).

In the three 2006 races, at least 83% of the African-American voters supported Gwendolyn T. Kennedy for Associate Justice of the Supreme Court, Place 2; at least 86.1% of the African-American voters supported John England, Jr. for Associate Justice of Supreme Court, Place 4; and at least 82.1% of the African-American voters supported Aubrey Ford for Court of Criminal Appeals, Place 1. However, Kennedy, England, and Ford (who are black) lost to their white opponents (Tom Woodall, Glen Murdock, and Greg Shaw, respectively), who received white votes of at least 67.6% (Woodall), 66% (Murdock), and 65.3% (Shaw). Handley Am. Expert Report (Table 3).

In the 2008 race for Court of Criminal Appeals, Place 1, Clyde Jones, who is black, received at least 89.6% of the black vote, but he lost to his white opponent (Beth Kellum), who received at least 73.1% of the white vote.

Dr. Handley opines that there was racial bloc voting in all eight of these general elections for appellate judges.

The two Democratic primaries for statewide appellate judicial races. Dr. Handley also examines two Democratic primary appellate judicial elections, one in 2000 and the other in 2008. She finds racially polarized voting in the 2008 election, but not in the 2000 election. In the 2000 Democratic primary, Yvonne Saxon, who is black, ran for Court of Criminal Appeals, Place 3, against Sue Bell Cobb, who is white. Cobb was the black-preferred candidate. She received 63.9% to 64.1% of the African-American vote, while Saxon received only 35.3% to 35.9% of the African-American vote. See Trial Tr. II, at 125 (Handley) (confirming that Cobb was "the choice of black voters" in the 2000 Democratic primary election against Saxon for the Court of Criminal Appeals). Cobb won the Democratic nomination (and the general election).

In the 2008 Democratic primary, Clyde Jones, who is black, ran for Court of Criminal Appeals, Place 1, against Steve Dodd, who is white.³² Jones won the primary with at least 74.8% of the black vote and at least 16.6% of the white vote, while his opponent, Dodd, lost with at least 23.3% of the black vote and at least 61.1% of the white vote. Dr. Handley concludes, though, that white voters were not able to vote as a bloc and defeat Jones in the primary, as they were able to do in the general election, because only 3% to 6% of white registered voters cast ballots in that Democratic primary. Handley Am. Expert Report (Table 5).

32 At trial, Dr. Handley corrected an error in her amended report concerning this election: In the 2008 Democratic primary for Court of Criminal Appeals, Place 1, Clyde Jones's opponent was Steve Dodd, not Beth Kellum. Trial Tr. II, at 135 (Handley); Handley Am. Expert Report, at 26, Appendix A (Pls.' Trial Ex. 121; Doc. # 78-1).

The three general elections in 2014 for statewide non-judicial races .Because an African American had not run for statewide judicial office since 2008, Dr. Handley analyzed the 2014 statewide general elections in which an African American ran for lieutenant governor, secretary of state, and auditor. In each race, the black candidate received a substantial portion of the black vote — at least 84.9% for James Fields (for lieutenant governor); at least 91.2% for Lula Albert-Kaigler (for secretary of state); and at least 89.8% for Miranda Joseph (for auditor) — but lost to the white opponent, each of whom received a substantial portion of the white vote — at least 82.3% for Kay Ivey (for governor); at least 85.3% for John Merrill (for secretary of state); and at least 83.3% for Jim Zeigler (for auditor). Handley Am. Expert Report (Table 6).

The following chart summarizes the results of Dr. Handley's statistical analyses of the thirteen elections.

Year	Contest	Judicial Office	Name &Race of theBlack-PreferredCandidateB = BlackW = White	Lowest % ofBlack Votesfor the Black-PreferredCandidate33	ERCorrelationCoefficients	Name &Race of theWinningCandidate	Lowest % ofWhite Votesfor theWinningCandidate34
2000	General	AssociateJustice ofSupremeCourt,Place 1	RalphCook(B)	83.9	.807	Lyn Stuart(W)	62.6
2000	General	AssociateJustice ofSupremeCourt,Place 3	JohnEngland,Jr.(B)	84.1	.831	TomWoodall(W)	64.8
2000	General	Court ofCriminalAppeals,Place 2	AubreyFord(B)	82.0	.798	Kelli Wise(W)	66.0
2002	General	Court of CivilAppeals	VickyUnderwoodToles(B)	78.3	.814	BillThompson(W)	68.0

Year	Contest	Judicial Office	Name &Race of theBlack-PreferredCandidateB = BlackW = White	Lowest % ofBlack Votesfor the Black-PreferredCandidate	ERCorrelationCoefficients	Name &Race of theWinningCandidate	Lowest % ofWhite Votesfor theWinningCandidate
2006	General	AssociateJustice ofSupremeCourt,Place 2	GwendolynT. Kennedy(B)	83.0	.828	TomWoodall(W)	67.6
2006	General	AssociateJustice ofSupremeCourt, Place 4	JohnEngland,Jr.(B)	86.1	.807	GlennMurdock(W)	66.0
2006	General	Court ofCriminalAppeals,Place 1	AubreyFord(B)	82.1	.778	GregShaw(W)	65.3
2008	General	Court ofCriminalAppeals,Place 1	ClydeJones(B)	89.6	.915	BethKellum(W)	73.1
2000	Democratic Primary	Court ofCriminalAppeals,Place 3	Sue BellCobb(W)	63.9	-.157	Sue BellCobb	72.4
2008	Democratic Primary	Court ofCriminalAppeals,Place 1	ClydeJones(B)	74.8	.708	ClydeJones	16.6
2014	General	LieutenantGovernor	JamesFields(B)	84.9	N/A	Kay Ivey(W)	82.3
2014	General	Secretary ofState	LulaAlbert-Kaigler(B)	91.2	N/A	JohnMerrill(W)	85.3
2014	General	Auditor	MirandaJoseph(B)	89.8	N/A	JimZeigler(W)	83.3

33 Consistent with Dr. Handley's approach, the lowest percentage of the ER and EI methods using statewide data is the percentage reflected in the fifth and eighth columns. Handley Am. Expert Report (Tables 1-6). This chart also includes the ER estimates with correlation coefficients. Handley Am. Expert Report, at 25-28 (App. A) (Pls.' Trial Ex. 121; Doc. # 78-1).

34 The "backside" of each of the numbers in this column is that, in all the general election races analyzed, the black candidate received between 27.9% and 37.4% of white votes consistently. Post-Trial Stipulations, at 48, ¶ 234 (Doc. # 176) ("According to Dr. Handley's analysis, around 30% of white voters supported the candidates of choice for African-American voters in all of the General Elections for judicial office that Handley analyzed, except for Clyde Jones, and he received about a quarter of the white vote."). Had Dr. Handley analyzed all the relevant races, the white vote for the black-preferred candidate would have been much higher.

33 Consistent with Dr. Handley's approach, the lowest percentage of the ER and EI methods using statewide data is the percentage reflected in the fifth and eighth columns. Handley Am. Expert Report (Tables 1-6). This chart also includes the ER estimates with correlation coefficients. Handley Am. Expert Report, at 25-28 (App. A) (Pls.' Trial Ex. 121; Doc. # 78-1).

34 The "backside" of each of the numbers in this column is that, in all the general election races analyzed, the black candidate received between 27.9% and 37.4% of white votes consistently. Post-Trial Stipulations, at 48, ¶ 234 (Doc. # 176) ("According to Dr. Handley's analysis, around 30% of white voters supported the candidates of choice for African-American voters in all of the General Elections for judicial office that Handley analyzed, except for Clyde Jones, and he received about a quarter of the white vote."). Had Dr. Handley analyzed all the relevant races, the white vote for the black-preferred candidate would have been much higher.

iii. Findings of Fact and Conclusions of Law: Racially Polarized Voting

Plaintiffs' evidence on racially polarized voting has noticeable weaknesses. Dr. Handley chose racially polarized voting in eight, hand-picked endogenous elections for a narrow period of time from 2000 to 2008. Oddly, Dr. Handley did not analyze the three statewide judicial races in 1982, 1988, or 1994, in which the African-American candidates won (i.e., Justice Adams and Justice Cook). She instead starts her analysis in 2000, the year Justice Cook and Justice England lost their races for the Alabama Supreme Court.³⁵ Justice Adams's and Justice Cook's candidacies yielded one opposed Democratic primary, one opposed Democratic primary runoff, and two opposed general elections, all with white opponents, that Dr. Handley could have analyzed. See SCLC, 56 F.3d at 1288 ("Justice Adams won reelection in statewide races against well-known white opponents in the 1982 Democratic primary, the 1982 primary runoff, and the 1988 general election."); Ala. Statewide Elections from 1970 to Present, at 13, 17, 22 (Defs.' Trial Ex. 103; Doc. # 151-1); Alabama Sec'y of State, Elections Data Downloads, Election Results Archive—State Supreme Court, https://www.sos.alabama.gov/alabama-votes/voter/election-data (last visited Feb. 5, 2020).³⁶ It is difficult to understand how the age of the only appellate court elections where the black-preferred candidates won renders them verboten. Dr. Handley's explanation for excluding these races is arbitrary and unpersuasive.³⁷ See Bonneau Expert Report, at 12, ¶ 29 (Defs.' Trial Ex. 24; Doc. # 85-1) (opining that focusing solely on races that include African-American candidates to determine racially polarized voting "leads to selection bias and potentially erroneous conclusions" and that "we need to look at how African-Americans in Alabama vote in all races, not just those where there are African-American candidates").

35 Plaintiffs' expert historian, Peyton McCrary, Ph.D., suggests there was not racially polarized voting during this time frame. Trial Tr. III, at 8-9 (McCrary) (testifying that in the 1980s and 1990s when Justices Oscar Adams and Ralph Cook were elected to the Alabama Supreme Court, racially polarized voting was "not legally significant in the sense" that racially polarized voting did not "result[] in the defeat of minority candidates for the appellate elections"). Translated, Justices Adams and Cook were elected by the same white "bloc" vote as defeated Justice Cook and Justice England just six years after Justice Cook was elected.

36 Of note, the court has had no difficulty accessing election data from the Alabama Secretary of State's website for elections prior to 2000.

37 Had there been a fundamental, structural change in the system for electing judges in Alabama as occurred in Florida (Nipper), the change might have rendered pre-2000 Alabama appellate judicial races non-probative on the issue of vote dilution. See Nipper, 39 F.3d at 1538 & n.83 (plurality opinion) (noting that the experts began their statistical analyses with Florida's judicial elections in 1972, the year Florida switched from partisan to non-partisan elections and implemented a gubernatorial appointment system for judicial vacancies; "[t]he implication [was] that elections that took place under the former system were sufficiently different so as to shed little light on whether the current scheme is diluting the voting strength of the minority group"). But there was no such change in Alabama.

Dr. Handley also excluded two general elections in which Sue Bell Cobb, who is white, was the black-preferred candidate. See Trial Tr. II, at 125 (Handley) (testifying that Cobb was "the choice of black voters" in the 2000 Democratic primary election against Saxon for the Court of Criminal Appeals); Trial Tr. II, at 26, 27 (Travis) (testifying that Justice Cobb was his choice for Chief Justice of the Alabama Supreme Court in 2006); Trial Tr. VI, at 28 (Bonneau) (testifying that "Sue Bell Cobb, who was the candidate of choice for African Americans in 2000 and 2006, also won her elections"); Trial Tr. IV, at 149-50 (Gray) (testifying that the ADC endorsed Cobb in her appellate judicial races in 2000 and 2006). Cobb won the 2000 general election for Court of Criminal Appeals, Place 3, defeating her Republican challenger, with 783,962 votes to his 746,785 votes. See Ala. Statewide Elections from 1970 to Present, at 27 (Defs.' Trial Ex. 103; Doc. # 151-1). She ran again in 2006, this time for Chief Justice, and beat the Republican incumbent, Drayton Nabers Jr. (634,494 votes to his 596,237 votes). See id. at 32. Dr. Handley's omissions are perplexing not only because Cobb was the black-preferred candidate in these races, but also because these elections fall squarely within the time period of the elections that Dr. Handley did analyze. In effect, Dr. Handley excluded the bases-loaded home runs from the RBI average of the black-preferred candidates.

Furthermore, Dr. Handley did not opine on whether there were any "'split preference' elections, in which black and white voters have preferred different white candidates." See Davis, 139 F.3d at 1417; see also id. at 1417 n.5 ("[A]n analysis of split-preference elections is also appropriate and relevant."). This omission leaves the record devoid of any evidence that in Democratic races, black and white voters preferred different candidates. For example, Dr. Handley excluded the near successes of Deborah Bell Paseur and Robert Smith Vance, Jr., white Democrats who were the black-preferred candidates.³⁸ See McCrary Expert Report, at 21, ¶ 41 (Pls. Trial Ex. 122; Doc. # 69-1) (providing that, in 2008, 90% of black voters voted for Paseur and 63% of white voters voted for her Republican opponent and that, in 2012, 95% of black voters voted for Judge Vance and 71% of white voters voted for his Republican opponent); Trial Tr. IV, at 156 (Gray) (testifying that the ADC endorsed Judge Vance for chief justice in the 2012 election). Paseur ran for the Alabama Supreme Court in 2008 and was thrown out at home plate, losing by only 12,892 votes out of more than 2 million cast (a .7% margin). See Ala. Statewide Elections from 1970 to Present, at 34 (Defs.' Trial Ex. 103; Doc. # 151-1). Judge Vance ran for Chief Justice of the Alabama Supreme Court in 2012 and was stranded on third, losing to his opponent by only a 3.6% margin. See Ala. Statewide Elections from 1970 to Present, at 37.

38 After this lawsuit was filed, Doug Jones, a white Democrat who the evidence revealed was "almost certainly" the black-preferred candidate, was elected to the U.S. Senate. He was the first Democrat to be so elected in Alabama in a quarter century. Trial Tr. V, at 75 (Bonneau) (Doug Jones "won in the 2017 Senate election" and "was almost certainly the preferred candidate of African Americans in the state of Alabama."); see also Trial Tr. V, at 28 (Bonneau) ("In 2017, I believe African Americans elected the candidate[] of their choice in the special election for the U.S. Senate."). Some would say that Doug Jones's election as a Democrat in a Republican-dominated Alabama in 2018 was a result of "special circumstances." But it is hard to quantify a political "special circumstance" (in the traditional out-of-the-ordinary definition of "special") in a state (1) where no Supreme Court Chief Justice has served a full term in the last six terms; (2) where the same Chief Justice has been elected twice, and removed from office twice, since 2002 (at least once, for defying federal court orders); and (3) where at a point in time four years ago, in a virtual three-branch trifecta, the Governor and Speaker of the House were both convicted of crimes resulting in removal or resignation from office, at the same time as the Chief Justice was getting the second boot.

Dr. Handley's omissions are conspicuous, but the State conceded the second and third Gingles preconditions in Plaintiffs' favor. See Post-Trial Stipulations, at 154, ¶ 29 (Doc. # 176). The State did not hire its own expert to analyze additional elections to counter Dr. Handley's opinions. And the State's expert, Dr. Bonneau, who opines on the causes of racially polarized voting, accepted that Dr. Handley correctly performed the statistical analyses of the races she chose. Trial Tr. V, at 17 (Bonneau). While the court also accepts the results of the statistical analyses of the elections Dr. Handley analyzed, the strength of Plaintiffs' case is weakened by the absence in the record of statistical analyses of the elections where the black-preferred candidate was white and of the successful elections of Justices Adams and Cook. Thus, Plaintiffs' statistical evidence of racially polarized voting is constrained by the thirteen elections that Dr. Handley analyzed.

Minority Political Cohesion. In each of the eight general elections for statewide appellate judges — in the years 2000, 2002, 2006, and 2008 — the African-American candidate received between at least 78.3% and at least 89.6% of the African-American vote. On the primary front, in the 2000 Democratic primary election for Court of Criminal Appeals, Place 3, the black-preferred candidate was the white candidate (Sue Bell Cobb), not the black candidate (Yvonne Saxon). Cobb received at least 63.9% of the African-American vote. In the 2008 Democratic primary, the African-American candidate (Clyde Jones) received at least 74.8% of the African-American vote.

In Gingles, the Supreme Court concluded that the "black voters' support for black candidates was overwhelming" where "black support for black candidates ranged from 71% to 92%" in eleven of the sixteen primary elections and from 87% to 96% in fourteen of the general elections. Gingles, 478 U.S. at 59, Appendix A. And in Solomon, involving a challenge by African-American citizens to the at-large method of electing a county commission and school board, the Eleventh Circuit found minority political cohesion where, under the homogeneous precinct analysis, between 75% and 100% of the African-American vote went to the African-American candidate in four of the six countywide elections for the offices at issue.³⁹ See 899 F.2d at 1019 & n.8 (Kravitch, J., concurring); see also id. at 1020 & n.10 (noting that the "uncontroverted statistical evidence," when compared to the "substantially similar" statistics in Gingles, "was sufficient to establish black political cohesiveness as a matter of law").

39 The four-judge concurring opinion in Solomon concluded that the other two elections, which involved the same black candidate, were "poor indicators of black political cohesiveness" based on the district court's finding that the "white establishment" backed this candidate in two of the school board elections. See 899 F.2d at 1019 n.8 (Kravitch, J., concurring). Notably, the fact that minority political cohesion was lacking in one-third of the endogenous elections did not preclude a finding that the voting patterns across multiple elections revealed minority political cohesion.

Excluding Sue Bell Cobb's 2000 Democratic primary race, in which Dr. Handley finds that voting was not racially polarized, Handley Am. Expert Report, at 15, the statistics here carry similar percentage ranges as in Gingles and Solomon. The low end of the range for African-American support for the African-American candidate exceeds that in both Gingles and Solomon (78.3% here compared to 71% in Gingles, and 75% in Solomon). And the high end of the range (89.6%) is 2.4 percentage points from matching that in Gingles. Moreover, the correlation coefficients, ranging from .708 to .915, between the proportion of African-American registered voters and votes cast for the African-American candidate are further proof of minority political cohesion. See Solomon, 899 F.2d at 1020 & n.9 (Kravitch, J., concurring) (observing that the ecological regression analysis produced "exceptionally strong positive correlation[s] between registered black voters and the number of votes received by the candidate expected to receive black support" in four of the six endogenous elections (ranging from .919 and .996) and in all ten of the exogenous elections (ranging from to .847 to .998)).

Finally, the exogenous 2014 statewide elections — for lieutenant governor, secretary of state, and auditor — are consistent with the foregoing. In each of these three races, the African-American candidate received a substantial portion of the African-American vote — ranging from at least 84.9% to at least 91.2%.

Overall, the percentages and correlation coefficients in nine of the ten endogenous are sufficient for purposes of the second Gingles precondition. See 478 U.S. at 56. Plaintiffs have shown minority political cohesiveness in these appellate-judgeship elections.

White Bloc Voting to Defeat the Black-Preferred Candidate. Plaintiffs also have proven that, in the elections Dr. Handley analyzed, the white bloc votes normally (but not always) defeated the combined strength of the black votes and white crossover votes such that the black-preferred candidate lost.

In the eight general elections for statewide appellate judges — occurring between 2000 and 2008 — the white candidates received between at least 62.6% and at least 73.1% of the white vote. These percentages reveal that white voters do not vote as cohesively as African Americans, whose bloc vote averaged 83.64% in these same eight elections. But the white votes cast for the white candidates, when averaged, are roughly two-thirds, the same as in Gingles, where the Court found that there was white bloc voting. Gingles, 478 U.S. at 59 ("[A]pproximately two-thirds of white voters did not vote for black candidates in general elections, even after the candidate had won the Democratic primary and the choice was to vote for a Republican or for no one."). Additionally, in each of these elections, the white crossover vote ranged from 26.9% to 37.4%, yet the black candidate lost each election. Handley Am. Expert Report, at 17, 22. Comparatively, in Gingles, white crossover voting for the black candidate ranged from 28% to 49% in general elections, and, in those races, the black-preferred candidate lost. See SCLC, 56 F.3d at 1304 (Hatchett, J., dissenting) ("[R]acial bloc voting often exists when the average white crossover vote is greater than twenty-one percent.").

Moreover, in the three 2014 Alabama statewide elections — for lieutenant governor, secretary of state, and auditor — the percentages of the white bloc vote were higher. In these three races, the white opponent received between at least 82.3% and at least 85.3% of the white vote; the white crossover vote for the black candidate ranged from 14.7% to 17.7%; and, in each election, the black-preferred candidate lost. These results are consistent with the pattern of white bloc voting in the endogenous elections and allow an inference to be made about the voting patterns in state appellate judge races.

However, the same pattern does not emerge in the two Democratic primaries. First, in 2000, Sue Bell Cobb received a substantial number of black votes (at least 63.9%) and white votes (at least 72.4%) for Court of Criminal Appeals, Place 3. The overwhelming majority of both black and white voters supported Cobb, and she won. There was not a white bloc vote (or stated differently, the white bloc elected a black-preferred candidate). Second, Clyde Jones, the black-preferred candidate, won the Democratic primary in 2008 for Court of Criminal Appeals, Place 1, with at least 74.8% of the black vote and at most 38.9% of the white crossover vote. Even though the percentages reveal that the white voters preferred Dodd and the black voters preferred Jones, Jones's ability to win the Democratic nomination over his white Democratic opponent is evidence that white voters did not try to vote as a bloc to defeat the black-preferred candidate. Bonneau Expert Report, at 16 ¶ 39 (Defs.' Trial Ex. 24; Doc. # 85-1). The inference of racially polarized voting that Dr. Handley derives from the white low-turnout in the Democratic primary is weak. Nevertheless, these two Democratic primaries do not negate what otherwise is a pattern of racial bloc voting in the limited elections Dr. Handley analyzed. See Gingles, 478 U.S. at 57 ("[A] pattern of racial bloc voting that extends over a period of time is more probative of . . . legally significant polarization than are the results of a single election.").

To sustain their burden under the third Gingles precondition, Plaintiffs do not have to prove that whites vote sufficiently as a bloc always to defeat the black-preferred candidate. They have to prove only that "whites vote sufficiently as a bloc usually to defeat the minority's preferred candidates." Id. at 56 (emphasis added). Plaintiffs have proven that, in the elections Dr. Handley analyzed, white bloc voting usually defeated the black-preferred candidate.

Summary of Racially Polarized Voting under Gingles. Dr. Handley analyzed eight general and two Democratic primary elections for appellate judgeships with African-American candidates on the ballots. Those elections occurred between 2000 and 2008. She also analyzed three exogenous statewide elections with African-American candidates on the ballots occurring in 2014.

Dr. Handley omits analyses of other relevant elections, without satisfactory explanation. She did not analyze the 2008 and 2012 appellate judicial races of Judge Paseur and Judge Vance, the black-preferred candidates. Dr. Handley also omitted analyses of Justice Adams's and Justice Cook's elections. The failure of Dr. Handley to analyze all relevant elections diminishes somewhat the overall weight of her opinions on racially polarized voting.

If those races had been added, there would have been thirteen endogenous statewide general elections between 1982 and 2012 in which the black-preferred candidate competed strongly (defined here as receiving over 44% of the statewide general election vote). Overall, in those thirteen elections, six ended in a winning cause⁴⁰ and seven in a losing cause, resulting in a 46% success for the black-preferred candidate.

40 An unopposed race is a "win" in its own right. See Milwaukee Branch of the NAACP, 116 F.3d at 1199.

1982	Justice Adams	Unopposed
1988	Justice Adams	57.5%
1994	Justice Cook	50.5%
1994	Judge Cobb	52.8%
2000	Justice Cook	47.3%
2000	Justice England	45.7%
2006	Judge England	45.0%

2000	Judge Cobb	51.2%
2006	Justice Cobb	51.5%
2000	Judge Ford	44.5%
2006	Judge Ford	44.9%
2008	Judge Paseur	49.6%
2012	Judge Vance41	48.1%

41 In 2018, Judge Vance also was on the ballot in the general election for Chief Justice of the Supreme Court and garnered 42.5% of the vote. This election occurred two days prior to the trial and more than eleven months after the completion of Dr. Handley's report. Statistical analysis of this election would not have been feasible for Dr. Handley or fair to the parties.

41 In 2018, Judge Vance also was on the ballot in the general election for Chief Justice of the Supreme Court and garnered 42.5% of the vote. This election occurred two days prior to the trial and more than eleven months after the completion of Dr. Handley's report. Statistical analysis of this election would not have been feasible for Dr. Handley or fair to the parties.

Notwithstanding that Dr. Handley did not canvass all relevant elections, the statistical results of the tested elections appear to be enough under existing precedent to demonstrate that African Americans vote cohesively, that black voters and white voters prefer different candidates, and that white voters usually are able to defeat the black-preferred candidate. See Solomon, 899 F.2d at 1021 (Kravitch, J., concurring) (stating that "six futile [endogenous] elections is enough" to prove racially polarized voting under Gingles.). And the smidgen of Plaintiffs' anecdotal evidence is not to the contrary. See, e.g., Trial Tr. II, at 41 (Muhammad). Plus, the fact remains that Dr. Handley's conclusions are uncontradicted by the State. Accordingly, Plaintiffs have proven, by a preponderance of the evidence, the second and third Gingles preconditions.

6. The Totality of Circumstances

Plaintiffs' § 2 vote dilution claim fails as to the AC plans on the first Gingles precondition. As will be discussed in Part V.A.7., Plaintiffs also have not shown that a feasible remedy can be fashioned, as is required in this circuit to fulfill the first Gingles precondition. Notwithstanding the lack of proof, the following analysis, in the interest of thoroughness, proceeds on the assumption that Plaintiffs have satisfied the three Gingles preconditions.

Proof of the three Gingles preconditions "does not end the inquiry." Solomon, 221 F.3d at 1225. The three Gingles preconditions establish an inference of a § 2 violation but are not enough alone to establish liability. This much the Supreme Court has settled:

[I]f Gingles so clearly identified the three as generally necessary to prove a § 2 claim, it just as clearly declined to hold them sufficient in combination, either in the sense that a court's examination of relevant circumstances was complete once the three factors were found to exist, or in the sense that the three in combination necessarily and in all circumstances demonstrated dilution.

Johnson, 512 U.S. at 1011. "As facts beyond the ambit of the three Gingles factors loom correspondingly larger, factfinders cannot rest uncritically on assumptions about the force of the Gingles factors in pointing to dilution." Id. at 1013. "[W]hat appears to be bloc voting on account of race may, instead, be the result of political or personal affiliation of different racial groups with different candidates." Solomon, 221 F.3d at 1225. As will become clear, the totality of circumstances points to factors other than race — most prominently, partisan politics and the decline of the Alabama Democratic Party — as causes of Alabama appellate judicial election results. And "[e]lectoral losses that are attributable to partisan politics do not implicate the protections of § 2." League of United Latin Am. Citizens, Council No. 4434 v. Clements, 999 F.2d 831, 863 (5th Cir. 1993) [hereinafter Clements] (en banc).

The key fact so far in Plaintiffs' favor is their inability to elect a black-preferred candidate since 2006 (Cobb) and an African-American, black-preferred candidate since 1994 (Cook). But, as Justice Stevens wrote in Chisom v. Roemer, "the inability to elect representatives of their choice is not sufficient to establish a violation unless, under the totality of circumstances, it can also be said that the members of the protected class have less opportunity to participate in the political process." 501 U.S. at 397; see also § 10301(b) ("A violation . . . is established if . . . it is shown that the political processes leading to nomination or election in the State . . . are not equally open to participation by members of [the protected class].").

Gingles aims to be helpful in consideration of the totality of circumstances by pointing to the "Senate factors" in the legislative history of § 2. Section 2 itself contains only 200 words; the Senate Report contains 130,892 words spread over 241 single-spaced pages. It is possible to find in those pages everything but plans to build a 747 aircraft. The Gingles Court essentially elevated the Senate factors to the status of law, though not legislative enactment. So, to comply with obligations established in the caselaw, the court has conducted a "comprehensive, not limited, canvassing of the relevant facts," Johnson, 512 U.S. at 1011, to determine if illegal vote dilution on account of race has occurred in Alabama appellate court races. That review includes the relevant Senate factors.

a. Senate Factors 1 and 5: Alabama's History of Discrimination and Its Lingering Effects

Alabama's history of discrimination against African Americans in all areas of life is long, well-documented, and undisputed. Though things have changed, the effects of that discrimination persist to some degree.

i. Senate Factor 1: History of Discrimination in Voting

The first Senate factor addresses Alabama's "history of official discrimination . . . that touched the right" of African Americans "to register, to vote, or otherwise to participate in the democratic process." S. Rep. No. 97-417, at 28.

Official actions designed to exclude African Americans from the State's political life began almost as soon as slavery ended. Summarized here because they are undisputed, those efforts were both debilitating and comprehensive. In 1874, white Democrats "redeemed" Alabama politics from the Republican "black rule" of Reconstruction with the help of "anti-black terrorism across the state." Norrell Expert Report, at 7 (Pls.' Trial Ex. 85; Doc. # 66-1). When Redeemer Democrats gained control of Alabama politics in 1874, they manipulated party rules and electoral devices to eliminate African-American influence in elections, including judicial elections. Norrell Expert Report, at 7-8. From that point on, African-American voting fell dramatically because it was simply too difficult and often dangerous. Many of the African-American votes that were cast were robbed by Redeemer Democrats, who stuffed and stole ballot boxes in the 1880s and 1890s.

One of the most effective of these early mechanisms of disfranchisement was the 1893 Sayre Law. The Sayre Law was a complex election statute designed to keep African Americans and poor whites from voting. Bolden v. City of Mobile, 542 F. Supp. 1050, 1062 (S.D. Ala. 1982), aff'd, 571 F.2d 238 (5th Cir. 1978), rev'd, 446 U.S. 55 (1980). The law "was quite successful" in reducing African-American voter turnout by 22% from 1892 to 1894 and in keeping it below 50%. Id.

In 1901, Alabama convened its infamous all-white constitutional convention, the proceedings of which began with its president declaring the convention's purpose "to establish white supremacy in this State." Hunter v. Underwood, 471 U.S. 222, 229 (1985) (quoting 1 Official Proceedings of the Constitutional Convention of the State of Alabama, May 21st, 1901 to September 3rd, 1901, at 8 (1940)). The convention achieved its purpose. The 1901 Constitution, which required voters either to be literate or to own property and pay a poll tax, removed all but 3,000 of the over-100,000 African Americans registered to vote in Alabama. Norrell Expert Report, at 13 (Pls.' Trial Ex. 85; Doc. # 66-1); see also Dillard v. Crenshaw Cty., 640 F. Supp. 1347, 1358 (M.D. Ala. 1986) ("The 1901 Constitution contained so many different voter qualifications that by 1909 all but approximately 4,000 of the nearly 182,000 black persons of voting age in Alabama had been removed from the rolls of eligible voters." (citation omitted)). In 1902, to further guarantee white supremacy, the Democratic Party, whose primary was the only election that mattered at the time and into the 1980s, excluded African Americans from its primary. Norrell Expert Report, at 13 (Pls.' Trial Ex. 85; Doc. # 66-1). The white primary lasted until it was outlawed by the Supreme Court in Smith v. Allwright, 321 U.S. 649 (1944), but the State did not willingly comply for several years thereafter.

After Smith, the State found new ways to keep African Americans relegated to second-class citizenship. In 1946, voters approved the Boswell Amendment, which required voters to "understand and explain" a provision of the Constitution, to prevent African Americans from registering. See Davis v. Schnell, 81 F. Supp. 872, 873 (S.D. Ala.) (three-judge district court), aff'd, 336 U.S. 933 (1949). After the Boswell Amendment was struck down in Davis, registrars in several counties continued their efforts to keep African Americans from voting by subjecting them to more rigorous procedures than whites.⁴² See United States v. Logan, 344 F.2d 290 (5th Cir. 1965) (per curiam) (Wilcox County); United States v. Mayton, 335 F.2d 153 (5th Cir. 1964) (Perry County); United States v. Atkins, 323 F.2d 733 (5th Cir. 1963) (Dallas County); Alabama v. United States, 304 F.2d 583 (5th Cir.) (Macon County), aff'd sub nom. Alabama v. United States, 371 U.S. 37 (1962); United States v. Parker, 236 F. Supp. 511 (M.D. Ala. 1964) (Montgomery County); United States v. Penton, 212 F. Supp. 193 (M.D. Ala. 1962) (Montgomery County).

42 In one case arising out of this district, the voter registrars of Bullock County supplemented spurious tests, playing "hide and seek" with African-American citizens by hiding themselves in the bowels of the courthouse basement behind closed doors when African-Americans showed up to register to vote. See Sellers v. Wilson, 123 F. Supp. 917 (M.D. Ala. 1954). The case was brought by two prominent African-American attorneys, Arthur D. Shores of Birmingham, Alabama, and Thurgood Marshall, who obtained a just ruling for their clients and for all African Americans: "The supreme law of this republic is that no tests can be required of a Negro applicant as a pre-requisite to registration as a voter that is not required of a white applicant; therefore, let no Board of Registrars try to devise any scheme or artifice to do otherwise." Id. at 920.

Of particular note, in the 1950s and 1960s, the Alabama Supreme Court played a key role in disfranchisement by drafting the questionnaire registrars used to disqualify African Americans. See Parker, 236 F. Supp. at 515; Penton, 212 F. Supp. at 196. That court was also complicit in the State's efforts to keep the National Association for the Advancement of Colored People ("NAACP") from operating in Alabama. See NAACP v. Alabama, 357 U.S. 449 (1958) (reversing the Alabama Supreme Court's contempt order against the NAACP for refusing to turn over membership records to state officials).

The State also sought to maintain white supremacy through racial gerrymandering, see Gomillion v. Lightfoot, 364 U.S. 339 (1960); Sims v. Baggett, 247 F. Supp. 96 (M.D. Ala. 1965), and selective enforcement of the poll tax, see United States v. Alabama, 252 F. Supp. 95 (M.D Ala. 1966) (three-judge district court). And when legal mechanisms were not enough, some public officials used intimidation and violence to keep African Americans from voting. See United States v. McLeod, 385 F.2d 734 (5th Cir. 1967); Clark v. Boynton, 362 F.2d 992 (5th Cir. 1966); Williams v. Wallace, 240 F. Supp. 100, 104 (M.D. Ala. 1965).

ii. Senate Factor 5: Discrimination in Education, Employment, and Health

The fifth Senate factor addresses "the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment and health, which hinder their ability to participate effectively in the political process." S. Rep. No. 97-417, at 29. Thus the court must find that (1) past discrimination in these areas existed, (2) that those effects are borne by African Americans today, and (3) that these effects hinder African Americans' present ability to participate in the political process. When plaintiffs have shown past discrimination and present socioeconomic disparities, "and where the level of black participation in politics is depressed, plaintiffs need not prove any further causal nexus between their disparate socioeconomic status and the depressed level of political participation." Id. at 29 n.114.

For at least the first seventy years of the twentieth century, segregation was a fact of life in Alabama. And nearly every aspect of public life was segregated. See, e.g., Norris v. Alabama, 294 U.S. 587 (1935) (juries); Smith v. Y.M.C.A., 316 F. Supp. 899 (M.D. Ala. 1970) (public pools), aff'd as modified sub nom. 462 F.2d 634 (5th Cir. 1972); Marable v. Ala. Mental Health Bd., 297 F. Supp. 291 (M.D. Ala. 1969) (three-judge district court) (mental health facilities); Washington v. Lee, 263 F. Supp. 327 (M.D. Ala. 1966) (three-judge district court) (jails and prisons), aff'd, 390 U.S. 333 (1968); Gilmore v. City of Montgomery, 176 F. Supp. 776 (M.D. Ala. 1964) (public parks), aff'd and modified, 277 F.2d 364 (5th Cir. 1960); Cobb v. Montgomery Library Bd., 207 F. Supp. 880 (M.D. Ala. 1962) (public libraries and museums); United States v. City of Montgomery, 201 F. Supp. 590 (M.D. Ala. 1962) (airports); Browder v. Gayle, 142 F. Supp. 707 (M.D. Ala.) (three-judge district court) (buses), aff'd, 352 U.S. 903 (1956), and aff'd sub nom. Owen v. Browder, 352 U.S. 903 (1956).

African Americans were long denied the same educational and employment opportunities as whites. Well after Brown v. Board of Education, 347 U.S. 483 (1954), Alabama state officials continued to resist integration of public schools, requiring further intervention by federal courts. See, e.g., Armstrong v. Bd. of Educ. of City of Birmingham, 323 F.2d 333 (5th Cir. 1963); Lee v. Macon Cty. Bd. of Educ., 231 F. Supp. 743 (M.D. Ala. 1964) (three-judge district court). The only public law school in the state, The University of Alabama School of Law, did not admit African-American students until 1964, had no African-American students to enroll between 1965 and 1968, and had no African-American graduates until 1972. See McCrary Expert Report, at 12, ¶ 24 (Pls.' Trial Ex. 122; Doc. # 69-1); see also Trial Tr. IV, at 73 (Cook) (testifying that, even during the decade after Brown v. Board of Education, "schools in Alabama were not extending invitations to persons of my ethnicity to attend [law school,] so I attended Howard University Law School"). Among that first group of African-American graduates of The University of Alabama's law school was Judge John England, who testified in this case. Trial Tr. I, at 30 (England). The State also discriminated against African Americans in public employment. Paradise v. Prescott, 585 F. Supp. 72 (M.D. Ala. 1983) (state troopers), aff'd sub nom. United States v. Paradise, 480 U.S. 149 (1987); United States v. Frazer, 317 F. Supp. 1079 (M.D. Ala. 1970) (finding race-based employment discrimination in six different state agencies).

This extensive and shameful history led Fifth Circuit Judge Richard T. Rives, himself a lifelong resident of Montgomery, to conclude for a three-judge court that, "from the Constitutional Convention of 1901 to the present, the State of Alabama has consistently devoted its official resources to maintaining white supremacy and a segregated society." United States v. Alabama, 252 F. Supp. at 101 (three-judge district court). And though Judge Rives made that statement in 1966, the effects of the past are still felt today, as black Alabamians have overall lower educational attainment levels than white Alabamians,⁴³ Trial Tr. II, at 177-78 (Cooper), and are more than twice as likely to be unemployed and to live in poverty, Trial Tr. II, at 178 (Cooper). See also Pls.' Proposed Findings of Fact, at 71, ¶¶ 67, 68 (Doc. # 176) ("76.5% of White Alabamian householders are homeowners[,] and 50% of African American Alabamian householders are homeowners," ¶ 67; "4.0% of White Alabamian households and 12.7% of African American Alabamian households are without a vehicle," ¶ 68.).

43 In Alabama, for individuals 25 years of age and older, a high school diploma (GED "or alternative") is the highest level of education for 33.2% of African Americans and 30.3% of whites. But more African Americans than whites have not earned even a high school diploma (18% versus 12.6%, respectively). For 31.1% of African Americans and 29.8% of whites, the highest level of education earned is "some college or [an] associate's degree." The largest educational gap between African Americans and whites occurs in the attainment of four-year college degrees and above: Whites lead African Americans 27.3% to 17.7% in this category. Pls.' Trial Ex. 89 (Ex. C1 to Cooper Expert Report); Trial Tr. II, at 193-94 (Cooper).

iii. Evaluation and Weighing of Senate Factors 1 and 5

As the United States Supreme Court has recognized, Alabama today is a vastly different place than it was even a half-century ago. See generally Shelby Cty. v. Holder, 570 U.S. 529 (2013). Overt discriminatory election devices have long been eliminated. Voter registration and turnout-rates among African-Americans and whites have reached parity. See id. at 535 (citing Dep't of Commerce, Census Bureau, Reported Voting and Registration, by Sex, Race and Hispanic Origin, for States (Nov. 2012) (Table 4b)); id. at 548 (citing S. Rep. No. 109-295, at 11 (2006)) (showing an almost identical percentage of white and black eligible voters who are registered in Alabama). In 2010 in Alabama, 74.38% of the black, voting-age population was registered to vote, compared to 74.35% of the white, voting-age population. Compare 2010 Census Population Statistics (Defs.' Trial Ex. 90), with Ala. Sec'y of State, Voter Registration Statistics — 2010, https://www.sos.alabama.gov/alabama-votes/voter/election-data (last visited Feb. 5, 2020). In 2017, Doug Jones became the first Democrat to win a U.S. Senate seat in Alabama in a quarter century, in an election in which African-American votes were decisive. See Trial Tr. II, at 46 (Muhammad); Trial Tr. II, at 62 (Harris); Trial Tr. V, at 75 (Bonneau); McKee Rebuttal Report, at 29 (Pls.' Trial Ex. 123; Doc. # 88). Plaintiffs simply have not shown that, in present-day Alabama, there are any barriers keeping African Americans from participating in the political process as voters. The level of black participation in the electoral process is not depressed.

But the evidence does show that socioeconomic disparities between black and white Alabamians hurt black candidates because it is harder for them to raise money. See Trial Tr. III, at 29-30 (McCrary). Evidence referenced in Part V.A.6.a.ii. supra lends support for this finding. In their 2000 campaigns, which they lost, Justices Cook and England were outspent by their white Republican opponents $1,254,450 to $437,481 and $1,107,838 to $500,680, respectively. Bonneau Expert Report, at 10, ¶ 24 (Defs.' Trial Ex. 24; Doc. # 85-1). However, few realities have only a sole cause. During this period, business interests were successfully spending large amounts of money to help Republicans win judicial races, making Alabama Supreme Court races some of the most expensive in the country. Gaylord Reply Report, at 10 (Defs.' Trial Ex. 27; Doc. # 93-2); Trial Tr. V, at 136 (Bonneau). Nevertheless, in intermediate appellate court races since 2000, the losing African-American candidates spent on average about five times less than the rest of the losing candidates (all of whom, with the exception of Sue Bell Cobb, were Democrats). Bonneau Expert Report, at 11-12, ¶ 28 (Defs.' Trial Ex. 24; Doc. # 85-1). The State argues that disparities in campaign spending help show that race-neutral reasons drive African-American electoral defeats, but the court cannot ignore the fact that the depressed socioeconomic status of African-American Alabamians is reflected in the amount of money African-American candidates are able to raise.

In sum, Alabama's long and sordid history of official discrimination against African Americans weighs in favor of Senate factor 1 because that history will never change. But the more recent evidence of African-American voter registration and turnout points toward a finding that Alabama's appellate judicial elections today are equally open to participation by African Americans. See § 10301(b). Plaintiffs put on very little evidence to the contrary. But, as to Senate factor 5, the effects of past discrimination are still felt today, negatively affecting African-American candidates' ability to raise money. Senate factor 1 weighs in favor of Plaintiffs, while Senate factor 5 weighs moderately in favor of Plaintiffs.

b. Senate Factor 2: The Extent of Racially Polarized Voting (Causation)

The second Senate factor examines "the extent to which voting in the elections of the state or political subdivision is racially polarized." S. Rep. No. 97-417, at 29. Some courts have identified this as one of the two most important factors, the other being the extent to which members of the protected class have been elected to office in the State or political subdivision. The evidence on racially polarized voting was analyzed in the discussion of the second and third Gingles preconditions. There, the inquiry focused solely on "how" black and white voters voted. The focus now, at the totality-of-circumstances stage, is on evidence of causation, which looks to "why" voters cast their ballots for certain candidates. That is to say, "what appears to be bloc voting on account of race [which is the inevitable result of satisfying the three Gingles preconditions], may, instead, be the result of political or personal affiliation of different racial groups with different candidates." Solomon, 221 F.3d at 1225 (alterations added).

The State counters Plaintiffs' evidence of vote dilution by showing that black-preferred candidates lose because they "are running as Democrats in a red State." Defs.' Post-Trial Br., at 6 (Doc. # 168). The evidence of this theory is substantial.

i. Party Alignment in Alabama

Since 2000, all African-American candidates for statewide office have run as Democrats, and all have lost, while two black-preferred (white) Democratic candidate have won a total of three races (Cobb and Doug Jones). However, "[t]he circumstance that a group does not win elections does not resolve the issue of vote dilution." LULAC, 548 U.S. at 428. After all, "the ultimate right of § 2 is equality of opportunity, not a guarantee of electoral success for minority-preferred candidates of whatever race." Johnson, 512 U.S. at 1014 n.11.

Alabama is indeed a ruby red state — "one of the most Republican states in the entire South." Trial Tr. VI, at 180 (McKee). That fact has made it virtually impossible for Democrats — of any race — to win statewide in Alabama in the past two decades. Since 2008, while all black Democrats have lost their statewide races, so have all white Democrats, with the exception of Doug Jones (2017 special U.S. Senate election). And, since 2000, Sue Bell Cobb has been the only victorious Democratic candidate for an appellate judgeship (Court of Criminal Appeals in 2000 and Chief Justice of the Supreme Court in 2006).

Within the Democratic Party, there is evidence that black Democrats get more votes in statewide appellate judicial races than white Democrats. Under a multivariate regression model in which the two independent variables are the percentage of registered voters who are African American and whether the losing candidate was African American, black Democratic candidates perform an average of 1.4 percentage points better than white Democratic candidates.⁴⁴ Bonneau Expert Report, at 15, ¶ 36 (Defs.' Trial Ex. 24; Doc. # 85-1). Excluding Sue Bell Cobb from this measure, that number increases to 2.1 percentage points. Id. And, since those results hold constant the percentage of registered voters who are African American, Trial Tr. V, at 66 (Bonneau), it is unlikely that a "lag" in white support for black Democrats is being disguised by a "boost" in black support for black Democrats. In fact, Dr. Bonneau found that, in areas with the highest percentage of African-American voters, black Democrats either perform worse than white Democrats or there are no statistically significant differences. Bonneau Expert Report, at 16, ¶ 38 (Defs.' Trial Ex. 24; Doc. # 85-1). From these results, it reasonably can be inferred that white Democratic voters give equal or greater support to black Democratic candidates as they do to white Democratic candidates. Plaintiffs do not challenge this fact, only its significance.

44 There were four African-American candidates who ran as Democrats during the time period of Dr. Bonneau's analyses. Bonneau Expert Report, at 9, ¶ 22 & App. A (Defs.' Trial Ex. 24; Doc. # 85-1).

The comparison of the electoral strength of white and black Democrats is significant considering that losing black candidates in general elections spent much less money in appellate court races than the rest of the losing candidates. Bonneau Expert Report, at 10-12, ¶¶ 24, 28 (Defs.' Trial Ex. 24; Doc. # 85-1). So despite devoting significantly less financial resources to their campaigns, black Democrats still slightly outperformed white Democrats against Republicans at the polls.

That white Democratic primary voters appear to give equal support to black Democratic candidates suggests that black candidates are not penalized in appellate judicial elections by their race alone. See Clements, 999 F.2d at 879 ("White voters' support for black Republican candidates was equal to or greater than their support for white Republicans. Likewise, black and white Democratic candidates received equal percentages of the white vote. Given these facts, we cannot see how minority- preferred judicial candidates were defeated 'on account of race or color.' Rather, the minority-preferred candidates were consistently defeated because they ran as members of the weaker of two partisan organizations."); Lopez, 339 F. Supp. 3d at 613 (evidence that, "whether running as Democrats or Republicans, Hispanic candidates have tended to slightly outperform non-Hispanic candidates with non-Hispanic voters" was evidence "that partisanship is a better explanation" for Hispanic defeats). The evidence is not as strong here as it was in Clements and Lopez, because neither Dr. Bonneau's data nor Dr. Handley's data included an election with an African-American Republican. Hence, there was no empirical evidence supporting or discrediting the State's theory that white voters would equally support a black Republican and a white Republican candidate.⁴⁵

45 It appears from the record that the only African American to run statewide for a judgeship in the general election as a Republican was Greg Griffin, who ran for a seat on the Court of Criminal Appeals in 1994 and lost to the venerable Sue Bell Cobb. See Ala. Statewide Elections from 1970 to Present, at 22 (Defs.' Trial Ex. 103; Doc. # 151-1).

There was evidence that an African-American Republican could win statewide in Alabama. Justice Murdock testified as much. On cross-examination, Dr. McKee, Plaintiffs' rebuttal witness, agreed that an African-American conservative Republican running statewide "would have a good chance of being elected." Trial Tr. VI, at 213-14 (McKee). The court recognizes also that a conservative black Republican might not be the black-preferred candidate. But it is important that, despite large spending disparities, losing black candidates receive a slight edge in their share of the vote over losing white candidates. Socioeconomic disparities between black and white Alabamians do make it harder for black candidates to raise money, and the court has not ignored that fact. But the notion that African-American candidates lose solely because of their skin color is not supported by the evidence. There is a strong case that party, not race, is driving election results in Alabama appellate judicial races.

ii. The Relative Strength of the Two Parties

That the Alabama Democratic Party is significantly weaker than its Republican counterpart makes it even harder for any Democratic candidate — white or black — to get elected. One need look no further than the past four general elections, in which Democrats put up candidates for only twelve out of forty-six statewide offices, and the failure of any Democratic candidate to qualify to run in the March 3, 2020 primary for six open appellate judicial seats, to see that the Alabama Democratic Party is on life support.

Several witnesses agreed that the Alabama Democratic Party is in very rough shape. See Trial Tr. I, at 69 (England) ("The Democratic Party's position is significantly weaker than it was 20 years ago on all levels [—] leadership, funds, ability to help their candidates. It's across the board weaker."); Trial Tr. II, at 33 (Travis) (explaining that the state Democratic Party's ability to raise money "is not as great" as the Republican Party's and that its few fundraising successes have been from out of state); Trial Tr. V, at 193-94 (Harwood) (describing recent losing Democratic candidates' intense complaints about party leadership and lack of fundraising or organizational support); Trial Tr. V, at 197 (Harwood) (agreeing that, during the 120-year transition from the dominance of the Democratic Party to the dominance of the Republican Party, there has "been a concomitant loss of influence in infrastructure and strength in the official Democratic Party in Alabama"); Trial Tr. V, at 197-98 (Harwood) (testifying that the late Paul Hubbard, former leader of the Alabama Education Association, was a "power" in the Democratic Party and that Hubbard's passing weakened the party); Trial Tr. VI, at 217-22 (McKee) (agreeing generally with those candidates' criticisms about the Democratic Party and describing the Democratic Party as being in a "moribund state"); Trial Tr. VI, at 215-16, 222 (McKee) (agreeing that white Democrats are having no more successes than black Democrats in Alabama elections, id. at 215-16, and stating that the Democratic Party in Alabama "is pretty darn electorally weak, and it has been for a long time," id. at 222).

The fractured state of the Alabama Democratic Party is now the subject of a state-court lawsuit in which one faction of the party has sued the other for party control. See Verified Complaint, Ala. Democratic Party, et al. v. Gilbert, et al., No. CV-2019-000531.00 (Circuit Court of Montgomery Cty., Ala. Oct. 30, 2019). Without a viable party behind them, Democratic candidates of any race have an uphill battle. Trial Tr. V, at 69-70 (Bonneau) (explaining that weak infrastructure makes it more difficult for a party to recruit candidates and support those candidates with volunteers and money).

An obvious consequence of the lack of vitality of the Alabama Democratic Party in the twenty-first century is its failure to compete in statewide judicial elections. The best empirical evidence of this impotency is raw facts about the number of elections and the results of elections. Supreme Court races in Alabama have in recent history elected, in staggered terms, one justice every six years (1990, 1996, 2002, 2008, 2014, and so on), three justices every six years (1986, 1992, 1998, 2004, 2010, 2016, and so on), and five justices, including the Chief Justice, every six years (1988, 1994, 2000, 2006, 2012, 2018, and so on).

Since and including 1980, there have been 128 elections for a position on one of Alabama's three appellate courts. From 1980 until 1994 (the year of the infamous Hornsby-Hooper race), Democrats were unanimous victors, winning all forty races, fifteen by forfeit (no Republican candidate). In 1994, Republicans won three of ten appellate court races: Hooper for the Supreme Court and two courts of appeals races. As mentioned before, Hooper was the first Republican in history elected to the Alabama Supreme Court, ending a 126-year drought. Republicans lost the other four Supreme Court seats that year, as well as three of five courts of appeals races. But they had begun to compete in earnest.

In 1996, Republicans skunked the Democrats, winning all five appellate races. In 1998, Democrats won three; Republicans won one. In 2000, a presidential election year in which Bush won (5-4, just barely), there were eleven appellate court races. Republicans won ten (all five on the Supreme Court) of the eleven places, with the aforesaid venerable (but increasingly lonely, in a strictly partisan sense) Sue Bell Cobb winning as a Democrat on the Court of Criminal Appeals. Since 2000, Sue Bell Cobb is the only Democrat to win an appellate judgeship in Alabama, when she was elected Chief Justice in 2006. And out of nineteen appellate judges, she was the only Democrat on the bench until her resignation in 2011. Republicans won all eight races combined in 2002 and 2004.

In all, since 2002, Republicans have shut out Democrats in eight of nine election cycles in appellate judicial races in Alabama. From 1998 until 2008, Democrats competed at a little less than par (in number of contested elections) with Republicans in statewide judicial races. In 2010, Democrats put up four candidates to Republican's five candidates for five seats. In 2012, Democrats put up only one candidate for eleven seats (Judge Vance for Chief Justice, running against Roy Moore — and Judge Vance took 48.12% of the vote). Democrats put up zero candidates for five places in 2014 and zero candidates for three places in 2016. Again in 2018, Judge Vance ran for Chief Justice, this time losing to Republican Justice Tom Parker. But Judge Vance was one of only two Democrats to take the field against eleven Republicans that year. See Ala. Statewide Elections from 1970 to Present, at 35-44 (Defs.' Trial Ex. 103; Doc. # 151-1).

Since Sue Bell Cobb won in 2006, Republicans have shut out Democrats thirty-nine to zero in six consecutive election cycles for thirty-nine appellate court slots. And the court takes judicial notice that, out of the six open seats for Alabama appellate judgeships in 2020, no Democrat has qualified to run in the March 3, 2020 primary. See Alabama Democratic Party, Certified Candidates for Primary Election, http://www.aldemocrats.org/blog/certified_candidates_for_2020_primary_election (last visited Feb. 5, 2020); see also SCLC, 56 F.3d at 1288 n.13 (taking judicial notice of the "current depiction of the composition of the bench" because that "information [was] not available at the time the district court rendered its decision").

No African-American candidate has stood for statewide judicial election in five consecutive election cycles, and the Democrats only put up two candidates in the last three election cycles for nineteen slots. Two for nineteen. That is evidence — strong evidence — of the current state of impotence of the Alabama Democratic Party. When African Americans run only as Democrats, their fate, of late, is sealed. That fate, unless you are Justice Sue Bell Cobb, is no different for white candidates running on the Democratic ticket for appellate judicial races.

Plaintiffs have failed to explain how they expect to win appellate judicial elections when they put up few and sometimes no candidates. Futility was their only argument. But that argument applies to the Democratic Party as a whole, not to the proper comparator of black-preferred candidates. Trial Tr. I, at 88 (England) ("If you aren't a Republican candidate, I don't think you really have a chance in 99 out of a hundred cases" to win an appellate seat). The evidence supports an inference that party, not race, causes election outcomes in Alabama.

Plaintiffs offered no evidence whatsoever to refute the dire circumstances of the Alabama Democratic Party. And they produced no evidence at all that this dismal state is "on account of race." In fact, the evidence in the case supports the opposite conclusion. The weakness of the Alabama Democratic Party in the twenty-first century is a circumstance that contributes significantly to the failure of African Americans to elect their candidates of choice.

iii. Partisan Judicial Elections

Partisan elections are a long-running part of Alabama's history for electing appellate judges. Candidates for appellate judicial office first compete in partisan primaries to represent their party in the general election. The primary winners appear on the general election ballot under their political party. Plaintiffs do not seek to eliminate the partisan nature of Alabama's elections for appellate judges. See Trial Tr. I, at 8 ("[I]n [Plaintiffs'] view, the partisan nature of the elections may or may not be lamentable; but it's not part of the remedy that we're seeking in this case. In other words, we're not trying to change the way the elections are held except for the fact that they should be held from districts.") (opening statements).

The fact that appellate judges must run under a party banner in a highly partisan state reinforces the theory that judicial election results are driven in part by the party of the candidate, not the race of the candidate. Judge England testified that partisan judicial elections give "[a] disincentive to look at the qualifications of candidates" because voters only consider "the R or the D in front of [the candidate's] name." Trial Tr. I, at 71 (England).

Judge England is right: Partisan elections lead to partisan results. In lower-profile races (like judicial races) where voters know less about the candidates, an R or D next to a candidate's name gives uninformed voters enough information to feel comfortable voting for that candidate. See Trial Tr. V, at 43-44 (Bonneau) (explaining that voter participation is higher in partisan than nonpartisan elections and voters can identify the candidate of their preferred party more often in partisan than nonpartisan elections); Trial Tr. V, at 59 (Bonneau) ("Particularly in down-ballot races . . . party ID can be a very helpful cue for voters.").

It may be true that, if partisan elections were eliminated, candidates would use "code symbol[s]" like "the individual's face, statements like I'm a conservative Christian," Trial Tr. I, at 109-10 (Jones), or "God or guns," Trial Tr. I, at 138-39 (White), to tell voters that the candidate is "basically a Republican and the other person is not," Trial Tr. I, at 110 (Jones). But when a large percentage of the Alabama electorate is voting straight ticket, Trial Tr. V, at 57 (Bonneau), those "code symbols" are unnecessary because the R or the D is sometimes all the information the voters need.

iv. Straight-Ticket Voting

Straight-ticket voting only exacerbates the phenomenon of partisan-driven election results. Many voters are driven to the polls because of races at the top of the ticket, then end up voting for down-ballot candidates, including judges, of the same party as their preferred top-of-the-ticket candidates. Trial Tr. V, at 73-74 (Bonneau). Between 2008 and 2014, about a quarter of total ballots cast in Alabama were straight-ticket Democrat, and another quarter of total ballots in Alabama cast were straight-ticket Republican. Bonneau Expert Report, at 7, ¶ 16 (Defs.' Trial Ex. 24; Doc. # 85-1); Trial Tr. V, at 57 (Bonneau). Moreover, the most recent numbers show that straight-ticket voting is even more prevalent today and decisively in the Republican party column. One witness for Plaintiffs testified that, after the 2018 election (which occurred during trial), he believed that two-thirds of Alabama voters voted straight ticket. Trial Tr. II, at 30 (Travis). That assertion is consistent with the statistics reported by the Alabama Secretary of State. According to the Alabama Secretary of State, 65% of voters in the 2018 general election voted straight ticket.⁴⁶ See Ala. Sec'y of State, Elections Data Downloads, 2018 General Voter Turnout Statistics, https://www.sos.alabama.gov/alabama-votes/voter/election-data (last visited Feb. 5, 2020). And the vote totals in general elections for state offices in the past ten years are consistent with widespread straight-ticket voting, as Republican candidates and Democratic candidates received a similar share of votes across the board in each election. See Ala. Statewide Elections from 1970 to Present, at 35-36, 39, 43-44 (Defs.' Trial Ex. 103; Doc. # 151-1); see also Trial Tr. V, at 117 (Bonneau) ("And when you add to that the straight-ticket voting option in the state of Alabama, it's not surprising to see a tight cluster of vote totals among candidates where voters just go in and vote all straight party, all for one party's candidates or all for the other party's candidates.").

46 Of the 1,725,877 total ballots cast, Republicans had 663,269 straight-ticket votes, and Democrats had 462,065 straight-ticket votes. While Republicans cast more straight-ticket votes than Democrats in the 2018 general election, the data on the Alabama Secretary of State's website does not include a breakdown of the total number of votes (straight-ticket voting plus split-ticket voting) cast by Republicans and Democrats. Hence, there is no evidence as to which party had a higher percentage of straight-ticket voters.

While the Republican party did not have a clear advantage in straight-ticket voting in the elections Dr. Bonneau cites, it is noteworthy that roughly two-thirds of the Alabama electorate is voting for a party, not necessarily for particular candidates. Thus, many straight-ticket voters are likely not looking at down-ticket candidates' qualifications (or, for that matter, their races or even their names), but merely under what party banner they are running.

v. The Competitiveness of Black-Preferred Candidates in Appellate Judicial Races

The ability of black-preferred candidates to compete well in appellate judicial races demonstrates the absence of vote dilution. Notwithstanding recent party weaknesses, the Alabama Democratic Party has backed black-preferred candidates who have won or competed strongly in races. As already discussed, since 1982, two African-American Supreme Court candidates, who were the black-preferred candidates, won three statewide general elections: two elections for Justice Oscar Adams (one unopposed and one opposed) and one for Justice Ralph Cook (opposed). Counting primaries, that would be two for Justice Adams (one unopposed and one opposed), one primary runoff for Justice Adams, and two primaries for Justice Cook (who had no opposition in 1994 and who beat his opponent in the 2000 Democratic primary, but lost the 2000 general election). Out of eight statewide primary and general races, head-to-head with white opponents, these two black incumbents won seven. The unopposed races are "wins" in their own right. See Milwaukee Branch of the NAACP, 116 F.3d at 1199. Justice England was appointed in 1999 to fill an unexpired term, won the 2000 Democratic primary, but lost in the general election of 2000. He ran again in 2006, won the Democratic primary and lost to his Republican challenger (Glenn Murdock) in the general election. Thus, in statewide elections, Judge England won two primaries and lost two general elections. In 2000 and 2006, Judge England garnered 45.7% and 45% of the statewide general election vote respectively, and daresay, respectably, in a state that is 26% African American, and, in both elections, he had more votes than the losing white Democratic candidates running for associate justice for other place numbers. See Ala. Statewide Elections from 1970 to Present, at 27, 32-33 (Defs.' Trial Ex. 103; Doc. # 151-1); Post-Trial Stipulations, at 31, ¶ 153 (Doc. # 176). In 2008, Judge Clyde Jones, the African-American, black-preferred candidate, won the Democratic primary but lost the general election to a Republican, Beth Kellum.

In the last ten years, no African-American candidate has stood for office in a statewide judicial race. Of course, the § 2 focus is not on the race of the candidate but the race of the voter, and African-Americans have a right not to have their vote for their preferred candidate diluted. There is evidence of fourteen endogenous statewide general elections since and including 1982 in which the black-preferred candidate competed strongly (defined here as receiving over 40% of the statewide general election vote), six in a winning cause and eight in a losing cause, resulting in a 38% success rate for the black-preferred candidate.

1982	Justice Adams	unopposed
1988	Justice Adams	57.5%
1994	Justice Cook	50.5%
1994	Judge Cobb	52.8%
2000	Justice Cook	47.3%

2000	Justice England	45.7%
2006	Judge England	45.0%
2000	Judge Cobb	51.2%
2006	Justice Cobb	51.5%
2000	Judge Ford	44.5%
2006	Judge Ford	44.9%
2008	Judge Paseur	49.6%
2012	Judge Vance	48.1%
2018	Judge Vance	42.5%

Additionally, the average percentage of the total vote of the black-preferred candidate is 47% over the 18-year span from 2000 to 2018. The percentage of African-American voters ranged from 24% to 26%, so there is evidence of a strong but not strong-enough coalition to elect a black-preferred candidate — if such a candidate would try.

vi. Whether Party Is a Proxy for Race

Plaintiffs correctly point out that the existence of extreme partisanship in the state does not matter unless one first explains why Alabama's partisan preferences are the way they are. Why do white Alabamians vote overwhelmingly Republican, and why do African-American Alabamians vote overwhelmingly Democratic? According to Plaintiffs, the answer is race. For this proposition, Plaintiffs rely heavily on the testimony of their rebuttal witness, Dr. McKee.

Dr. McKee's impressive credentials no doubt render him an expert on topics such as southern political history and partisan realignment in the south, but Dr. McKee is not an expert on the voting behavior of Alabamians, particularly their subjective voting motivations. No judge has ever qualified him as such, and he has never — until he took the stand here, and as far as this court can tell — claimed to be an expert on the subjective voting motivations of Alabamians. The record does not support that Dr. McKee knows any Alabama voter outside those he worked with professionally on this case. Yet on direct examination he opined that whites in the south are Republican for racial reasons because he "can't think" of any other reason. Trial Tr. VI, at 157, 174. He characterized "conservative values" as "racialized." Trial Tr. VI, at 157, 173, 196. He espoused that in Alabama race is the reason for the realignment of the Republican Party as the dominating party and that "to me" this conclusion does not "take[] any in-depth analysis"; "it's face validity."⁴⁷ Trial Tr. VI, at 131. He discarded tort reform and high profile statewide races as legitimate forces in Alabama for party alignment. See Trial Tr. IV, at 170 ("[Voters are] not going to realign over tort reform"); Trial Tr. IV, at 169-70 (rebuffing the premise that the Hooper/Hornsby race was a catalyst for the realignment of the political parties in Alabama and quipping, "I know that the Hooper/Hornsby dispute gets a lot of attention in this courtroom"). On cross-examination, he doubled down, proclaiming, "I'm starting to laugh when I hear Hooper/Hornsby again," Trial Tr. IV, at 196, and testifying that a white voter (presumably, every white voter) who votes Republican does so because of race, Trial Tr. VI, at 231.

47 This testimony harks the undersigned back to law practice days when a party, who was asked if he had any evidence of a fact that he had asserted, testified: "I don't need evidence. It's true."

Many witnesses were offered by both sides who are qualified to give lay opinions on the motivations of Alabama voters over time and in the context of the state's rich political history. Justices Cook, Murdock, and Harwood and Judges England and Jones are clearly qualified to say in their opinion, as lifelong Alabama citizens, voters, and direct participants in numerous political campaigns, what a subjective motivation for voting behavior in Alabama might or might not be. Justices Cook and Harwood both testified that they have campaigned in all 67 Alabama counties, and benefitted from it; no doubt the others have the same experience. Justice Murdock has litigated as a private attorney one of the most noteworthy political cases in Alabama's history, the Hooper-Hornsby case so maligned by Dr. McKee. Justice Murdock has attended meetings of the state Republican committee for twenty-five years. Trial Tr. VI, at 57. Moreover, even local counsel for all parties would be abundantly qualified to testify on voter motivations in Alabama; many are nationally recognized experts, authors, and speakers on that and on related topics. Alas, as to testifying on the subjective motivations of the voting behavior of Alabama voters, Dr. McKee is not qualified.

Dr. McKee's conclusion that race best explains the present results in Alabama judicial elections has at least two flaws. First, his explanation for voter behavior unpersuasively writes off ideology as a cause of partisan sorting. Second, Dr. McKee did not consider facts specific to Alabama judicial elections that speak to the results of those elections. He did not include any state judicial races in his empirical analysis.

The main part of Dr. McKee's historical analysis will take a few pages to sort out. It begins with the 1964 presidential election, when he says the national parties switched positions on civil rights. Democratic President Lyndon B. Johnson championed the Civil Rights Act of 1964, but Republican presidential nominee Barry Goldwater voted against it. Goldwater's vote signaled which party, at least on a national level, "took up the cause of civil rights." Trial Tr. VI, at 130 (McKee). Dr. McKee's testimony is notably devoid of anything other than one Senator's vote out of a hundred that shows the parties "permanently reversed their positions on black civil rights." McKee Rebuttal Report, at 9 (Pls.' Trial Ex. 123; Doc. # 88). In fact, he acknowledges that the GOP's success with white southerners that year was "ephemeral." Id. at 10. Dr. McKee vastly oversimplifies an historic cultural shift in white votes. Whites going overboard the Alabama Democratic ship did not pass African Americans abandoning the Republican ship in 1964. Whites took a long time to jump.

It is beyond dispute that the Democratic Party dominated both national and state elections in Alabama from 1874 until 1964 and continued to dominate state and local elections in Alabama from 1874 until 2000. The national Democratic Party's stance on civil rights issues had a lot to do with starting the transition in the 1960s. But the story of political realignment in Alabama is a long and complicated one,

After the Civil Rights Act of 1964, continued Dr. McKee, the passage of the Voting Rights Act in 1965 was the next major event in political realignment. Southern Democrats who had opposed civil rights had to decide whether to embrace the new African-American voters in their party or make a switch. According to Dr. McKee, this is what happened: Southern Democrats figured out that they could keep winning if they made separate overtures to white and black voters. From the mid-1970s to the early 1990s, Democrats were able to win with biracial coalitions in Alabama through "stealth reconstruction" — preaching socially conservative positions to white voters and civil rights issues to African-American voters. Stealth reconstruction blocked Republican successes for decades. Trial Tr. VI, at 152, 154-56 (McKee). This theory is borrowed from the writings of Democrat Glenn Browder, a former Alabama Secretary of State and congressman. Trial Tr. VI, at 155 (McKee). The court does not credit it.

Dr. McKee's narrative is inconsistent. First, he says that generational changes in voter ideology ended Democrats' ability to form biracial coalitions. As younger generations of white southerners came of voting age, "they were much better aligned, in terms of conservative values and conservative views, . . . with the Republican Party. And Ronald Reagan was clearly one of those moments where it became that much clearer to any white southerner which party they should belong to." Trial Tr. VI, at 157 (emphasis added). These younger generations started off as either independents or Republicans, while their parents had been Democrats and became Republicans, if at all, later. Trial Tr. VI, at 157.

Dr. McKee goes on to argue concepts of "relative advantage" and "group conflict theory." According to Dr. McKee, "relative advantage" and "group conflict theory" explain the causative relationship in Alabama between "black growth and then the response in terms of white Republicanism." Trial Tr. VI, at 150, 163. As African Americans "become a greater presence in the Democratic Party," whites, feeling their "white power" threatened, leave the Democratic Party to take over the Republican Party so they can maintain power without the need to form biracial coalitions. Trial Tr. VI, at 161-62 (McKee). Oddly, the relative advantage and group conflict theories do not explain why whites supported black candidates, Justice Oscar Adams, who was elected twice, and Justice Ralph Cook, in the 1980s and 1990s. Nor does Dr. McKee explain why these theories leapt over the Adams/Cook successes, delaying twenty years (the 1980s and 1990s) before coming into influence in Alabama in 2000.

But Dr. McKee also said that younger white voters, who had driven the realignment, "recognized the better fit of issue alignment between a conservative ideology and the contemporary Republican Party." McKee Rebuttal Report, at 13 (Pls.' Trial Ex. 123; Doc. # 88) (emphasis added). Although relative advantage may explain why some older whites left the Democratic Party, it does not explain why younger, conservative white voters, who drove the realignment (and who were never Democrats), became reliable Republicans. As Dr. McKee testified, conservative ideology was the driving force. Trial Tr. VI, at 157.

Despite this testimony, when confronted with the possibility that conservative ideology, not race, determines party preference, Dr. McKee contradicted himself, repeatedly rejecting the notion that Alabama voters know which political party reflects their values. Trial Tr. VI, at 196-97 (McKee) ("If I were to give you a random set of issues that mattered to most voters, they couldn't connect the dots and tell you what the conservative and liberal position is."); Trial Tr. VI, at 207 (McKee) ("Most voters in Alabama cannot do what you're asking me. They cannot correctly identify the alignment of those parties with those issues.") This testimonial turnaround, and the way it was delivered and received into evidence "live," left no doubt that Forrest Gump had entered the courtroom as a hapless, confused Alabama voter — but confused only on party preference. Forrest was sure of his ideology, but not his party; yet, according to Dr. McKee, Forrest was capable of "using the concepts of 'relative advantage and group conflict theory'" in order to maintain power without the need to form biracial coalitions. Because of these inconsistencies and because Dr. McKee offered no evidence that he knows even one registered Alabama voter other than Plaintiffs and counsel, his testimony has no credibility on the motivations of Alabama voters.

Strongly to the contrary, the evidence shows no issues or party confusion among Alabama voters. Because voters must approve constitutional amendments on a statewide basis, the results of voting on those amendments provide a snapshot into Alabamians' ideology.⁴⁸ Here are a few examples:

48 To the extent the State objects to the court's use of statewide amendment results without the aid of expert testimony, that objection is sustained. As demonstrated below, the court has only viewed those results under the guidance of one strain of Dr. McKee's testimony about voter behavior.

• Dr. McKee acknowledged that people recognize the GOP as a pro-life party. Trial Tr. VI, at 207. In 2018, voters approved by a wide margin amending the state constitution to provide that the State "supports the sanctity of unborn life and does not protect abortion." 2018 Election Statewide Amends., at 64 (Defs.' Trial Ex. 103; Doc. # 151-1).

• Dr. McKee acknowledged that the concept of "right to work" is a recognizably Republican position. Trial Tr. VI, at 211. In 2016, voters approved a "right-to-work" amendment by a wide margin. 2016 Election Statewide Amends., at 62 (Defs.' Trial Ex. 103; Doc. # 151-1).

• Dr. McKee acknowledged that people recognize the Republican Party as a pro-Second Amendment party. Trial Tr. VI, at 207. In 2014, voters approved by a wide margin an amendment that "protects [the] right to bear arms" and subjects any restriction of that right to strict scrutiny. 2014 Election Statewide Amends., at 61 (Defs.' Trial Ex. 103; Doc. # 151-1).

• Dr. McKee acknowledged that a platform plank supporting "[t]raditional marriage and family" is a recognizably Republican position. Trial Tr. VI, at 211. In 2006, Alabama voters approved by a wide margin a constitutional amendment defining marriage as between a man and a woman. 2006 Election Statewide Amends., at 55 (Defs.' Trial Ex. 103; Doc. # 151-1); Ala. Const. art. I, § 36.03, ruled unconstitutional by Obergefell v. Hodges, 135 S. Ct. 2584 (2015).

The court finds that Dr. McKee painted all Alabama voters with a broad stroke. Contrary to his assessment, a large majority of Alabamians have no confusion about which political party agrees with them on these issues.

Dr. McKee thought that political issues that are neutral on their face may become "racialized." His only example, itself a "maybe," is welfare, which he says (with no empirical evidence) many whites view "as being given disproportionately to undeserving blacks." Trial Tr. VI, at 173 (McKee). He does not explain how issues like abortion, the Second Amendment, the "right to work," or traditional marriage and family are viewed by Alabama voters as racial issues. And the court finds no evidence to support a racial component to them.

The empirical evidence for Dr. McKee's subjective opinion that most Alabama voters view the Democratic Party as the party of African Americans and the Republican Party as the party of whites is equally unconvincing. He gives conclusory support for that opinion with reliance on "voting behavior, party affiliation, and the racial composition of partisan officeholders," McKee Rebuttal Report, at 18-19 (Pls.' Trial Ex. 123; Doc. # 88), but he never connects the evidentiary dots. It is not news that, in Alabama and in the Deep South generally, African Americans "constitute the majority of the Democratic Party's voters and officeholders in federal and state legislative positions." McKee Rebuttal Report, at 19. But nowhere does Dr. McKee point to evidence that white voters are sensitive to that fact, let alone that the racial composition of the Democratic Party, as opposed to the actual issues and policies promoted by the party, determines how whites perceive it. Moreover, whites in Alabama vote Democratic in significant, if not controlling, numbers, as evidenced by polarization percentages discussed supra, and by recent election successes (Sue Bell Cobb and Doug Jones) and near misses (Deborah Bell Paseur and Robert Vance.).

As Dr. McKee acknowledged on cross-examination, one cannot know why individuals voted the way they did without asking them, and there are no such surveys in his report. Trial Tr. VI, at 203-04; see also Bonneau Reply Report, at 1-2 (Defs.' Trial Ex. 26; Doc. # 93-1). "[I]nferring motivations from observed behavior" is a tricky business, as is assessing the influence of race, which cannot be varied for experimental purposes, as a causal variable. Bonneau Reply Report, at 4, ¶ 6; Trial Tr. V, at 74 (Bonneau).

The second flaw in Dr. McKee's opinion is that this is a case about appellate judicial elections in Alabama, and he considers no evidence specific to partisan realignment on Alabama's appellate courts. Gingles calls for an "intensely local appraisal" of the electoral system, 478 U.S. at 79 (citation and internal quotation marks omitted), along with a "functional" view of the political process, id. at 48 n.15. The court must therefore look beyond elections for high-profile state and federal offices, which form the main part of Dr. McKee's analysis, to see how partisan realignment played out in down-ballot, lower-profile state appellate judicial races.

Tort reform played a key role in the transition from an all-Democrat to all-Republican Supreme Court beginning in the 1990s.⁴⁹ Excessive jury verdicts and plaintiff-friendly judging gained Alabama a national reputation as "tort hell" for business defendants. The case of BMW of North America, Inc. v. Gore, 517 U.S. 559 (1996), in which a Jefferson County jury awarded $4 million in punitive damages for a $4,000 undisclosed repainting of a new sports car, became the poster child for Alabama's reputation for "jackpot justice." Gaylord Reply Report, at 6 (Defs.' Trial Ex. 27; Doc. # 93-2); Trial Tr. VI, at 9-10 (Murdock). On appeal, the Alabama Supreme Court remitted the punitive damages award to $2 million, but with little explanation for upholding an award that was still 500 times the size of the compensatory damages award. See Gore, 517 U.S. at 567. Subsequently, the U.S. Supreme Court struck down the $2 million punitive damages award as "grossly excessive" in violation of the Due Process Clause of the Fourteenth Amendment. See id. at 574-86. The expansion of Alabama tort doctrines led the defense bar to nickname the Alabama Supreme Court the "Tort of the Month Club." Trial Tr. VI, at 10 (Murdock); see, e.g., Johnson v. State Farm Ins. Co., 587 So. 2d 974 (Ala. 1991) (adopting a more lenient standard of "justifiable reliance" for a plaintiff to prove fraud).

49 To give context to the Republican switch, after 1874, no Republican nominee for the Alabama Supreme Court won the general election until Perry Hooper in 1994. In 1970, Democrats were elected to all statewide offices in Alabama, and the Republican Party fielded only one candidate for statewide office. In 1972, only Democrats appeared on the ballot for appellate judicial offices. In 1974, Democrats were elected to all statewide offices in Alabama, and the Republican Party fielded no candidates for statewide office. In 1978, Democrats were elected to all statewide offices. Republicans fielded no candidates for appellate judicial offices. Again, in 1980, Republicans fielded no candidates for appellate judicial offices. See Post-Trial Stipulations, at 32-33, ¶¶ 160-65 (Doc. # 176).

In response, the legislature passed tort reform measures in 1987, including a cap on punitive damages. Many of those measures were challenged and struck down by the Democratic Alabama Supreme Court. See, e.g., Henderson v. Ala. Power Co., 627 So. 2d 878, 893-94 (Ala. 1993) (striking down the punitive damages cap for violating the right to trial by jury).

Business interests decided it was time to do something about "tort hell." In 1994, the Business Council of Alabama hired political consultant Karl Rove, who had successfully advised pro-tort reform judicial election campaigns in Texas, to do the same in Alabama. Gaylord Reply Report, at 5-6 (Defs.' Trial Ex. 27; Doc. # 93-2); see also Trial Tr. III, at 100-04 (Gaylord); Trial Tr. I, at 135-36 (White); Trial Tr. VI, at 11-12 (Murdock). The 1994 campaign was characterized by "the same pro-business, tort reform themes that proved successful in Texas." Gaylord Reply Report, at 5 (Defs.' Trial Ex. 27; Doc. # 93-2); see also Trial Tr. IV, at 69-70 (Thibodeaux) (explaining that, "[i]n 1996, it was the height of tort reform in Louisiana" with statewide conflicts among the state business counsel, the oil and gas groups, and trial lawyers). Republican challengers emphasized that Democratic incumbents had been soliciting and receiving donations from trial lawyers who were appearing in front of them. Gaylord Reply Report, at 7-8 (Defs.' Trial Ex. 27; Doc. # 93-2).

Their strategy worked. Hugh Maddox, a conservative Democrat incumbent who defeated a trial lawyer-backed Democrat in the primary, and Perry Hooper, Sr., a Republican, were elected to the Alabama Supreme Court in 1994. Gaylord Reply Report, at 5, 8 (Defs.' Trial Ex. 27; Doc. # 93-2). Hooper was the first Republican elected to the Alabama Supreme Court in at least 129 years. He defeated incumbent Chief Justice Sonny Hornsby, a former president of the Alabama Trial Lawyers' Association, who "was widely regarded as sort of the leader of the Court when it came to striking down tort reform," Trial Tr. VI, at 13 (Murdock), in a ferociously contested election that was not decided until a year later, Trial Tr. VI, at 26 (Murdock).⁵⁰ These Republican successes were followed by the GOP gaining Supreme Court seats with Harold See's 1996 election and Champ Lyons's 1998 appointment. Gaylord Reply Report, at 8 (Defs.' Trial Ex. 27; Doc. # 93-2); Trial Tr. VI, at 12 (Murdock). In 2000, Alabama voters elected three first-time Republican candidates — Lyn Stuart, Tom Woodall, and Bernard Harwood (who testified in this case) — to give the GOP a solid majority on the Alabama Supreme Court. Gaylord Reply Report, at 8-9. After the 2004 election cycle, the Republican takeover was complete: The court had gone from 9-0 Democratic to 9-0 Republican in just ten years. Gaylord Reply Report, at 9.

50 The rest of Karl Rove's "slate" in 1994 court races lost. Gaylord Reply Report, at 8 (Defs.' Trial Ex. 27; Doc. # 93-2); Trial Tr. VI, at 117 (Murdock).

Alabama was not alone in this transformation. The same was occurring in Louisiana and Texas. Trial Tr. IV, at 69-70 (Thibodeaux); Gaylord Reply Report, at 4-5 (Defs.' Trial Ex. 27; Doc. # 93-2); see also LULAC, 548 U.S. at 411-12 (explaining that "[t]he 1990's were years of continued growth for the Texas Republican Party," that "by the end of the decade it was sweeping elections for statewide office," and that, in 2003 "Texas Republicans" had gained control of both houses of the state legislature).

During this era, Alabama judicial elections were the battleground for plaintiffs' lawyers and business groups. Trial Tr. I, at 148 (White). Campaign spending increased dramatically, and Alabama judicial races are now among the most expensive in the country. Gaylord Reply Report, at 10 (Defs.' Trial Ex. 27; Doc. # 93-2); Trial Tr. V, at 136 (Bonneau). The elections during the "tort war" years were characterized by recognizably conservative themes of pro-business, ending judicial overreach, and defending the rule of law. See Gaylord Reply Report, at 9. The fight over tort reform was a major cause of the transition from an all-Democratic to an all-Republican Alabama Supreme Court. See Gaylord Reply Report; see also Trial Tr. II, at 29 (Travis) (agreeing that tort reform was "the cause" or "a primary cause" of the flip from Democrat to Republican on the court). That Dr. McKee does not consider this localized evidence at all — indeed, his demeanor and comments in court indicated he disdained the very idea — further undermines the credibility of his opinions on Alabamians' voter behavior.

The contentious 1994 race for chief justice also fueled the party switch on the Alabama Supreme Court, though to a lesser extent. The Hooper-Hornsby race, which lasted about a year and garnered extensive media attention across the state, put voter fraud and judicial integrity on voters' minds. Trial Tr. VI, at 23-27 (Murdock); see Joint Tr. Exs. D-8 through D-20.

On election night in 1994, Hornsby's lead was about 600 votes, prompting a recount. Counties began to correct their vote totals, and Hooper gained a 262-vote lead. Trial Tr. VI, at 15 (Murdock). Not included in this count were some 1,700 absentee ballots, mostly from the Democratic Black Belt, that were not counted because they were not witnessed and notarized as required by state statute. Newspapers and pundits speculated those absentee-ballot votes would fall heavily for incumbent Democrat Chief Justice Sonny Hornsby. The parties took the fight to court. Hornsby's team filed a lawsuit in state court to validate these absentee ballots. For eleven months the issue was kicked between state and federal courts. Initially, a Democratic state-court judge agreed with Hornsby that the ballots must be counted because the ballots substantially complied with the statutory requirements. Hooper's team responded with a federal action in Mobile to prevent the counting of the illegal ballots. Their legal theory in federal district court was that it violated due process to have one rule about what ballots should count before the election and another one after the election. Trial Tr. VI, at 17-21 (Murdock).

In the appeal of the federal district court's ruling preliminarily enjoining the Alabama Secretary of State and election officials from complying with the state court order to count the contested absentee ballots, the Eleventh Circuit certified a question to the post-election Alabama Supreme Court, asking whether the contested ballots should be counted. See Roe v. Alabama, 43 F.3d 574, 583 (11th Cir. 1995). The Alabama Supreme Court held, 6-1, that substantial compliance with the statute was enough for the ballots to be valid. See Roe v. Mobile Cty. Appointment Bd., 676 So. 2d 1206, 1226 (Ala. 1995). In what can easily be described in hindsight as a shortsighted opinion heavy with political rhetoric, the Supreme Court — sans Chief Justice Hornsby but including four justices who not only were his colleagues and fellow Democrats, but who had all contributed to Hornsby's campaign and he to at least one of theirs (Justice Ralph Cook) — took some forty-six pages to blast the whole effort by the Republicans, telling the federal appeals court that the ballots should be counted. Newly elected Justice Ralph Cook wrote a stinging concurrence. Two justices recused. See id. at 1234-42 (statement of Justice Houston explaining his recusal decision). And two did not participate for unexplained reasons. The justices who did not recuse, including Justice Ralph Cook, took heat for participating in a decision that could be decisive in whether their colleague was reelected. Trial Tr. VI, at 25 (Murdock). Not satisfied with the Alabama Supreme Court's answer, the Eleventh Circuit Court of Appeals remanded the case to U.S. District Judge Alex T. Howard in Mobile to have an evidentiary hearing on the merits. The federal district court found for the Republican challenger, and that decision was affirmed. See Roe v. Mobile Cty. Appointing Bd., 904 F. Supp. 1315, 1336 (S.D. Ala.), aff'd sub nom. Roe v. Alabama, 68 F.3d 404 (11th Cir. 1995).

That is how Hooper ended up winning the election, but the issue of voter fraud remained. Governor Fob James called the legislature into a special session in 1996 on the issue of election reform. Glenn Murdock (later Justice Murdock), who had been part of Hooper's legal team, would use the issue of election integrity in his own 1998 and 2000 judicial races. Not until Republicans took over the Alabama Supreme Court about five years later did voter fraud subside as a significant concern in judicial elections. Trial Tr. VI, at 29-31, 89-93 (Murdock).⁵¹

51 There were several significant voter fraud convictions during this period. Justice Murdock testified about an absentee-ballot buying operation that resulted in voter-fraud convictions against prominent Republicans in Winston County. In another case, two individuals forged 600 absentee ballots and were convicted on voter-fraud charges in Wilcox County in the wake of the 1994 election. Trial Tr. VI, at 29, 89-93 (Murdock). And a Democratic U.S. Attorney and Republican Alabama Attorney General jointly investigated absentee ballot fraud after five suitcases containing forged absentee ballots were recovered at the Eutaw, Alabama post office on election day in 1994. Convictions were obtained against the individuals who were responsible for the "ballot box stuffing party at the Greene County Jail" and the delivery of the suitcases. Trial Tr. VI, at 91 (Murdock).

While the effects of the Hooper-Hornsby election were relatively short term, Justice Murdock gave credible testimony on the role it played in the politics of judicial elections during the transition period and the effect the contentious judicial race had on the flip from Democrats to Republicans on the appellate bench. See, e.g., Trial Tr. VI, at 23, 25-26 (testifying that "the election of [a] chief justice with the potential for the first Republican since Reconstruction . . . was in all the papers," that "[i]t was front and center," id. at 23, and that the six or seven justices who did not recuse when the Eleventh Circuit certified a dispositive question of state law to the Alabama Supreme Court "took heat" from the voters "for participating in a case about the colleague [whom] they worked with"). Dr. McKee did not consider this evidence in forming his opinion that race drives election results in Alabama. This failure to consider relevant history and his flippant remarks during his testimony further undermine his credibility. See, e.g., Trial Tr. VI, at 196 (McKee) ("I'm starting to laugh when I hear Hooper/Hornsby again.").

A witness for Plaintiffs testified that he believed that an earlier event — the politically heated 1986 Baxley/Graddick Democratic primary runoff — was the catalyst for the eventual switch in Alabama to the one-party Republican state that he believes Alabama is today. Trial Tr. IV, at 148-49 (Jerome Gray); see supra Part III (discussing the 1986 gubernatorial election of Republican Guy Hunt, who defeated Democratic challenger, William J. Baxley). The court finds, at a minimum, that the 1986 governor's race left Alabama voters highly sensitized to party differences, contributing to a civic awakening that bloomed in the 1990s.

vii. Weighing of Senate Factor 2

Plaintiffs' statistical evidence accounted only for the fact of racially polarized voting, not its cause. Defendants have presented evidence — bearing on the totality of circumstances — of reasons why black-preferred candidates are losing that are unrelated to race. The State's evidence is not conclusive, but it well supports the idea that African Americans are not losing appellate judicial elections "on account of race or color." § 10301(a). Senate factor 2 weighs heavily in favor of the State.

c. Senate Factor 3: Election Devices that Enhance Potential for Discrimination

The third Senate factor asks whether the State "has used unusually large election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority group." S. Rep. No. 97-417, at 29. Plaintiffs point to three election procedures in Alabama's judicial elections they say potentially enhance the opportunity for discrimination. First, the numbered-post requirement: In 1927, the Alabama legislature passed a numbered-post requirement for the appellate courts. Post-Trial Stipulations, at 6, ¶ 37 (Doc. # 176). A numbered-post requirement effectively eliminates single-shot voting because voters may elect only one candidate per post. But, as discussed more fully in the analysis of Plaintiffs' constitutional claim, see infra Part V.B., the numbered-post law was about giving advantage to incumbents against Klan-backed factions within the Democratic Party, not about racial discrimination. See SCLC, 56 F.3d at 1285-86.

Moreover, the State has an important interest in ensuring incumbents are not required to run against each other, which would negatively affect collegiality on the appellate courts. See S. Christian Leadership Conference of Alabama v. Evans, 785 F. Supp. 1469, 1476 (M.D. Ala. 1992), af f' d sub nom. SCLC v. Sessions, 56 F.3d 1281 (11th Cir. 1995) (en banc) ("Certainly, Alabama should be allowed to have a system for electing judges where they do not compete against each other at election time."). SCLC dealt with trial judges. The State has an even stronger interest in keeping members of an appellate court, who make decisions in tandem, from having to run against each other. Indeed, Plaintiffs have identified no states that use at-large elections for multimember courts that do not use numbered places.

Second, Plaintiffs argue that Alabama's primary election procedures are potentially discriminatory. In 1931, the legislature adopted a statute pertaining only to primary elections that eliminated second-choice and single-shot voting and required the winning candidate to receive a majority of the vote. Post-Trial Stipulations, at 7, ¶ 38 (Doc. # 176); Trial Tr. III, at 253 (Norrell). But there was no evidence that these procedures reduce the potential for African-American candidates to win party primaries. In fact, the evidence showed that, since 2000, African-American candidates routinely win the Democratic Party primary for appellate judicial races.⁵²

52 In 2000, Ralph Cook, John England, and Aubrey Ford were the Democratic nominees for their respective races. In 2002, Vicky Underwood Toles was the Democratic nominee for a place on the Court of Civil Appeals. In 2006, Gwendolyn Kennedy, Aubrey Ford, and John England won the Democratic nominations for their respective races. In 2008, Clyde Jones was the Democratic nominee for a seat on the Court of Criminal Appeals. All were African Americans. See Handley Am. Expert Report (Tables 1-4). And, although Yvonne Saxon, an African American, lost the Democratic nomination for Court of Criminal Appeals in 2000, she was not the black-preferred candidate: Sue Bell Cobb (who also won the general election) was. Handley Am. Expert Report (Table 5). No African American has run for a seat on any of Alabama's appellate courts since 2008. Post-Trial Stipulations, at 39, ¶ 203 (Doc. # 176).

Third, and finally, Plaintiffs argue that Alabama's adopting at-large, statewide elections is an example of an "unusually large election district." Plaintiffs cite no legal authority that statewide elections are implicated by this aspect of Senate factor 3,⁵³ and such an interpretation abuses the normal use of the term "district." A "district" is "[o]ne of the territorial areas into which an entire state or country, county, municipality or other political subdivision is divided, for judicial, political, electoral, or administrative purposes." District, Black's Law Dictionary (5th ed. 1979) (emphasis added). District implies a division that has already occurred. That is clearly not the case with statewide elections. Dr. McCrary's observation amounts to a redefinition of the word, and it is not credited. Moreover, concerns that a state could manipulate the size of districts to dilute minority votes are simply not present when the "district" encompasses the entire state. This is a significant definitional blow to the dilution argument.

53 Dr. McCrary testified that a statewide election could be considered an "unusually large election district," Trial Tr. III, at 30-31, but cited no authority for the idea that Congress contemplated statewide elections for purposes of this factor. In fact, he admitted that statewide elections were "not a part of what the Congress had in mind in terms of what's reflected in the record of the House and Senate hearings in 1981 and '82." Trial Tr. III, at 31. The only other federal court to have ruled directly on a challenge to at-large, statewide elections for appellate judges observed that, in Texas, "[t]he size of the election district is a function of the size of the state." Lopez, 339 F. Supp. 3d at 614. In Lopez, the size of the election district was a neutral factor in the analysis of the totality of circumstances. See id. However, the State of Texas's linkage interest in at-large, statewide elections weighed heavily in favor of the State. Id. at 619.

Furthermore, Alabama adopted statewide elections for appellate judges in 1868 to increase the opportunities for African Americans to elect their preferred candidates, and there is insufficient evidence to conclude that this practice was maintained for a discriminatory purpose. See supra Part V.A.4.a.i. and infra Part V.B. And, as discussed elsewhere, the State has a strong interest in linking appellate judges' jurisdiction with their electoral base. See infra Part V.A.7.d.; Lopez, 339 F. Supp. 3d at 614 (noting the state's linkage interest for purposes of Senate factor 3).

Plaintiffs also argue that statewide elections are more expensive, so socioeconomic disparities between whites and African Americans would be felt more acutely in those elections. Without citing empirical support, Dr. McCrary opined as much. See McCrary Expert Report, at 22-23 (Pls.' Trial Ex. 122; Doc. # 69-1). But Dr. Bonneau, who performed an empirical study of the issue, testified that statewide and districted judicial elections were equally expensive. Trial Tr. V, at 49-50 (Bonneau). Dr. Bonneau's view is credited.

In short, although Alabama has used discriminatory election devices in the past, there is insufficient evidence that any current procedures in Alabama appellate judicial elections were adopted or maintained for discriminatory reasons. Based on this lack of evidence, the State has a weighty interest in maintaining those procedures. Additionally, there is an absence of evidence that the 1931 changes to party primaries have a discriminatory effect. For these reasons, Senate factor 3 is weighed moderately toward a finding that Alabama's appellate judicial elections are equally open to participation by African Americans.

d. Senate Factor 4: Slating

No formal slating process exists in Alabama. Nonetheless, Plaintiffs argue that, in the mid-1990s, Karl Rove effectively "slated" white candidates for Alabama judicial races. Mark White, who ran campaigns for Justices Oscar Adams and Ralph Cook during this period, testified that business leaders, unhappy with some Alabama Supreme Court decisions, privately met with Rove in 1994 to select Republican judicial candidates. According to Mr. White, none of the meeting's participants or the selected candidates was African American. Trial Tr. I, at 135-37 (White).⁵⁴

54 It is unclear from the record whether Mr. White was present at this meeting or simply heard about it after the fact.

Plaintiffs' argument is unconvincing. Broadly speaking, "access to slating" refers to "the ability to run for office." United States v. Marengo Cty. Comm'n, 731 F.2d 1546, 1569 (11th Cir. 1984). Candidate slating may occur when an influential group of political leaders selects candidates that it recommends to voters as a slate. Trial Tr. III, at 52-54 (McCrary). For example, in White v. Regester, the Court found it probative that the Dallas [Texas] Committee for Responsible Government, "a white-dominated organization that [was] in effective control of Democratic Party candidate slating in Dallas County," had only ever slated two African-American candidates. 412 U.S. 755, 766-67 (1973).

Even assuming an out-of-state political consultant's spearheading candidate-recruitment efforts in one or two election cycles constitutes "slating," Plaintiffs have not shown that African Americans were denied access to this process or that it was designed "to winnow out candidates." Hall v. Holder, 117 F.3d 1222, 1227 (11th Cir. 1997). Plaintiffs produced no evidence that African Americans are unable to compete in either party's nominating process or to receive political endorsements from influential groups. Indeed, several witnesses were entirely dismissive of the idea. Having attended meetings of the Alabama Republican Party Executive Committee for twenty-five years, Justice Glenn Murdock testified that the Alabama Republican Party actively recruits African-American candidates. Trial Tr. VI, at 57, 101. Justice Ralph Cook, one of two African Americans to be elected to the Alabama Supreme Court, won in 1994 with the support of the business community. Trial Tr. IV, at 92 (Cook); Trial Tr. VI, at 36 (Murdock). In terms of the ability to run for office, "there does not appear to have been any substantial formal or informal impediment to black candidacies." Marengo Cty. Comm'n, 731 F.2d at 1569. This factor weighs toward a finding that Alabama's judicial races are equally open to participation by African Americans. See § 10301(b).

e. Senate Factor 6: Racial Appeals

The sixth Senate factor addresses "whether political campaigns have been characterized by overt or subtle racial appeals." S. Rep. No. 17-497, at 29 (emphasis added). "As overtly bigoted behavior has become more unfashionable, evidence of intent has become harder to find." Marengo Cty. Comm'n, 731 F.2d at 1571 (quoting Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1290 (7th Cir. 1977)). Thus, while "[e]vidence of racism can be very significant," the absence of evidence supporting this factor "should not weigh heavily against a plaintiff proceeding under the results test of section 2."⁵⁵ Id.

55 But since Marengo and Village of Arlington Heights, bigoted racial behavior has been supplanted by bitter partisan behavior. That is a topic that eludes a consensus, reasoned explanation. But it tends to support a finding that partisanship has much to do with recent election outcomes generally.

Plaintiffs produced evidence of only two candidates in Alabama elections whose ads contained racial appeals. One was in the 2018 election for Chief Justice of the Alabama Supreme Court. In one of his campaign commercials, Chief Justice Tom Parker, a white Republican running against a white Democrat, Robert Vance, Jr., touted his opposition to Roe v. Wade, 410 U.S. 113 (1973) and Obergefell v. Hodges, 135 S. Ct. 2584 (2015), and stated that he had "taken on and beaten the Southern Poverty Law Center" ("SPLC"). Pls.' Trial Ex. 135. See generally Parker v. Judicial Inquiry Comm'n of Ala., 295 F. Supp. 3d 1292 (M.D. Ala. 2018) (issuing a preliminary injunction against the judicial inquiry commission's investigation, prompted by the SPLC, into Parker's criticism of Obergefell v. Hodges, 135 S. Ct. 2584 (2015)). The ad says that Parker opposes "the leftist mob tr[ying] to destroy our society" and features a clip of Congresswoman Maxine Waters, a controversial African-American Democrat from California. See Pls.' Trial Ex. 135.

The SPLC is known primarily for its advocacy on race issues. See Trial Tr. III, at 228 (Hall); Trial Tr. IV, at 113-14 (Cook). When juxtaposed with images of an African-American Democratic congresswoman from California who had no other reason to appear in an ad for an Alabama judicial race, the court concludes that one of the motives of the ad was to draw attention to race. Representative Laura Hall and Justice Cook came to the same conclusion. Trial Tr. III, at 227-28 (Hall); Trial Tr. IV, at 113-15 (Cook).

Another of Justice Parker's ads is racially tinged, but mostly toward illegal aliens. This commercial begins, "It's an invasion. What happens if they make it to Alabama?" and shows what appear to be non-citizens attempting to cross the border. It then turns to Parker's opponent, Vance, "a liberal social justice activist who will always side with the illegal alien lobby" and who is supported by George Soros, the SPLC, the ACLU, Planned Parenthood, and "every liberal newspaper in Alabama." The ad concludes that Justice Parker "stand[s] up for what we believe" and "stand[s] with us." See Pls.' Trial Ex. 136 (emphasis added).

The "illegal-alien invasion" ad prominently features an issue whose relevance to an Alabama judicial race is hard to see objectively, but the racial element is suggested in the context of the attacks on the SPLC and signaling to "we" and "us."⁵⁶ While invoking the SPLC, illegal immigration, and Maxine Waters may not be objectively racial in other political contexts, these gratuitous references were made in a race for Chief Justice of the Alabama Supreme Court, which supports an inference that race was involved.

56 Anecdotally, Representative Hall also testified that, against the backdrop of the whole commercial, she interpreted Parker's slogan, "I am fighting for us," to mean "fighting against black folks" and "fighting for the white people." Trial Tr. III, at 228 (Hall).

The "we" and "us" signals have been used before in Alabama. Ironically, it was in a racial appeal to white voters that was run by an African-American candidate. Plaintiffs admit that Justice Cook's 1994 campaign commercial featuring NASCAR driver Bobby Allison telling voters that Cook is "one of us" was a racial appeal to white voters. Trial Tr. I, at 126-28 (White); Trial Tr. IV, at 88-90 (Cook); Pls. Proposed Findings of Fact, at 72, ¶ 29 (Doc. # 176) ("Judge Cook in 1994 ran a campaign commercial in which famous race car driver Bobby Allison, who was White, assured viewers that Judge Cook is 'one of us,' which was understood as a racial appeal to White voters." (emphasis added)).⁵⁷

57 The court does not necessarily discount the possibility of a "reverse" racial appeal where a minority candidate plays on white prejudice to his or her political advantage. After all, it may indicate race consciousness in the electorate. But Plaintiffs have cited no authority supporting this theory. And without more context, it is not clear, despite the admission, that this was a racial appeal. As part of his "stealth campaign," Trial Tr. IV, at 88, 92, the ad did not feature Cook's picture. It is therefore likely that, in a lower-profile election, many voters were unaware of his race, and there would thus be no apparent racial element to the ad. Compare Jordan v. Winter, 604 F. Supp. 807, 813 & n.8 (N.D. Miss.) (citing as a "racial campaign tactic" a "campaign television commercial sponsored by the white candidate whose slogan was 'He's one of us' [that] opened and closed with a view of Confederate monuments accompanied by [an] audio message" that was racially charged), aff'd sub nom. Miss. Republican Exec. Comm. v. Brooks, 469 U.S. 1002 (1984), with Ga. State Conf. of NAACP v. Fayette Cty. Bd. of Comm'rs, 950 F. Supp. 2d 1294, 1319 (N.D. Ga. 2013) (candidate's statement that he was running to "maintain and preserve the heritage we have in our county" was not, by itself, a racial appeal), rev'd in part, vacated in part on other grounds, 775 F.3d 1336 (11th Cir. 2015); see also Clements, 999 F.2d at 879 (observing that the African-American candidate against whom a racial appeal had been used ended up winning the election).

Other than those ads, there was no credible evidence that there have been racial appeals in appellate judicial races in at least the past twenty-five years. Plaintiffs' characterization of the other examples of racial appeals is unpersuasive. First, Plaintiffs argue that the Republican Party omitted African-American candidate Greg Griffin's picture from its campaign literature to signal to white voters that the GOP was "branding itself as a White party." Pls.' Trial Br., at 41 (Doc. # 167); Trial Tr. I, at 57-58 (England). (Griffin ran for the Court of Criminal Appeals in 1994, and it appears from the record that he is the only African-American Republican to have run for a seat on one of Alabama's appellate courts.) This is an upside down, facially illogical argument. Plaintiffs condemn that as a "branding" effort, when Justice Cook and Justice England, both African-American Democrats, did the same thing in campaigns. Trial Tr. I, at 36 (England) (1994 campaign for circuit judge); Trial Tr. IV, at 88, 92 (Cook) (1994 campaign for chief justice).

Three ads in two statewide campaigns in 24 years — two by a 2018 white candidate and one by a black candidate directed to white voters — cannot be said to make that conduct characteristic of judicial campaigns. Three bricks do not a house make.

The rest of Plaintiffs' arguments have no traction. They argue that Lyn Stuart, Justice Cook's opponent in the 2000 race for Alabama Supreme Court, made a racial appeal when she ran a commercial touting her being "tough on crime" and having sentenced defendants to death as a circuit judge. Trial Tr. IV, at 96-97 (Cook). The court finds that this was not a racial appeal. There is no evidence that voters thought this was a racial appeal. Jerome Gray, who retired from the Alabama Democratic Conference in 2006 after serving nearly three decades as a field director, said he would remember if there had been racial appeals in this judicial race and that he did not recall any. Trial Tr. IV, at 140-41, 150 (Gray). Indeed, Justice Oscar Adams made a similar pitch to voters in his 1988 campaign: "Oscar Adams believes the criminal must pay the price of his crimes. He has proven in his performance on the Court that he is tough on the criminal and is on the side of law enforcement in protecting our society." Justice Adams's Campaign Advertisement, circa 1982 (Joint Trial Ex. P-84). A candidate's being tough on crime is a legitimate concern for voters, especially in judicial races. Without more, there is no racial component to such an appeal.

Plaintiffs' other examples of racial appeals are not probative of current conditions because they are not recent and did not occur in statewide races. Until 1966, the Alabama Democratic Party's primary ballot stated, "White Supremacy, For the Right." Pls.' Trial Ex. 133; Trial Tr. III, at 266 (Norrell). No more explicit, abhorrent racial appeal could be imagined. But this message has not been on the ballot in half a century. It no longer has any signaling effect to voters, as African-American Alabamians overwhelmingly vote Democratic and a strong majority of white Alabamians vote Republican.

Plaintiffs point to implicit racial appeals in local judicial races. In 1994, Judge England's (unsuccessful) opponent for a circuit court judgeship in Tuscaloosa County ran ads stating that, as a civil rights leader, England "couldn't be fair." Trial Tr. I, at 38 (England). Judge England interpreted the ads as racial; there was no evidence that voters saw them that way. In 2000, Judge Jones's (successful) opponent in a local judicial race distributed fliers in white areas of Jefferson County stating that, if Jones won, Richard Arrington "would be running the courthouse." Trial Tr. I, at 76 (Jones). (Arrington was the first African-American mayor of Birmingham.) These appeals in Tuscaloosa and Jefferson counties, however, have slight weight because they are not recent, not intrinsically racial, and did not occur in statewide races. See Thomas v. Bryant, 366 F. Supp. 3d 786, 807 n.71, 808 n.74 (S.D. Miss. 2019) (giving less weight to racial appeals that are remote in time and occurred in exogenous elections).

Considering these facts, the court finds that Senate factor 6 — whether campaigns have been characterized by racial appeals — carries no weight in Plaintiffs' favor. Two candidates since 1980 — thirty-eight years — in 128 statewide judicial races producing no doubt hundreds of political ads, do not demonstrate a pattern, practice, or routine of racial appeals across the election landscape. See Clements, 999 F.2d at 879 (noting the lack of evidence that racial appeals "were anything more than isolated incidents"). Plaintiffs admit as much: Parker's ads "are virtually unprecedented in modern-day politics . . . ." Pls.' Proposed Findings of Fact, at 73, ¶ 80 (Doc. # 176) (emphasis added).

There is no evidence that Alabama political campaigns generally, nor statewide judicial races specifically, are characterized by racial appeals.

f. Senate Factor 7: Extent of Minority Success

The seventh Senate factor examines "the extent to which members of the minority group have been elected to public office in the jurisdiction."⁵⁸ Gingles, 478 U.S. at 37 (quoting S. Rep. No. 97-417, at 28-29). The relevant "jurisdiction," for purposes of this factor, is defined in § 2: It is either "the State" or a "political subdivision" of the State. § 10301(b) ("The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered . . . . " (emphasis added)). This case is about the entire State as the relevant political unit. The Eleventh Circuit has not addressed the State as a whole as the relevant political jurisdiction. It has, however, set the territorial parameter for a political subdivision of a state. See Stallings, 829 F.2d at 1560 (holding that "the extent to which minority candidates have been elected to office in the jurisdiction, obviously must be measured by elections in the political jurisdiction in question, in this case Carroll County," which was defending a § 2 vote dilution claim to its at-large electoral system for its commissioner). The Senate had in mind districted, legislative elections, not at-large, statewide judicial elections, in discussing this factor. See supra note 53; see also Nipper, 39 F.3d at 1529 ("[S]erious problems [may] lie ahead in applying the 'totality of circumstances' factors developed in section 2 cases involving legislative elections to the election of judges." (quoting Chisom, 501 U.S. at 403)). So the language, "The extent to which members of a protected class have been elected to office in the State . . . is one circumstance which may be considered" is ambiguous: Does it mean "in statewide elections," or does it mean "in all elections in the State?" There is a difference in statewide elections and elections statewide.⁵⁹

58 Section 2 uses "State or political subdivision"; this Senate factor changes the language to "jurisdiction." Section 2 controls. See infra note 59.

59 This is an example of the Senate Report language being at odds with the specific language of the Act. The Act says "the State or political subdivision"; the Senate Report says "in the jurisdiction," which is, of course, much less specific — more ambiguous — than the statute. The Stallings court quotes "in the jurisdiction" from the Senate Report, even though the Act passed by Congress and signed by the President says "political subdivision." The distinction is of little moment in the Stallings case; it was all about a political subdivision of a state, and nothing about a "jurisdiction" of any sort, as far as § 2 is concerned. But the reference to "State" in the statute — left out of the Senate Report — matters much in this case.

Opting for a plain language interpretation, and under compulsion to conduct an expansive consideration of all the circumstances, the court will resolve the ambiguity by analyzing both the results of at-large, statewide elections, for which there is ample evidence in the record, and evidence of successes of "members of a protected class" in elections throughout Alabama insofar as there is evidence in the record or legitimate inferences therefrom.⁶⁰

60 The federal courts' treatment of relevant elections under Senate factor 7 has varied. See, e.g., Ellen Katz et al., Documenting Discrimination in Voting: Judicial Findings Under Section 2 of the Voting Rights Act Since 1982 Final Report of the Voting Rights Initiative, University of Michigan Law School, 39 U. Mich. J.L. Reform 643, 720 (2006) ("Many courts compared minority electoral success in endogenous elections to elections for governing bodies not challenged in the same suit. For some courts, the success of minority candidates in exogenous elections was sufficient evidence of minority electoral success, even where minority candidates did not win any office in the challenged jurisdiction. Most, however, emphasized that such exogenous elections were less probative of electoral difficulty or success. Some courts accorded almost no weight to exogenous electoral evidence, and several appellate courts reversed district court decisions which found plaintiffs had failed to meet Factor 7 based on exogenous electoral success." (internal citations omitted) (collecting cases)).

The seventh Senate factor expressly focuses on the electoral successes of the "members of the minority group," indicating that the electoral success of black-preferred white candidates in statewide elections is not the proper consideration at this juncture. Presumably for this reason, the parties have not included the electoral successes of Sue Bell Cobb in two statewide appellate judicial races (2000 and 2006) or of Doug Jones in the recent statewide senate election (2017) — both of whom were the black-preferred candidates — in the seventh factor's count. Neither has the court.

This brings the discussion back to a deeper analysis of campaigns in which African Americans won at-large, statewide elections. As noted, only two African Americans have been elected to statewide office in Alabama, judicial or otherwise: the late Justice Oscar W. Adams, Jr. and Justice Ralph D. Cook. In both instances, the office was associate justice of the Alabama Supreme Court. Justice Adams was elected to two consecutive statewide terms, in 1982 and 1988. Significantly, in 1982, Justice Adams was the first African American ever to run for an appellate court seat in Alabama, and he won. Post-Trial Stipulations, at 33, ¶¶ 166-167, at 34 ¶ 172 (Doc. # 167).

In the 1982 Democratic primary, Justice Adams garnered more votes than his white opponents, Jim Zeigler and George Williams, and defeated Zeigler in the 1982 Democratic runoff. See Ala. Sec'y of State, Election Results Archive — State Supreme Court, https://www.sos.alabama.gov/alabama-votes/voter/election-data (last visited Feb. 5, 2020); see also SCLC, 56 F.3d at 1288 ("Justice Adams won reelection in statewide races against well-known white opponents in the 1982 Democratic primary, the 1982 primary runoff, and the 1988 general election."). Justice Adams ran unopposed in the 1982 general election. See Ala. Statewide Elections from 1970 to Present, at 13 (Doc. # 151-1). He also ran unopposed in the 1988 Democratic primary and went on to defeat his well-known, white Republican opponent (Harry Lyons) in the 1988 general election with 675,216 votes to Lyons's 499,058 votes. See id.

Justice Cook won statewide election to the Supreme Court in 1994. Trial Tr. IV, at 91; Post-Trial Stipulations, at 36, ¶ 187 (Doc. # 176). He ran unopposed in the 1994 Democratic primary. Trial Tr. IV, at 87 (Cook). In the 1994 general election, Justice Cook defeated his white Republican challenger. See Ala. Statewide Elections from 1970 to Present, at 22 (Defs.' Trial Ex. 103; Doc. # 151-1).

The only other African American to serve on the Alabama Supreme Court is Justice John H. England, Jr., but he was not elected. He reached the state's highest court in 1999 through a gubernatorial appointment and was unable to retain his seat, losing his race for reelection in 2000.⁶¹ Post-Trial Stipulations, at 37, ¶¶ 191, 194 (Doc. # 176).

61 Judge England ran again for a seat on the Alabama Supreme Court in 2006. He ran unopposed in the statewide Democratic primary, but he lost in the general election to his white Republican opponent. Post-Trial Stipulations at 39, ¶ 202 (Doc. # 176).

Plaintiffs contend that not only is this an unimpressive number of minority electoral successes, but that the successes of Justice Adams and Justice Cook are tempered by their initial appointments to the bench, which gave them the "benefit [of] incumbency." Pls.' Post-Trial Br., at 35 (Doc. # 167). Here, Plaintiffs' evidence focuses on Justice Cook's 1994 election after his initial appointment.

Justice Cook gave credible testimony about the positive impact he believes his initial appointment to the bench had on his subsequent electoral success in 1994. As Justice Cook testified, Justice Adams strategically retired a year prior to the expiration of his second term to permit Justice Cook's appointment to the vacant seat. Justice Cook espoused his opinion, one that Justice Adams shared, that incumbency, even when secured through gubernatorial appointment, was an important factor for an African American's electoral success to statewide office. Trial Tr. IV, at 85-87, 93 (Cook); Pls.' Post-Trial Br., at 35 (Doc. # 167); Post-Trial Stipulations, at 34, ¶ 173, at 36, ¶ 184 (Doc. # 176).

But Justice Cook also testified that the benefits of incumbency inure to all candidates, regardless of race. Trial Tr. IV, at 86 (testifying that incumbency "gave lawyers an opportunity to see what kind of justice I would be if I was elected to full term" and noting that, when he ran, a judicial candidate's reputation among lawyers "went a long way with the electorate toward getting you elected"). Incumbency likely played a role that helped Justice Cook win the office in a tight election. Post-Trial Stipulations, at 36, ¶ 187 (Doc. # 176) (stipulation that, in 1994, Justice Cook defeated his white opponent in the general election by a narrow margin of 551,042 votes to 539,947 votes). But incumbency was not the only factor.

In addition to incumbency, Justice Cook credits his success over his white opponent to his "good reputation" as a district and circuit court judge, the 45,000 campaign miles he traveled to meet people in every county of the state, advertising, his diverse base of campaign contributors, and his ability to form biracial coalitions in the electorate. Trial Tr. IV, at 91-92, 119 (Cook); Trial Tr. I, at 135 (White); Post-Trial Stipulations, at 36 ¶¶ 185, 188 (Doc. # 76). Plaintiffs simply have not shown that it was "the sheer power of incumbency," Askew, 127 F.3d at 1384 n.18, that resulted in Justice Cook's electoral success in 1994, and Justice Cook certainly did not testify that it was. There was much merit behind his candidacy; he was highly qualified; and he campaigned aggressively.

Additionally, Plaintiffs have not proven that the 1994 race for Supreme Court justice between Justice Cook and his white opponent was a "polarized contest." Gingles, 478 U.S. at 57. Plaintiffs bear the burden of "controverting [Justice Cook's] success" with evidence of a "special circumstance." Solomon, 221 F.3d at 1231 (citations and internal quotation marks omitted). Plaintiffs' expert on racially polarized voting presented no statistical evidence as to this election: Dr. Handley arbitrarily chose not to reach back that far in her analysis. Importantly, Plaintiffs did not introduce evidence of how many white votes Justice Cook received in the 1994 general election as compared to the white votes received by his white opponent. And Plaintiffs' own expert testified that this race was not racially polarized. See Trial Tr. III, at 8-9 (McCrary).

Nor have Plaintiffs demonstrated that the initial appointments of Justice Adams and Justice Cook to the bench are "special circumstances" that negate their electoral successes. Gingles, 478 U.S. at 57. Notably, the fact that, after his initial appointment, Justice Adams ran unopposed in the 1982 general election suggests that he was highly qualified and that he had garnered wide support from white voters. See Milwaukee Branch of the NAACP, 116 F.3d at 1199 ("One good measure of white voters' willingness to support black candidates is the failure of white candidates to present themselves for election even when a majority of the electorate is white. Potential opponents concede the election only when they face certain defeat. That 6 black candidates ran without opposition therefore is highly informative. No decision of the Supreme Court suggests otherwise.").

Nonetheless, the stark statistic remains that, in Alabama's history, only two African Americans have been elected to statewide office for a total of three general elections and three primary elections. At the statewide level, this factor weighs in favor of Plaintiffs.

As noted, there is a difference in statewide elections and elections statewide. Here is where the Senate legislative language trumps Senate factor 7 "Legislative History" language: For purposes here, the relevant language and geography is "State": "The extent to which members of a protected class have been elected to office in the State . . . ." In a plain reading, there is no political subdivision of the State at issue here. In statewide elections, members of the protected class have fared poorly; but in elections statewide, they have fared much better, particularly in judicial races.

In the legislative branch today, Alabama has proportional representation by race in the state legislature. Post-Trial Stipulations, at 40, ¶ 210 (Doc. # 176); Defs.' Post-Trial Br., at 33 (Doc. # 168) (citing Rep. Hall's trial testimony). Turning to the judicial branch, the State criticizes Plaintiffs' expert, Dr. McCrary, for failing to consider the myriad post-1990 races where African Americans have been elected to circuit judgeships and other local offices. Defs.' Post-Trial Br., at 33. At present, African Americans have been elected to 46 of 150 state trial judgeships (circuit and district judges) as of the 2018 election cycle, for a 30.67% rate in a state with a 7% rate of African-American attorneys.⁶² Post-Trial Stipulations, at 41, ¶ 217 (Doc. # 176). Jefferson County, Alabama's largest county, is majority white but has a majority of African-American state judges. Nine of the state's sixty-eight probate judges (13.24%) are African American. The vast majority of legal decision making is local, and Alabamians have elected African-American lawyers to local trial benches in numbers over four times their proportion of all Alabama lawyers, and four percentage points higher than the African-American population of the state. Considering Alabama's notorious history regarding voting, this is a remarkable turnaround.

62 Justice Cook ran in five general elections, two statewide and three in Jefferson County (by far Alabama's most populous county) and won four of those five. In each instance, the electorate was majority white. Trial Tr. IV, at 79, 119 (Cook).

Moreover, twelve of Alabama's sixty-seven sheriffs (17.91%) are African American. The mayors of two of Alabama's largest cities, Birmingham and Montgomery, are African American, as are numerous other mayors, council members, county commissioners, and school board members. By any measure, and irrespective but not dismissive of "not there yet" sentiments, Alabama has made remarkable progress in the election of people of color to public office in the last fifty years.

There is no political subdivision at issue here. These non-statewide races are "in the State" and therefore probative under the seventh Senate factor. To the extent that the Eleventh Circuit authority speaks otherwise, see Stallings, 829 F.2d at 1560, it is focused on a challenge involving a political subdivision, not a State as a whole. Thus, these local, districted elections are entitled to some weight for purposes of the seventh senate factor.

Senate factor 7 weighs in favor of Plaintiffs on statewide elections but weighs in favor of the State on elections statewide.

g. Senate Factor 8: Responsiveness to Minority Needs

Senate factor 8 asks whether elected officials are sensitive to the "particularized needs of the members of the minority group." S. Rep. No. 97-417, at 29. Plaintiffs offered no evidence of any lack of responsiveness, taking the position that "Factor 8 is not applicable to the facts of this case." Pls.' Post-Trial Br., at 21 n.11 (Doc. # 167). This factor, thus, does not factor into the totality-of-circumstances analysis.

h. Senate Factor 9: Whether the State's Policy Is Tenuous

Senate factor 9 asks whether the State's policy behind its system "is tenuous." S. Rep. No. 97-417, at 29. As extensively discussed above and below, and not repeated here, Alabama's well-established policy of at-large, statewide, numbered-place elections for appellate judges is the very opposite of tenuous: It is weighty. See supra Part V.A.4., V.A.6.; infra Part V.A.7., V.B. This factor, while not dispositive, weighs heavily in favor of the State.

i. Conclusion: Totality of Circumstances

The court has weighed the totality of the circumstances in light of the Senate factors and has conducted "an intensely local appraisal of the design and impact" of the State's at-large, statewide method of electing appellate judges. Gingles, 478 U.S. at 79 (quoting Rogers v. Lodge, 458 U.S. 613, 622 (1982)). It has assigned appropriate weight to each factor. As explained, Senate factor 1 weighs in favor of Plaintiffs, and Senate factor 5 weighs moderately in favor of Plaintiffs. Senate factor 7 weighs in favor of Plaintiffs on statewide elections and in favor of the State on elections statewide. Senate factors 2, 3, 4, and 9 weigh moderately to strongly in favor of the State. Senate factors 6 and 8 have no weight.

Race has undoubtedly been a dominant factor in Alabama elections in the past. Today, though, the evidence of African-American voter registration and turnout indicates that Alabama's appellate judicial elections are equally open to participation by Plaintiffs and the protected class of voters. Notwithstanding the recent minimal representation of black-preferred candidates on Alabama's appellate courts, the evidence demonstrates that reasons other than race better explain the defeat of black-preferred candidates in Alabama's appellate judicial races. The decline of the Alabama Democratic Party over the past two decades has resulted in the defeat of all Democratic candidates since 2000 — both white and black — who have run for appellate judgeships in Alabama, with the single exception of Sue Bell Cobb. As a result, the Alabama Democratic Party has put up few candidates for appellate judicial races — only three candidates for thirty-three open seats since 2012. In a state where partisan elections have been an integral, long-running component of Alabama's history for electing appellate judges, it is not surprising that black-preferred candidates for appellate judgeships, when they are running at all, are losing because they are campaigning as Democrats.

The lack of Democratic candidates on the ballot, the proactive push in the late 1990s and early 2000s to change the politics of the Alabama Supreme Court for the purpose of implementing massive tort reform, the explosive 1986 Governor's race, the 1994 high-profile Hooper-Hornsby race, and Alabama's conservative ideology, considered cumulatively, are non-racial factors demonstrating that "what appears to be bloc voting on account of race [is instead] . . . the result of political or personal affiliation of different racial groups with different candidates." Solomon, 221 F.3d at 1225.

Based on the totality of circumstances and a "comprehensive, not limited, canvassing of relevant facts," Johnson, 512 U.S. at 1011, the court finds that Plaintiffs have not proven that the State's century-and-a-half-old (for the Supreme Court) and 109-year-old (for the courts of appeals) at-large, statewide voting structure for the election of its appellate judges denies Alabama's African-American citizens the right "to vote on account of race." § 10301(a). Plaintiffs have not shown that African Americans lack equal access to the political process in elections for Alabama's appellate judges. Thus, Plaintiffs have failed to prove a § 2 vote dilution claim.

This conclusion is even more compelling when the State's interests are considered. The discussion now turns to those interests as part of the remedy discussion.

7. Remedy: Balancing the State's Interests Against Plaintiffs' Interests in Remedial Districting

To reiterate, in the Eleventh Circuit, "the existence of a proper remedy" is essential for establishing liability under § 2. Davis, 139 F.3d at 1419 (emphasis added) (first citing SCLC, 56 F.3d at 1289, 1294-97; and then citing Nipper, 39 F.3d at 1530-31 (plurality opinion), 1547 (Edmondson, J., concurring)). "In assessing a plaintiff's proposed remedy, a court must look to the totality of the circumstances, weighing both the state's interest in maintaining its election system and the plaintiff's interest in the adoption of his suggested remedial plan." Id. at 1419-20 (citing Houston Lawyers', 501 U.S. at 426). The State has at least three interests at stake: (1) maintaining its long-standing system of electing appellate judges established by its Constitution; (2) linking an appellate judge's statewide jurisdiction and electoral base; and (3) guaranteeing the greatest possible pool of potential judicial candidates from which voters can choose. The balancing of these interests against Plaintiffs' proposed districting remedy weighs strongly in the State's favor.

a. Interference with Alabama's Constitutional Model

Statewide election for appellate judges is neither a recent development nor an accident in Alabama. Statewide election for appellate judges has been the State's policy since 1868, and that policy is undergirded by compelling justifications.

"Tradition, of course, does not make right of wrong, but we must be cautious when asked to embrace a new revelation that right has so long been wrong." Clements, 999 F.2d at 872. Abolishing Alabama's consistent, longstanding rule of statewide elections for appellate judges would substantially interfere with the State's constitutional model.

"[A] state has an interest in maintaining the judicial selection model established by its constitution." Davis, 139 F.3d at 1420 (citing Nipper, 39 F.3d at 1531 (plurality opinion), 1547 (Edmondson, J., concurring); see also Shelby Cty., 570 U.S. at 543 ("[T]he Framers of the Constitution intended the States to keep for themselves, as provided in the Tenth Amendment, the power to regulate elections." (quoting Gregory v. Ashcroft, 501 U.S. 452, 461-62 (1991)); Boyd, 143 U.S. at 161 ("[E]ach State has the power to prescribe the qualifications of its officers and the manner in which they shall be chosen."); SCLC, 56 F.3d at 1294 ("[A] state's interest in maintaining the challenged system is a legitimate factor to be considered." (citing Houston Lawyers', 501 U.S. at 426-27)). "Implicit in this first Gingles requirement is a limitation on the ability of a federal court to abolish a particular form of government and to use its imagination to fashion a new system." Nipper, 39 F.3d at 1531. In the judicial context, "single-member districts may run counter to the state's judicial model." Id.

The Alabama Constitution provides that "[a]ll judges shall be elected by vote of the electors within the territorial jurisdiction of their respective courts." Ala. Const. art. VI, § 152. So any elected judge — appellate, circuit, district, probate, or municipal — who has authority over a geographical area must be elected by the voters living there. Courts refer to this idea as "linkage" — connecting a judge's territorial jurisdiction with her electoral base. See Davis, 139 F.3d at 1421-22; see also Trial Tr. III, at 85 (Gaylord) (A state's "linkage interest relates to the jurisdiction of the court and the electoral base of that court so that those two are coextensive with one another."). Because the appellate courts have jurisdiction over the entire State, linkage is relevant to statewide elections for those judges.

The implementation of districted elections to remedy a § 2 violation would require both a rewrite of this constitutional provision by the Alabama legislature and a statewide vote on the new provision. And, although the boundaries of the circuit and district courts are prescribed by the legislature, see Ala. Const. art. VI, §§ 142, 151, Alabama's constitution has no mechanism for drawing districts for appellate courts. So not only would the legislature have to pass a bill proposing districted elections for the appellate courts, it would also have to pass a bill authorizing itself (or someone else) to draw the district lines. Both bills would require the approval of a majority of voters. See Ala. Const. art. XVIII, § 285. This is no small amount of federal court-ordered constitutional restructuring and intrusion into the State's election system for its appellate courts.

"[T]he authority of the people of the States to determine the qualifications of their most important government officials . . . lies at the heart of representative government." Gregory, 501 U.S. at 463 (citation and internal quotation marks omitted). "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." Clements, 999 F.2d at 872. The people of Alabama have "a substantial interest in defining the structure and qualifications of their judiciary," and they have chosen to define the office of appellate judge as one chosen on a statewide basis. Id.

In our federal system of government, states are entitled to structure the form of their state judicial elections as they see fit. "'[F]ederalism secures to citizens the liberties that derive from the diffusion of sovereign power.'" Bond v. United States, 564 U.S. 211, 221 (2011); s ee also Milwaukee Branch of the NAACP, 116 F.3d at 1201) ("In a federal system, states are entitled to do things differently; Illinois' willingness to use subdistricts no more obliges Wisconsin to do so than the other way around.").

Alabama is not unique in its selection. Like Alabama, thirty-three other states use at-large, statewide elections to select their appellate judges. Those states span the four (Census Bureau) regions of the United States: from the West — Alaska, Arizona, California, Colorado, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming; across the Midwest — Indiana, Iowa, Kansas, Michigan, Missouri, Minnesota, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin; in the South — Arkansas, Florida, Georgia, Maryland, Oklahoma, North Carolina, Tennessee, Texas, and West Virginia; and to the Northeast — Pennsylvania. But even within the consistency of at-large, statewide elections, these states implement varied components to their system. Seven states use partisan direct elections; fifteen use non-partisan direct elections; and sixteen use retention elections after initial appointment. See Post-Trial Stipulations, at 12-13, ¶ 67 (Doc. # 176).

Only four states — Illinois, Kentucky, Louisiana, and Mississippi — use districts, and each of those four states' districting selections is widely varied in its set-up. Stipulated evidence was received concerning the structure and history of appellate judicial elections in these four states. Post-Trial Stipulations, at 12-31, ¶¶ 67-151 (Doc. # 176). And trial testimony was received on Louisiana's and Mississippi's districted elections for appellate judges. See generally Trial Tr. IV, at 10-31, 64 (Thibodeaux) (discussing Louisiana's single-member districts for the election of appellate judges); Pls.' Trial Ex. 139B, at 5-6, 9-10, 14-15 (Banks); Trial Tr. IV, at 9 (Banks) (discussing Mississippi's multi-member districts for the election of appellate judges). There is a marked distinction between Alabama's constitutional judicial election model and the constitutional models of those four states. Districted elections for appellate judges have been entrenched in those states' constitutions for more than 150 years. See Post-Trial Stipulations, at 14, ¶ 78 (Illinois Constitution of 1848); at 17, ¶ 88 (Kentucky Constitution of 1850); at 20, ¶ 98 (Louisiana Constitution of 1852); at 28, ¶ 135 (Mississippi Constitution of 1832). Not so here. Imposing districted elections for the appellate courts would destroy Alabama's long-standing constitutional model of statewide elections for appellate judges and thus would substantially interfere with Alabama's judicial model. And there is no evidence in this record that districting is a proper comparator to at-large, statewide elections in appellate judicial races.

This factor — the State's interest in maintaining its judicial selection model — weighs heavily against Plaintiffs' districting remedial plan.

b. Linkage Between Appellate Judges' Jurisdictions and Electoral Base

The Eleventh Circuit has repeatedly recognized a state's weighty interest in linking a judge's territorial jurisdiction with her electoral base. See Davis, 139 F.3d at 1421-22; SCLC, 56 F.3d at 1296-97; Nipper, 39 F.3d at 1547 (Edmondson, J., concurring). In other words, states may legitimately adopt a policy that all voters in a judge's territory have a say in who the judge is. When "there is no evidence that a state is administering its judicial election system in a racially discriminatory manner, the state's interest in preserving linkage between judges' jurisdictions and electoral bases is even weightier." Davis, 139 F.3d at 1421. It is appropriate, first, to discuss the evidence concerning the administration of justice in Alabama's appellate courts. After that, the policy goals of linkage and how they relate to Alabama's election system for appellate judges will be addressed.

i. Impartial Administration of Justice in Alabama's Appellate Courts

There is no objective, firsthand evidence that justice in Alabama's appellate courts is being administered in a racially discriminatory manner. While Alabama NAACP President Bernard Simelton lamented that some of the group's members "don't feel comfortable . . . appealing their case . . . because they seem to know the results," Trial Tr. I, at 163-64, neither he nor any other witness gave an example of such a case. Plaintiffs have identified no recent appellate decision⁶³ that was harmful to African Americans or reflects insensitivity to their interests. See Lopez, 339 F. Supp. 3d at 617 (observing that plaintiffs "have not identified any examples of decisions that are non-responsive to minority needs or wrongly decided in the overall interest of equal justice").

63 Plaintiffs did introduce a list of appellate rulings that supported white supremacy in the State. See Pls.' Trial Ex. 131. But none of those decisions is any more recent than 1963, so they are entitled to no weight as evidence of twenty-first century realities.

Relatedly, there is no evidence that justice in the State's appellate courts is not being administered impartially. The perception that justice is being administered impartially is of critical import to both the State and Plaintiffs, and Alabama's presently all-white appellate judiciary could foster a negative perception in the African-American community of its ability to be fair to minorities. But the court has considered this fact with respect to the objective benchmark of Senate factor 7 — the extent of minority success in elections. To the extent Plaintiffs argue that they do not trust their appeals to appellate courts on the basis of the judges' skin color, that argument is unsubstantiated and rejected.

Moreover, one of the Plaintiffs acknowledged that the court experiences with which he and his friends are familiar are at the trial-court level, not at the appellate level. See, e.g., Trial Tr. II, at 61-62 (Harris). This testimony makes sense. The judges whom litigants, jurors, witnesses, and spectators are most apt to see — face to face in the flesh — are those elected inside prescribed jurisdictions: circuits for circuit court and counties for district court and probate court. These judges are the ones who live among the people in every county or circuit of counties, who have practiced law privately or publicly, who coach ball teams, and who serve on school, church, chambers of commerce, charity, and bank boards. These are the judges actually known personally by the people, and who are directly accountable to the people they know and serve. These judges are the ones whose courthouse doors are open wide and whose courtrooms are overflowing. Figures from the Alabama Unified Judicial System's annual report for fiscal year 2018 indicate that there were 205,691 cases filed in Alabama circuit courts, and 604,286 in Alabama district courts (as compared to 1,216 case filings in the Alabama Supreme Court, 573 in the Court of Civil Appeals, and 1,248 in the Court of Criminal Appeals). See Ala. Admin. Office of Courts, Alabama Unified Judicial System: Fiscal Year 2018 Annual Report and Statistics 143, 154 (2019), http://www.alacourt.gov/Publications.aspx.; Supreme Court of Alabama, Annual Statistics for the Fiscal Year Ending September 30, 2019, http://judicial.alabama.gov/Appellate/SupremeCourt. Hence, justice at the "retail level" is administered by state circuit and district judges, and to some extent, probate judges. Any dissatisfaction in the justice system for most voters more likely would be felt at the trial court level, where there are over 800,000 filings per year.

There being no evidence that justice in Alabama's appellate courts is being administered impartially or in a racially discriminatory manner, "[t]he purpose to be served by [districting] . . . would be, for the most part, to assist in the election of black [appellate] judges." SCLC, 56 F.3d at 1295. "[T]he effect of having black judges accountable primarily to the black section of the [state] and white judges answerable to the white section of the [state] . . . would threaten the administration of justice — race would become the linchpin of the judicial system." Id. Accordingly, districting would "undermine the existing public confidence in judges as fair and impartial decisionmakers." Id.

ii. Judicial Accountability and Independence

Linkage promotes two main policy goals: judicial accountability and independence. The first policy promoted by linkage is judicial independence. Linkage ensures that judges are accountable to all those who will be most affected by their decisions. "The Alabama model reflects a belief that voters should have the right to hold a judge accountable for his or her performance." SCLC, 56 F.3d at 1296-97. "Were a judge to be answerable to an electorate smaller than his [or her] jurisdiction, the judge would have an incentive, however unethical, to engage in 'home cooking,' favoring litigants from his [or her] election district over others." Davis, 139 F.3d at 1421 (alterations added). Moreover, the Eleventh Circuit has recognized a state's "interest in avoiding even the appearance that its judges may harbor 'home cooking' biases." Id.

The second policy promoted by linkage is judicial independence, from the legislative and executive branches and from various factions within the electorate. Alabama used legislative appointment for the Supreme Court in its early years, but rejected that model in favor of statewide elections in 1868, and has maintained them ever since. Gaylord Expert Report, at 27-30 (Defs.' Trial Ex. 25; Doc. # 87-1). Alabama was not alone. Alabama's decision to adopt elections for its appellate courts was part of a broad-based, nationwide trend predicated on concerns over legislative and executive overreach. Id. at 4.

Statewide elections support judicial independence in several ways. First, statewide elections make judges accountable not to those whose laws they may strike down, but to the people who constitute the State. As an empirical matter, there was evidence that elected judges exercise their power of judicial review to strike down unconstitutional laws much more often than appointed judges. Trial Tr. V, at 45-46 (Bonneau). The Alabama Supreme Court is no stranger to judicial review. During the "tort wars" of the 1990s, the court struck down legislative tort reforms, which led to business interests becoming more involved in Alabama judicial elections. See, e.g., Henderson, 627 So. 2d at 878 (striking down a legislative cap on punitive damages as violating the right to a jury trial); Gaylord Reply Report, at 6-8 (Defs.' Trial Ex. 27; Doc. # 93-2).

Second, statewide elections further the important interest in maintaining separation of powers by removing legislative control over the judiciary. See Ala. Const. art. III, § 42 ("To the end that the government of the State of Alabama may be a government of laws and not of individuals, . . . the legislative branch may not exercise the executive or judicial power . . . ."); Lopez, 339 F. Supp. 3d at 618 (stating that single-member districts would "inject legislative control into the drawing of judicial districts, contrary to the State's sovereign interest in a governing model that includes separation of powers"). Statewide elections remove the legislature's temptation to manipulate districts for political purposes.

Third, statewide elections promote the idea that judges should not be subservient to the parochial interests of a distinct geographical area. Accountability and independence are implicated here. In The Federalist No. 10, James Madison endorsed a similar idea. Madison thought that the influence of "factions," or special interest groups, could be kept in check by increasing the size of the polity and by taking in more competing interests, thereby preventing one faction from becoming too powerful and "capturing" public officials. See generally The Federalist No. 10 (James Madison); Trial Tr. III, at 86-87 (Gaylord). "Extend the sphere," said Madison,

and you take in a greater variety of parties and interests; you make it less probable that a majority of the whole will have a common motive to invade the rights of other citizens; or if such a common motive exists, it will be more difficult for all who feel it to discover their own strength, and to act in unison with each other.

The Federalist No. 10 (James Madison). For Madison, structure is paramount in counteracting the mischiefs of democratic self-government. See id.; see also The Federalist No. 51 (James Madison) (explaining the importance of the separation of powers in mitigating the pitfalls of human nature). Alabama has decided to structure its judiciary so that those who have final say on matters of state law administer justice in view of the interests of fishermen and harbor workers in the Gulf region, farmers and laborers across the Wiregrass, Black Belt, and Tennessee Valley, public servants in Montgomery, and engineers in Huntsville. Federal courts may not lightly overturn that decision.

Plaintiffs' response to these two policy goals is twofold. First, citing their judicial witnesses, they say there is no evidence that a judge elected from a district would promote the interests of his or her constituents over others. Madison would disagree, and his argument is human nature: "It is vain to say that enlightened statesmen will be able to adjust these clashing interests, and render them all subservient to the public good. Enlightened statesmen will not always be at the helm." The Federalist No. 10 (James Madison). "Judicial integrity is . . . a state interest of the highest order," Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 889 (2009) (quoting Republican Party of Minn. v. White, 536 U.S. 765, 793 (2002) (Kennedy, J., concurring)), and Alabama "has an interest in avoiding even the appearance that its judges may harbor 'home cooking' biases," Davis, 139 F.3d at 1421 (emphasis added). See also Caperton, 556 U.S. at 888 (noting that states may legitimately take measures "to eliminate even the appearance of partiality"); Chisom, 501 U.S. at 400-01 ("The fundamental tension between the ideal character of the judicial office and the real world of electoral politics cannot be resolved by crediting judges with total indifference to the popular will while simultaneously requiring them to run for elected office."); Milwaukee Branch of the NAACP, 116 F.3d at 1201 ("Perhaps the belief that judges favor those who elect them is unwarranted — though the diversity jurisdiction of the federal courts rests in part on a belief that state judges highly value the interests of that state's citizens and thus are potentially biased against citizens of other states."). Put simply, the judicial oath is not a foolproof guard against human nature. The State may legitimately structure its judiciary to remove any appearance that its appellate judges serve anyone but the entire state.

Second, Plaintiffs argue that, under the current statewide system, the problem of "capture" is already felt because appellate judges need not serve the interests of the state's African-American population to get elected. But, as the Eleventh Circuit has repeatedly recognized, their remedy would make matters even worse. Were the court to impose districts for appellate judgeships, it would further dilute the votes of racial minorities within each district. "In white subdistricts the voting power of blacks would be diluted; in black subdistricts the voting power of whites would be diluted." SCLC, 56 F.3d at 1297; see also Davis, 139 F.3d at 1422 (quoting with approval the same passage from SCLC). If appellate judges do not need African-American votes now, they would certainly not need them if half the African-American population were wedged into one or two districts.

Dr. McKee's report on the matter for which he is an expert confirms this phenomenon. He notes that Gingles itself "tore[] asunder" the need for candidates to form biracial coalitions to win. When the State was ordered to create a black-opportunity district for the United States House of Representatives for the 1992 election, drawing the new District 7 forced neighboring District 6 to reduce its African-American, voting-age population from 31.5% to 8.3% of the district. The veteran Democratic incumbent then lost to a Republican that year. McKee Rebuttal Report, at 13-14 (Pls.' Trial Ex. 123; Doc. # 88). Of course, the diminution of minority influence outside black-opportunity districts is a natural consequence of the Gingles remedy, which has indeed increased minority representation. McKee Rebuttal Report, at 14. But the Eleventh Circuit seems to think this is a special concern for judges (as opposed to legislators), and this court is bound to agree. The court therefore concludes that Plaintiffs' interest in remedying vote dilution is not well-served by diminishing minority influence in elections outside majority-black districts. See Clements, 999 F.2d at 872-73.

Not only would racial minorities within each district lose influence in appellate judicial elections, districts would also "disenfranchise every voter residing beyond a judge's [district], thus rendering the judge accountable only to the voters in his or her [district]." Nipper, 39 F.3d at 1543 (plurality opinion) (alterations added); see SCLC, 56 F.3d at 1297; Davis, 139 F.3d at 1422 (agreeing with the trial court's ruling that plaintiffs' proposed subdistricting remedy "would be akin to compelling the state to disenfranchise every voter residing in the two jurisdictions, but outside the subdistrict" (quotation omitted)). For example, in the five-district hypothetical plans for the intermediate courts of appeals, approximately 42.9% of all African-American Alabamians live in District 1, the majority-black district. This means, assuming racially polarized voting, more than half of African Americans who live elsewhere in the State would have little opportunity to elect their preferred candidate. See Bonneau Expert Report, at 17, ¶ 43 (Defs.' Trial Ex. 24; Doc. # 85-1). The majority-black district in the eight-district supreme court plan contains 25.1% of the African-American population in Alabama, which disfranchises 75% of the African-American electorate living outside this district. Under the hypothetical plans, over half of the black population would be disfranchised by being spread among four (in the AC plans) or seven to eight (in the SC plans) majority-white districts. See Bonneau Expert Report, at 17, ¶ 44.

Furthermore, in the current system, all Alabama voters vote on all nineteen appellate positions. Plaintiffs' proposed remedy would force all Alabama voters to give up their right to vote for sixteen out of nineteen appellate judges (or fifteen out of nineteen if the chief justice were elected statewide). Yet all nineteen — not just the three they voted on — would exercise jurisdiction over all citizens statewide. That fact also weighs against Plaintiffs' proposed remedy. While now Alabama voters vote on 100% of the appellate judges, districted elections would diminish that voting power to a mere 16% (or 21%, if the chief justice election is statewide). And the appellate courts' current practices of deciding a portion of supreme court cases in panels of five and criminal appeals in panels of three would further deprive Alabama voters of their interest in holding appellate judges accountable at the ballot box.

Plaintiffs correctly note that Nipper, SCLC, and Davis were all about trial judges, not appellate judges. And unlike trial judges, who act alone, appellate judges exercise authority as a group. Thus, Plaintiffs argue, voters' interest in picking every judge is less weighty. See SCLC, 56 F.3d at 1297 (explaining in dicta that, with subdistricting, the loss of minority influence on judicial decision making "would be more pronounced in [the] context of a lone decisionmaker, a trial judge, who would lack input from the colleagues elected by the rest of the citizenry of the jurisdiction").

This argument loses much of its force in view of Alabama appellate courts' current procedures, evidence that was not available at the motion-to-dismiss stage. The Alabama Supreme Court decides about a third of its cases in divisions of five: four associate justices plus the chief justice. Trial Tr. VI, at 49 (Murdock); Trial Tr. IV, at 108-09 (Cook); see Ala. R. App. P. 16. Thus, if justices were elected by districts, half of Alabamians would have no "representative" associate justice in a third of the cases, not including cases in which justices recuse.⁶⁴ The same would be true for the Court of Criminal Appeals, which may also issue decisions in panels of three. See Ala. Code § 12-3-2(c); Trial Tr. VI, at 50-51 (Murdock). In short, in all decisions rendered by less than the full court, half of the electorate would have no ability to hold those judges accountable, nor would those voters have a "representative[] of their choice." § 10301(b).

64 It is true that, in one of Plaintiffs' remedial plans, the chief justice is elected statewide, so voters would theoretically have a say in all decisions in which the chief justice participates. But under Plaintiffs' theory of the case, that is a poor substitute for a judge elected from the voter's home district having a say in the decision — especially when that voter has had her vote diminished from nineteen votes to three, or in the case of the Supreme Court, from nine votes to one or two.

African Americans would suffer this loss of accountability even more strongly. When the Supreme Court decides cases in divisions, black-preferred justices may, simply by the seniority-based allocation of seats, be excluded from the decision-making process altogether in a significant number of cases. See Trial Tr. VI, at 50 (Murdock) (explaining how the senior-most justice heads the first division, the next-senior-most justice heads the second, "[a]nd then you go back and forth across the table by seniority until you fill in all the seats"). And on the Court of Criminal Appeals, African Americans will lose their elected judge in all the cases decided by a panel for which his or her name is not drawn. Plaintiffs have identified no authority under which a federal court may order a state court to alter its internal decision-making procedures to remedy this issue. In any event, such an order would present enormous federalism concerns.

Regardless, if the concern that districted elections would disfranchise voters is less pronounced for appellate judges than for trial judges, it is only slightly less so. That is because what power appellate judges lose in their inability to issue solo judgments, they gain in the weight and finality of their judgments. Unlike the single-judge trial court, appellate judges sit together (in panels or en banc) because they find and apply legal rules, binding themselves, lower courts, and federal courts applying state law. Setting precedent is too important a task to leave in the hands of one judge. The Alabama Supreme Court is the final interpreter of the constitution, statutes, regulations, rules of court, and common-law doctrines that govern the entire state. A trial court's legal errors are correctable by appeal, but overturning a wrong appellate decision is a much more difficult task. As Justice Robert Jackson pithily quipped, "We are not final because we are infallible, but we are infallible only because we are final." Brown v. Allen, 344 U.S. 443, 540 (1953) (Jackson, J., concurring in the judgment). Alabama has a weighty interest in preserving all of its citizens' right to have a say in who the state's most powerful judges are.

Moreover, it is far from clear that African Americans' ability to influence appellate decision making would be enhanced by district-based judgeships. As the Lopez court explained: "The collegial nature of a multimember body does not prevent it from, in practice, running roughshod over any given minority of judges. Single-member districts could have the unintended effect of increasing the power of a majority of judges elected from districts with wider polarization levels in favor of white voters." 339 F. Supp. 3d at 619.

Linkage promotes judicial accountability and judicial independence from legislative and faction control, and these policy goals are equally applicable to trial and appellate judges. Plaintiffs' proposed remedy of districting would sever the State's legitimate interest in linking an appellate judge's statewide jurisdiction and electoral base.

iii. A Matter of Law , Not Fact

Plaintiffs' final response to the State's linkage interest is that there was no evidence supporting it. This argument echoes that of the two dissenters in Nipper. See Nipper, 39 F.3d at 1557 (Hatchett, J., joined by Kravitch, J., dissenting) ("The opinion holds, with no evidentiary corroboration, that Florida's interest in maintaining its current at-large electoral systems for judges precludes the acceptability of any of the appellants' proposed remedies . . . . Instead of evidence, the majority apparently relies upon disembodied conjecture to support its ultimate determination."). But Judges Hatchett and Kravitch were, in fact, dissenters. Judges Tjoflat and Anderson did not seem to think that evidence of linkage was important or required, and the four concurring judges explicitly held that the state's "legitimate interest in maintaining linkage between jurisdiction and the electoral bases of its trial judges is, as a matter of law, great and outweighs . . . whatever minority vote dilution that may possibly have been shown here." See Nipper, 39 F.3d at 1547 (Edmondson, J., concurring) (emphasis added).

That linkage is a matter of law, not fact, is confirmed in SCLC and Davis. See Davis, 139 F.3d at 1421 (noting that a state "has an interest in avoiding even the appearance that its judges may harbor 'home cooking' biases" (emphasis added)); SCLC, 56 F.3d at 1296-97 (holding, without citing evidence, that linkage "serves to preserve judicial accountability" and to prevent the specter of "home cooking"); see also Clements, 999 F.2d at 871 ("The plaintiffs urge that the weight or substantiality of Texas's linkage interest is an issue of fact for the district court to decide in the first instance, reviewable only for clear error. We disagree."); Lopez, 339 F. Supp. at 617 (finding that linkage "is a question of law" and dismissing the plaintiffs' criticism of "the State for not submitting evidence of its linkage interest" (citing Clements, 999 F.2d at 871)). The legitimacy and weight of Alabama's interest in linking appellate judges' jurisdiction with their electoral base are issues of law, and the State need not provide evidence that these policy justifications are in fact promoted by linkage. Linkage is a philosophical and policy judgment about what makes a judge and how a judiciary should be structured.

Whether the State in fact had these interests, and not discriminatory motives, in mind when it adopted or maintained its present system is a different matter. That question is one of fact, and thus evidence is needed. See Clements, 999 F.2d at 871. As explained in the discussion of the constitutional claims, infra, there is no evidence to conclude that Alabama's policy of linking judges' territorial jurisdiction with their electoral base disguises an invidious motive.

c. The Pool of Candidates

Statewide elections also ensure that voters have a large pool of qualified candidates to choose from, a state interest that is undermined by districting. "The State's interest in having a pool of black lawyers large enough to give the voters a reasonable choice from which to select a qualified candidate must be considered when evaluating suggested remedies." SCLC, 56 F.3d at 1296. To serve as a judge on one of Alabama's appellate courts, one must have been a member of the state bar (or another state's bar) for at least ten years. Ala. Code §§ 12-2-1, 12-3-1.

Plaintiffs' remedy severely limits the pool of potential candidates for seats on the appellate courts. For example, District 7 in both SC1 and SC2 has about 600 lawyers (of all races) living there. But District 5 in the same two plans has more than 4,000 lawyers (of all races). Post-Trial Stipulations, at 41-42, ¶ 220 (Doc. # 176). Considering the number of African-American lawyers in each district (if it is assumed that African-American voters prefer African-American judges), the pool is even more limited. It also is presumptuous to assume that all or most Alabama lawyers — no matter their race — desire to campaign, run for, and be an appellate judge. In all of Plaintiffs' Supreme Court plans, District 1, which encompasses the Black Belt, has no more than 264 African-American attorneys (189 of whom are in Montgomery County).⁶⁵ Defs.' Trial Ex. 102; Defs.' Proposed Findings of Fact, at 135-36, ¶ 262 (Doc. # 176). There are nearly twice that many African-American attorneys in Jefferson County alone. In fact, 767 of the 1,052 African-American lawyers in Alabama (72.9%) practice in three counties: Jefferson, Montgomery, and Mobile. See Number of Members in Good Standing - by Race and Gender, Nov. 16, 2018 (Defs.' Trial Ex. 101). So not only will Plaintiffs' remedy dilute African-American voting power in the districts in which African Americans are not a majority, but it will force all but about 300 potential African-American candidates into just one or two of nine districts in the Supreme Court plans, or into one of five districts for the intermediate appellate courts. That makes it even less likely that an African American will be elected from elsewhere in the State. Under the current structure, an African American from any of Alabama's sixty-seven counties could potentially run, not just for one to three seats, but for any of the nineteen seats.

65 It should be noted that 264 is the number of total attorneys, not the number of attorneys eligible for an appellate court seat, which would be much less than 264.

Plaintiffs suggest that there need not be a requirement that a lawyer reside in the district he or she seeks to represent. That may be so, but to quote Justice Murdock, "[G]ood luck getting elected in a district that you don't live in." Trial Tr. VI, at 103; see also Trial Tr. IV, at 62-63 (Thibodeaux) (explaining that it is "political suicide" to run outside the district where one lives); Trial Tr. V, at 40 (Bonneau) (explaining that a nonresident candidate would open herself to attacks that she is an outsider). Eliminating a residency requirement therefore does little to enhance the feasibility of Plaintiffs' remedy.

In sum, districted elections would limit the pool of judicial candidates from whom the voters in districts could choose.

d. Weighing of the Interests

Plaintiffs' proposed remedy of districted elections is not a feasible alternative to the at-large, statewide system that has been in place in Alabama for more than a century and a half.

First, in its 1868 Constitution, Alabama chose at-large, statewide elections for its Supreme Court judges, and when it created an intermediate Court of Appeals in 1911, Alabama also selected at-large, statewide elections. The at-large elections link appellate judges' jurisdiction to the whole of the judges' electoral base. Plaintiffs' proposed remedy of districted elections runs counter to the State's more than century-and-a-half-old method that the people of Alabama selected for electing appellate judges.

Second, Plaintiffs' proposed districting remedy would sever the State's legitimate linkage interest; it would abolish the correlation between the statewide territorial jurisdiction of appellate judges and their electorate. The policy goals linkage promotes — democratic accountability, judicial independence from legislative and faction control, and the appearance of justice — are equally applicable to trial and appellate judges. And Plaintiffs have not shown that the mere circumstance that appellate judges make decisions as a group diminishes the State's interest in giving all voters a right to choose those judges. Districted elections would undermine these policy goals by compromising the public's confidence in the appellate judiciary's dispensation of justice, by disfranchising voters who live outside a judge's district, and by depriving Alabama voters of having a say in who is elected to all nineteen appellate seats. A federal court should be loath to interfere with that interest when, as here, there is no evidence that the system is administered in a racially discriminatory manner.

Third, districted elections would hamper the State's interest in ensuring a wide pool of qualified judicial candidates from whom Alabama voters could choose.

Fourth, the expanse of § 2 jurisprudence has evolved in the context of legislative elections, not judicial elections or, more specifically, not at-large, statewide judicial elections. This is significant because, while in 1991 the Supreme Court held that § 2, as amended in 1982, applied to the use of multimember districts in state appellate judicial elections, it did not then and has not in twenty-eight years since delineated a standard for § 2's application to judicial elections, recognizing only that "serious problems [might] lie ahead" in § 2's application to state judicial elections. Chisom, 501 U.S. at 403. In the concurring opinion in SCLC, Judge Edmondson succinctly summed up this circuit's law on § 2's application to at-large elections of trial judges in Alabama. His pronouncements on the evidence and on the State's interests apply equally in this case:

Alabama as a matter of law has the right — a right is an extremely weighty interest — to structure its judicial branch pretty much as Alabama thinks best. That structure now (and for a considerable time) involves at-large elections of trial [and appellate] judges from the jurisdictions they serve. This approach is neither irrational nor, in this country, rare. And, no evidence in this case shows that either the structure of the judicial system as a whole or specific judicial boundaries were drawn up to discriminate against any Alabama citizen on account of race. None of the judicial districts are unusual or strange. These few facts have great legal force. With a single stroke, they strike, in favor of Alabama, the balance of the circumstances.

SCLC, 56 F.3d at 1298 (Edmondson, J., concurring) (brackets added) (internal citations omitted). SCLC was decided in 1995, but it still is the Eleventh Circuit's controlling word on § 2 vote dilution claims challenging a state's judiciary model. Three years later, in Davis, the Eleventh Circuit confirmed the inability of a § 2 plaintiff in this circuit to offer a workable remedy when challenging a state's at-large judicial election system: "Together with Nipper, SCLC, and the additional case of White v. Alabama, we will with this decision have disallowed redistricting, subdistricting, modified subdistricting, cumulative voting, limited voting, special nomination, and any conceivable variant thereof as remedies for racially polarized voting in at-large judicial elections." Davis, 139 F.3d at 1423. There has been no Eleventh Circuit decision on the topic since.

In view of the evidence and of the state's political interests, districting for Alabama's appellate judges only makes sense if the people chose it. In most states, they have, like Alabama, made other sensible choices. There is no benchmark for the court to choose among all the various election systems for state appellate judges. See Holder, 512 U.S. at 880 (There must be a "reasonable alternative practice as a benchmark against which to measure the existing voting practice."). Additionally, Alabama's "interest in maintaining its judicial model and in preserving such linkage outweigh[] the plaintiffs' interest in ameliorating the effects of racial polarization in at-large judicial elections." See Davis, 139 F.3d at 1423. Accordingly, Plaintiffs' § 2 claim fails for the additional reason of a lack of a feasible remedy. B. Constitutional Claims

Plaintiffs challenge as unconstitutional two aspects of Alabama's model for electing appellate judges: (1) statewide elections and (2) the numbered-place requirement. For the reasons explained below, Plaintiffs have not carried their burden to show that racial discrimination was a motivating factor in the adoption or maintenance of any of the aspects of Alabama's judicial elections.

The Fourteenth and Fifteenth Amendments prohibit racial discrimination in voting. See Shelby Cty., 570 U.S. at 560. A successful equal protection claim requires "[p]roof of racially discriminatory intent or purpose," Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 265 (1977), along with discriminatory results, see Johnson v. DeSoto Cty. Bd. of Comm'rs, 72 F.3d 1556, 1564 (11th Cir. 1996). Claims that a law was adopted or is being maintained with the purpose of minority vote dilution are "subject to the standard of proof generally applicable to Equal Protection Clause cases." Rogers v. Lodge, 458 U.S. 613, 617 (1982). Plaintiffs must "prove by a preponderance of the evidence that racial discrimination was a substantial or motivating factor" in the adoption or maintenance of the challenged aspects of Alabama's appellate judicial elections. Hunter, 471 U.S. at 225 (quoting Underwood v. Hunter, 730 F.2d 614, 617 (11th Cir. 1984)).

A discriminatory purpose "may often be inferred from the totality of the relevant facts, including the fact . . . that the law bears more heavily on one race than another." Washington v. Davis, 426 U.S. 229, 242 (1976). Courts should also consider (1) "the historical background of the decision," (2) "the specific sequence of events leading up to the challenged decision," (3) departures from normal procedures or substantive policies, and (4) "[t]he legislative or administrative history . . . especially where there are contemporary statements by members of the decisionmaking body." Vill. of Arlington Heights, 429 U.S. at 266-68. Evidence of historical discrimination in voting is also relevant, particularly "where the evidence shows that discriminatory practices were commonly utilized, that they were abandoned when enjoined by courts or made illegal by civil rights legislation, and that they were replaced by laws and practices which, though neutral on their face, serve to maintain the status quo." Rogers, 458 U.S. at 625. Upon a showing of discriminatory motive, "the burden shifts to the law's defenders to demonstrate that the law would have been enacted without this factor." Hunter, 471 U.S. at 228.

Plaintiffs have not carried their burden to show that any aspect of Alabama's elections for appellate judges was adopted for a discriminatory purpose. Statewide elections for Alabama's appellate judges were adopted in the 1868 Reconstruction Constitution, which was "written and ratified primarily by Republicans committed to equality for African Americans." Norrell Expert Report, at 6 (Pls.' Trial Ex. 85; Doc. # 66-1). Dr. Norrell testified that, in 1868, at-large, statewide elections for Alabama Supreme Court justices were "not instituted for the purpose of discriminating against blacks. You might say [they] w[ere] instituted in order to help African Americans." Trial Tr. III, at 280; see also Gaylord Expert Report, at 29 (Defs.' Trial Ex. 25; Doc. # 87-1) (opining that "the shift to judicial elections in Alabama's 1868 Constitution was meant to promote the values underlying the widespread movement of States to popularly elect judges and to preclude the use of appointments to perpetuate racial discrimination in the judiciary"); Trial Tr. III, at 114, 145-46 (Gaylord). Thus, statewide elections were initially adopted to enhance African Americans' chances of having their preferred candidates achieve judicial office.

And, as the trial court found in Southern Christian Leadership Conference of Alabama v. Evans, 785 F. Supp. 1469, 1477 (M.D. Ala. 1992), af f 'd sub nom. SCLC v. Sessions, 56 F.3d 1281 (11th Cir. 1995) (en banc), a case in which Plaintiffs' historian Dr. Norrell also testified, the evidence shows that numbered-place requirements were not motivated by a discriminatory purpose. Beginning in 1874, when white "Redeemer" Democrats⁶⁶ took back control from Reconstruction Republicans, the justices of the Alabama Supreme Court were effectively chosen by the all-white Democratic primary, a process which was later officially sanctioned, then struck down by the United States Supreme Court in 1944, see Smith v. Allwright, 321 U.S. 649 (1944). Norrell Expert Report, at 7, 22 (Pls.' Trial Ex. 85; Doc. # 66-1). In 1901, the infamous Alabama Constitution was adopted, effectively removing all but 3,000 of the 100,000 African Americans who were on the voter rolls. Norrell Expert Report, at 13 (Pls.' Trial Ex. 85; Doc. # 66-1). The numbered-place system was adopted in 1927, at a time when the appellate courts were selected in the all-white Democratic primary and when all but a few African Americans were disfranchised. See SCLC, 56 F.3d at 1286. Moreover, in 1927, there were no African-American lawyers in Alabama, so no African Americans would have been eligible for a seat on the appellate courts. See id.; see also Trial Tr. III, at 285 (Norrell). Dr. Norrell stated that he had no reason to disagree with SCLC's finding that there were no African-American lawyers in Alabama in 1927. Trial Tr. III, at 285 (Norrell).

66 They were so named because their goal was to "redeem" Alabama from the "black rule" of Reconstruction. Norrell Expert Report, at 3 (Pls.' Trial Ex. 85; Doc. # 66-1).

As Dr. Norrell testified and the court found in SCLC, the goal of numbered places was incumbency protection. In 1926, the Ku Klux Klan-backed faction of the Democratic Party fielded successful candidates, threatening Alabama with the possibility of Redeemer rule. Norrell Expert Report, at 16-18 (Pls.' Trial Ex. 85; Doc. # 66-1); see SCLC, 56 F.3d at 1286. The Klan-backed candidates who won in 1926 were Bibb Graves, Hugo Black (yes, that Hugo Black), and Charlie McCall, who were elected governor, U.S. senator, and attorney general, respectively. Joel Bascom Brown, who also was supported by the Klan, was elected to the Alabama Supreme Court in the same year.⁶⁷ Norrell Expert Report, at 17. In 1927, the legislature adopted a numbered-place requirement for the Alabama Supreme Court and Alabama Court of Appeals to give an advantage to incumbent Redeemer Democrats against other Democratic factions, none of whom was concerned with the interests of African Americans. Trial Tr. III, at 285-86 (Norrell).

67 Klan political influence peaked in the 1920s, and not just in the South. Colorado had a Klan-backed Governor and a Klan legislative majority. Trial Tr. III, at 202 (Norrell).

Plaintiffs have not pointed to any evidence that commands a conclusion other than that of the court in SCLC. With only one exception,⁶⁸ Plaintiffs have not argued that Dr. Norrell's data and conclusions in SCLC about Alabama's numbered-place system in judicial elections were different from his data and conclusions in this case. See Johnson, 72 F.3d at 1561 n.5 (noting that, if the evidence in a prior case is "identical to the evidence" in the pending case, then the appellate court's rulings on the evidence in the prior case "would be binding precedent" in the pending case).

68 Plaintiffs contend that Dr. Norrell has relied upon "evidence that was missing in" SCLC, namely, a December 31, 1961 article in which columnist Bob Ingram wrote that Act 221 "was aimed at Negroes as much as anything else." Pls.' Proposed Findings of Fact, at 67, ¶ 48 (Doc. # 176). But Act 221 was passed in 1961 and, thus, "did not change the judicial electoral process." SCLC, 56 F.3d at 1286; see also SCLC, 785 F. Supp. at 1488 ("Whatever one may find as to the reason for Act 221 requiring at large, numbered places for elections generally, the Act simply did not affect judicial elections."), aff'd sub nom. SCLC v. Sessions, 56 F.3d 1281 (11th Cir. 1995) (en banc). --------

Plaintiffs have also not carried their burden to show that any aspect of Alabama's appellate judicial elections was maintained for a discriminatory purpose. Judicial reform efforts in the late 1960s and 1970s were met with hostility by incumbent judges who resisted "changes that might undermine the advantages of their incumbency." Norrell Expert Report, at 28-29 (Pls.' Trial Ex. 85; Doc. # 66-1). Reforms were eventually made, however. In 1973, Constitutional Amendment 328, which is in effect today, shifted control of the judiciary from the legislature to the Alabama Supreme Court. Amendment 328 kept judicial elections in place, which was politically necessary for the amendment to pass. Norrell Expert Report, at 29-30; see also SCLC, 56 F.3d at 1286 ("The Judicial Article and reform movement of the 1970s also did not attempt changes from a popularly elected judiciary. Moving toward an appointment system would have garnered opposition from many Alabamians and would have lead to the defeat of the Judicial Article because enactment required voter approval."). Dr. Norrell opined that the objection to an appointment system, advocated by judicial reformers Howell Heflin and C.C. "Bo" Torbert, was that it would resemble the federal judiciary, which had been so demonized by Governor George Wallace. Norrell Expert Report, at 29-30. So Heflin's and Torbert's reform efforts kept judicial elections intact in exchange for Wallace's support. Trial Tr. III, at 270-71 (Norrell).

But Wallace was not the only reason Alabama chose to keep judicial elections. As noted by one of Dr. Norrell's sources (but not mentioned by him), "[Chief Justice] Heflin knew that even among Alabama's reform-minded judges there was little [political] support for doing away with direct judicial election and so, ultimately, the principle was not incorporated in the proposed amendment." Tony A. Freyer & Paul M. Pruitt, Jr., Reaction and Reform: Transforming the Judiciary Under Alabama's Constitution, 1901-1975, 53 Ala. L. Rev. 77, 123 (2001).

The connection between Wallace's demonization of the federal judiciary and Heflin's and Torbert's later reform efforts is simply too tenuous for Plaintiffs to carry their burden to prove that Alabama rejected judicial appointments for discriminatory reasons. At the time of the reform efforts, judicial elections had been a fact of life in Alabama for over a century. It is therefore no surprise that Heflin and Torbert could not garner the necessary support to do away with them. Convincing people that they should give up the right to vote for an entire branch of government is hard to sell. And there is nothing in the record to support an inference that decisionmakers rejected judicial appointments because of the possibility that they would result in more African Americans on the bench or more pro-civil rights decisions. According to another of Dr. Norrell's sources, Heflin indicated that, by 1974, "court reform no longer possessed a racial stigma that politicians such as Wallace could profitably exploit." Tony A. Freyer, Paul M. Pruitt, Jr. & R. Volney Riser, Clement Clay Torbert and Alabama Law Reform, 63 Ala. L. Rev. 867, 884 (2012). Dr. Norrell failed to mention that Chief Justice Heflin apparently thought his judicial reform efforts were not tainted by race issues. See id. (quoting Chief Justice Heflin's pronouncement that his platform of court reform "wasn't an issue of race").

Nor is there any evidence that 2001 and 2004 efforts by African-American legislators to implement district-based elections for appellate judges were rejected for racial reasons. Alabama Representative Laura Hall, an African-American Democrat, backed legislation in 2000 and 2004 in the then-Democratic-controlled Alabama Legislature, to elect supreme court justices by district, but the bills never made it out of committee. Trial Tr. III, at 217-22 (Hall). But 306 other House bills introduced in 2000 suffered the same fate, as well as 336 other House bills introduced in 2004. Woodard Decl. (Defs.' Trial Ex. 3). Representative Hall acknowledged that bills often die in committee for many reasons unrelated to race, and she said she had no evidence (other than the fact that the bills did not make it out of committee) that a racial reason was behind those bills' defeats in a Democrat-controlled Legislature. Trial Tr. III, at 230 (Hall). Without more, Plaintiffs cannot prevail on their constitutional claims of intentional discrimination in the defeat of reform legislation.

Alabama's strong, consistent policy of statewide elections with numbered places also militates against a finding of discriminatory motive. "Under an intent test, a strong state policy in favor of at-large elections, for reasons other than race, is evidence that the at-large system does not have a discriminatory intent." Marengo Cty. Comm'n, 731 F.2d at 1571. As previously discussed, the State has weighty interests in linking appellate judges' jurisdiction with the electorate, preventing any judges from being captured by parochial interests, averting legislative interference with district lines, and ensuring that incumbents do not have to run against each other. The policy of statewide elections has been applied in appellate judicial races since 1868, and the numbered-place requirement has been applied in appellate judicial races and Public Service Commission races since 1927. Trial Tr. III, at 250- 52 (Norrell). Plaintiffs have not carried their burden to show a discriminatory motive in either of these aspects of Alabama's appellate judicial election system.

VI. CONCLUSION

Plaintiffs have not proven that the State's at-large, statewide and numbered-place features of Alabama's method for electing appellate judges — a method initiated for the Alabama Supreme Court in 1868 and sustained with the creation of an intermediate appellate court in 1911 and continuously in use to date — dilutes African-American voting strength in violation of § 2 or is racially discriminatory in violation of the Fourteenth and Fifteenth Amendments. Plaintiffs lose for seven primary reasons.

First, Plaintiffs have not proven that Alabama's at-large, statewide system of electing appellate judges was "imposed or applied . . . in a manner which results in a denial or abridgment of the right of any citizen . . . to vote on account of race or color." § 10301(a). They failed in their evidentiary burden to prove vote dilution on account of race in the totality of circumstances. Alabama's political process for the election of appellate judges is equally open to participation by African-American voters in that these voters have an equal opportunity to participate in the political process and to elect representatives of their choice. Proportional representation is not the standard. See § 10301(b) ("[N]othing in this section [§ 2 of the Voting Rights Act] establishes a right to have members of a protected class elected in numbers equal to their proportion in the population." (alterations added)).

Second, there are reasons other than race that account for election losses of candidates favored by African-American voters in appellate judicial elections in Alabama. The recent decline of the Alabama Democratic Party and a failure of black-preferred candidates to compete generally are two primary culprits. Plaintiffs presented no evidence that race is the reason for the Party's current failings or the reason why black-preferred candidates are not running for statewide appellate judge seats.

Third, the law of § 2 and of the Eleventh Circuit is entirely unfavorable to § 2 vote dilution claims involving state courts at either the trial or appellate level.

Fourth, assuming vote dilution on account of race (which was not proved), Plaintiffs have failed to prove a feasible § 2 remedy that outweighs the State's strong interests. Even if restructuring Alabama's election method on a district basis would give African Americans more control over the selection of one or two or three appellate judges in majority-black districts, that does not make districting a feasible remedy. Districted elections would sever the State's clear preference to link the appellate courts' jurisdiction and electoral base and would substantially interfere with Alabama's long-standing constitutional model of statewide elections for appellate judges. The State's interest in linkage is just as compelling for its appellate judges as it is for trial judges, given the authority of Alabama's appellate judges to render final and weighty decisions on every facet of the law arising under its constitution, statutes, and common-law. Moreover, districting has its own dilutive effect on all voters.

Fifth, there is no evidence of movement nationally among the states to change to districted elections of state appellate judges, with thirty-four states, including Alabama, having some form of an at-large election component in the selection of appellate judges and with only four states using districted elections. Nor is there evidence that the districting remedy is a fair comparator for at-large, statewide appellate judicial elections.

Sixth, public policy, including federalism concerns and a state's strong constitutional interest in determining its own form of government, overrides claims of and remedies for vote dilution on account of race.

Seventh, Plaintiffs have not proven that racial discrimination was a motivating factor in the adoption or maintenance of Alabama's at-large, statewide, numbered-place system for electing appellate judicial elections.

Alabama has come a long way because it had a long way to go. Much of Plaintiffs' evidence was of the "turn back the clock" variety which, while undeniable as historical fact, must have an evidentiary connection to present realities: losing statewide judicial elections in the present day. It turns out that blaming election losses in Alabama appellate judicial races on vote dilution is merely a "euphemism for political defeat at the polls." Whitcomb, 403 U.S. at 153.

African Americans have served at the highest reaches of state government, and they can do so again. There is a time for everything. Lest it be forgotten, Oscar Adams was appointed to the Supreme Court two years before George Wallace's last election as governor, and Justice Oscar Adams ran and won in 1982 on the same ballot with George Wallace. Justice Adams's success was the result of visionary political thinking by a strongly unified Democratic Party that rejected the politics of the past. But that time has slipped away; in politics, time victimizes times. Not so the law; it eschews the noise of now.

Based on the evidence, Alabama's at-large, statewide system of electing appellate judges today is benign of racial hostility, either overt or covertly lurking in the recesses of § 2, and is not racially discriminatory either in its adoption or maintenance.

Final judgment will be entered in favor of the State Defendants and against Plaintiffs.

DONE this 5th day of February, 2020.

/s/ W. Keith Watkins

UNITED STATES DISTRICT JUDGE

Image materials not available for display.