Opinion
Civil Action 3:22-CV-734-DPJ-HSO-LHS
07-02-2024
MISSISSIPPI STATE CONFERENCE OF THE NATIONAL ASSOCIATION FOR THE ADVANCEMENT OF COLORED PEOPLE, ET AL. PLAINTIFFS v. STATE BOARD OF ELECTION COMMISSIONERS, ET AL. DEFENDANTS
Before Southwick, Circuit Judge, Jordan, Chief District Judge, and Ozerden, District Judge .
Per Curiam.
The Mississippi State Conference of the National Association for the Advancement of Colored People and numerous individual black Mississippi voters brought this suit against the Mississippi State Board of Election Commissioners and other Mississippi officials. They challenged the State's 2022 redistricting maps for electing members of the state legislature. The Plaintiffs allege the 2022 maps dilute black Mississippian votes in violation of Section 2 of the Voting Rights Act of 1965, 52 U.S.C. § 10301, and contain unconstitutional racial gerrymanders in violation of the Fourteenth Amendment of the United States Constitution.
We find the Plaintiffs have satisfied their burden under Thornburg v. Gingles, 478 U.S. 30 (1986), by presenting three illustrative districts that satisfy the requirements of Section 2. See 52 U.S.C. § 10301. We conclude, however, that the Plaintiffs did not establish the redistricting maps are unconstitutional racial gerrymanders. We will provide the Mississippi Legislature an opportunity to enact revised maps.
Before explaining our ruling, we wish to state the court's appreciation to all counsel for their professionalism. Voting-rights litigation can be contentious. In this lawsuit, though, lawyers and witnesses were respectful and measured in the expression of the disagreements about the law and the evidence. Some very wise and experienced Mississippi voting-rights lawyers appeared on both sides of this dispute. New and quite able lawyers from within and beyond the state's borders appeared as well. All are to be commended.
I. FACTUAL AND PROCEDURAL BACKGROUND A. Parties
The Mississippi State Conference of the National Association for the Advancement of Colored People (“Mississippi NAACP”) is the only organizational plaintiff in this suit. It is a subsidiary organization of the National Association for the Advancement of Colored People, Inc., a nonprofit organization founded in 1909. Stipulations [199] App. A at 1. Some Mississippi NAACP members are registered voters who reside in the five legislative districts that the Plaintiffs challenge as unconstitutional racial gerrymanders. Id.; Doc [220], 6.
Fourteen individual Mississippians joined the Mississippi NAACP as Plaintiffs in this case. Doc [199], 2-5. These individuals are all registered voters who live in one of the challenged legislative districts and consider themselves to be Democrats. Id. Several individual Plaintiffs are also members of the Mississippi NAACP. Id. We refer to these parties collectively as the Plaintiffs.
The Individual Plaintiffs named in the complaint are Dr. Andrea Wesley, Dr. Joseph Wesley, Robert Evans, Gary Fredericks, Pamela Hamner, Barbara Finn, Otho Barnes, Shirlinda Robertson, Sandra Smith, Deborah Hulitt, Rodesta Tumblin, Dr. Kia Jones, Angela Grayson, Marcelean Arrington, and Victoria Robertson. Doc [199], 2-5. Angela Grayson withdrew as a Plaintiff in September 2023. Doc [84] (Order Granting Withdrawal).
The Mississippi State Board of Election Commissioners is composed of the Attorney General, the Mississippi Governor, and the Secretary of State. Miss. Code Ann. § 23-15-211. That board, Mississippi Attorney General Lynn Fitch, Mississippi Governor Tate Reeves, and Mississippi Secretary of State Michael Watson were all named as Defendants in this suit. Id. at 5-6. The Mississippi Republican Executive Committee moved to intervene as a Defendant, and the motion was granted in May 2023. Doc [32] (Motion); Text-Only Order May 19, 2023 (Granting Motion). We refer to these parties collectively as the Defendants.
B. The Mississippi Legislature's Actions
The United States Census Bureau conducts a census during the first year of each decade and releases the results the following year to every state. See 13 U.S.C. § 141. Every ten years following the release of the census data, the Mississippi Legislature is to “apportion the state in accordance with the Constitution of the state and of the United States into consecutive numbered Senatorial and Representative districts of contiguous territory.” Miss. Const. art. XIII, § 254. The Mississippi Legislature's Standing Joint Legislative Committee on Reapportionment and Redistricting (“Standing Joint Committee”) is the legislative body responsible for drafting each Mississippi reapportionment plan for state senate and state house membership once the decennial census data is delivered. See Miss. Code Ann. §§ 5-3-91 to 5-3-103. The Standing Joint Committee must draw plans “according to constitutional standards” and state guidelines. §§ 5-3-93, 5-3-101. The state guidelines are:
(a) Every district shall be compact and composed of contiguous territory and the boundary shall cross governmental or political boundaries the least number of times possible; and
(b) Districts shall be structured, as far as possible and within constitutional standards, along county lines; if county lines are fractured, then election district lines shall be followed as nearly as possible.§ 5-3-101. Once the redistricting maps are drafted and approved by the Standing Joint Committee, each house must approve a joint resolution that sets out the maps. Miss. Const. art. XIII, § 254. The Mississippi Governor has no official role in the approval of the maps.
In response to COVID-19, the Census Bureau was required to adapt and delay some of its 2020 Census operations. See U.S. Census Bureau, U.S. Dep't of Com. and Labor, 2020 Census Operational Adjustments: Changes Due to COVID-19 (2020), https://www.census.gov/content/dam/Census/library/factsheets/2023/d ec/operational-adjustments-covid-19.pdf. The release of redistricting data had an original statutory deadline of March 31, 2021, but it was not released to states until August 12, 2021. Id.; Stipulations [199] App. A at 10. The 2020 Census data showed that over the last ten years, Mississippi's overall population declined from 2,967,297 to 2,961,279 - a 6,018-person decrease from the 2010 Census. Stipulations [199] App. A at 11. The non-Hispanic White population decreased by 83,210 persons, and the Any-Part Black population increased by 7,812 persons. Id. The Any-Part Black population constitutes 37.94 percent of the state's total population and 36.14 percent of the voting-age population, making it the largest minority population in Mississippi. Id. Based on this population data, the ideal district size for a state senate district is 56,948, and the ideal size for a state house district is 24,273. Id. at ¶ 57.
Following the release of the 2020 Census data, the Mississippi Legislature convened the Standing Joint Committee, which conducted nine public hearings and held four public meetings over the course of nine months. Doc [199], 10-11; Stipulations [199] App. A at 12. Members of the public were invited to participate in the public hearings held between August 5, 2021, and August 23, 2021, and to provide input on the redistricting process. Stipulations [199] App. A at 12. No proposed redistricting maps were revealed to the public at these hearings, nor was the public given the opportunity to comment on the maps. Id.
In a 15-minute November 2021 open meeting, the Standing Joint Committee adopted the following criteria for drawing the state legislative districts:
(1) Each district's population should be less than 5 [percent] above or below the ideal population of the district.
(2) Districts should be composed of contiguous territory.
(3) The redistricting plan should comply with all applicable state and federal laws including Section 2 of the Voting Rights Act of 1965, as amended, and the Mississippi and United States Constitutions.Doc [199], 10; Stipulations [199] App. A at 12; JTX-013. Thus, the population must be within 5 percent of 56,948 for a state senate district and within 5 percent of 24,273 for a state house district. Stipulations [199] App. A at 11-12. In addition to these three criteria, separation of incumbents into different districts is a legally available redistricting standard. Bush v. Vera, 517 U.S. 952, 964-65 (1996) (plurality opinion). The Standing Joint Committee also considered precinct-level voting-age and voting-age racial-demographic information as required by the Voting Rights Act when devising districts. Stipulations [199] App. A at 12-13; see 52 U.S.C. § 10301.
The Standing Joint Committee held its final meeting on March 27, 2022, and it publicly revealed the proposed maps for the state legislative districts. Stipulations [199] App. A at 13. At five o'clock that same day, the Standing Joint Committee voted to adopt both proposed maps, with Representative Bo Brown, Senator Angela Turner-Ford, and Senator Derrick Simmons voting against the proposed state house map. See Id. at ¶ 63; JTX-014.
On March 29, both the full state house and state senate voted to adopt their respective districting plans. Stipulations [199] App. A at 13. Two amendments were proposed to the state house districting plan, JR 1. JTX-010, 31:20-24, 23:25-46:20. The first amendment was adopted by voice vote to separate the districts for two black Democratic representative incumbents, and the second amendment supplying an alternative map portraying an additional five black-majority districts failed by a vote of 77 to 39 as an admitted racial gerrymander. Id. at 47:20-21. The two amendments proposed to the state senate districting plan, JR 202, also failed. Those amendments would have added four black-majority districts. Democratic Senator Simmons, who advocated for adding four majority-black senate seats, emphasized that the black population in Mississippi had grown and the white population had shrunk. Therefore, he argued, a “map that maintains the status quo simply dilutes black voting strength in Mississippi.” JTX-011, 98:09-99:12; see also Stipulations [199] App. A at 13; Doc [219], 9 n.5. Republican Senator Dean Kirby responded by contending that Senator Simmons's proposed map “is not a map this state needs.” JTX-011, 99:19- 20. The senate approved the map released by the Standing Joint Committee. JTX-011, 194:13-14; see also Doc [219], 9 n.5; Doc [220], 4.
On March 31, the state house approved JR 202, the state senate approved JR 1, and the maps (the “Enacted Plans”) became law. Stipulations [199] App. A at 13. The Enacted Plans were determined to be in compliance with the Mississippi Code, containing only one senate district and three house districts where two incumbents were paired together. See Miss. Code Ann. § 5-3-101; see also Bush, 517 U.S. at 964-65.
The Mississippi Legislature consists of 52 state senate districts and 122 state house districts. Stipulations [199] App. A at 13, 15. Of the 52 senate districts, the Enacted Senate Plan renders 15 black-majority districts, and of the 122 house districts, the Enacted House Plan renders 42 black-majority districts based on the 2020 Census data. Id. at 14-15. Both Enacted Plans retained the same number of black-majority districts used for the last state legislative elections prior to the 2020 Census. Doc [220], 2. In doing so, the floor debates of both houses contained generalized statements from Senator Kirby and Representative Beckett that “maintaining the political performance” of Republicans and Democrats throughout the state was an “important consideration in developing th[e] proposed plan[s],” yet those statements were not made in relation to the specific districts in this case, nor was there any data mentioned that was used to achieve that general goal. JTX-010, 14:25-15:06; see JTX-011, 12:06-11. The only specifics provided were that years of service and incumbency were considered factors of “political performance.” JTX-010, 20:09-24. The Defendants use that phrase without specifically defining it. We take from the record, though, that years of service and incumbency are aspects of political performance. If political performance is distinguishable from incumbent protection, the differences would seem to be minor.
The legislature enacted plans for both houses in 2012 following the release of the 2010 Census data; in 2019, the legislature adopted a revised plan for the senate as a result of a court order. See Thomas v. Bryant, 938 F.3d 134, 140, 143 (5th Cir. 2019). The 2012 State Senate Plan contained 15 black-majority districts, and the 2012 State House Plan contained 42 black-majority districts. Stipulations [199] App. A at 14-15 ¶¶ 67, 79. The United States Department of Justice precleared the plans as was then required under Section 5 of the Voting Rights Act. See Shelby County v. Holder, 570 U.S. 529, 556-57 (2013). The 2019 Senate Plan that created one more black-majority district was used in the 2019 elections, but the senate districts reverted to the 2012 geographical boundaries following the Fifth Circuit's vacating the 2019 Senate Plan because the litigation that caused its creation was moot. Thomas v. Reeves, 961 F.3d 800, 801 (5th Cir. 2020).
The Standing Joint Committee played no further role in the redistricting process once the Enacted Plans became state law. It is the responsibility of various other state and local officials to enforce and implement the law and to administer elections. See Miss. Code Ann. § 23-15-211, et seq. The Enacted Plans established the current makeup of the Mississippi Legislature as of January 15, 2024: 16 Democrats and 36 Republicans in the state senate; 41 Democrats, 79 Republicans, and 2 Independents in the state house; 14 black senators and 38 white senators; and roughly 40 black representatives and 82 white representatives. Stipulations [199] App. A at 16.
C. The Litigation
The Plaintiffs filed this civil action in December 2022, challenging the Enacted Plans as violations of both the Equal Protection Clause of the Fourteenth Amendment and Section 2 of the Voting Rights Act and seeking declaratory and injunctive relief. Doc [219], 13; Doc [220], 4; Doc [1] (Complaint). Two former defendants, State Senator Dean Kirby and State Representative Dan Eubanks, filed a motion to dismiss for lack of jurisdiction over them, after which the Plaintiffs amended their complaint and named only the above-mentioned Defendants. Docs. [17], [18], [24], [27]. The Mississippi Republican Executive Committee later intervened as a Defendant. Doc [32] (Motion); Text-Only Order May 19, 2023 (Granting Motion).
As part of their Section 2 claim, the Plaintiffs assert Mississippi's black population requires at least four additional black-majority senate districts and at least three additional black-majority house districts be drawn. Doc. [27], 3. In some of the same areas where these alleged districts can be drawn, the Plaintiffs contend two enacted senate districts and three enacted house districts are unconstitutional racial gerrymanders that were drawn with race as the predominant factor. Id. at 4; Doc [220], 4-5 (see chart on pg. 5). To support their contention, the Plaintiffs proposed two Illustrative Plans - one Illustrative Senate Plan and one Illustrative House Plan - that contained seven illustrative majority-minority districts to be added to the current number of majority-minority districts. PTX-001, 26-79, Ex. M & AH. According to the Defendants, were the Mississippi Legislature to accept the Plaintiffs' seven illustrative districts, it would cause ripple effects in more than 70 current electoral districts. PTX-001, 28 ¶ 53, 60 ¶ 122.
An order for a trial was entered in June 2023, but the trial did not begin until February 26, 2024, before a three-judge panel as required by statute. See 28 U.S.C. § 2284(a); Doc [3], 40, 44; see Text-Only Order June 23, 2023 (Setting Trial Deadlines). The trial lasted eight days, during which fourteen Plaintiff witnesses and three Defense witnesses testified. See generally Docs [212], [215]. Thereafter, the Plaintiffs and Defendants, along with the Intervenor Defendant, filed post-trial briefs and proposed findings and conclusions of law. Docs. [218]-[221].
II. GENERAL LEGAL PRINCIPLES
Legislative reapportionment “is primarily the duty and responsibility of the States, not the federal courts.” Allen v. Milligan, 599 U.S. 1, 29 (2023) (quotation marks and citation omitted). However, a State's actions throughout its redistricting process can be challenged under both the Fourteenth Amendment and the Voting Rights Act. The Equal Protection Clause of the Fourteenth Amendment prohibits states from denying “any person within [their] jurisdiction the equal protection of the laws.” U.S. Const. amend. XIV, § 1. “Its central purpose is to prevent the States from purposefully discriminating between individuals on the basis of race.” Shaw v. Reno, 509 U.S. 630, 642 (1993). Improper racial gerrymanders used in legislative redistricting - like those alleged here - fall within the Fourteenth Amendment's prohibition. Cooper v. Harris, 581 U.S. 285, 291 (2017).
“‘Electoral districting is a most difficult subject for legislatures,' requiring a delicate balancing of competing considerations.” Bethune-Hill v. Va. State Bd. of Elections, 580 U.S. 178, 187 (2017) (quoting Miller v. Johnson, 515 U.S. 900, 915 (1995)). Section 2 of the Voting Rights Act prohibits any “standard, practice, or procedure” imposed by a State throughout its electoral districting process that “results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color.” 52 U.S.C. § 10301(a). “The essence of a [Section] 2 claim is that a certain electoral law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives.” Gingles, 478 U.S. at 47. “Such a risk is greatest ‘where minority and majority voters consistently prefer different candidates' and where minority voters are submerged in a majority voting population that ‘regularly defeats' their choices.” Milligan, 599 U.S. at 18 (quoting Gingles, 478 U.S. at 48). The Supreme Court thus requires Section 2 vote-dilution claims be analyzed using the Gingles three-part framework. Id. at 17. If a court concludes that plaintiffs do not have an equal opportunity to elect their preferred candidate under a challenged districting map, there likely is a Section 2 violation. See Gingles, 478 U.S. at 44; Robinson v. Ardoin, 86 F.4th 574, 589 (5th Cir. 2023).
The Plaintiffs presented evidence at trial under both the Equal Protection Clause and Section 2. We will discuss each theory separately.
III. EQUAL PROTECTION CLAIM
Racial gerrymandering occurs when race improperly motivated the drawing of electoral districts such that it “rationally cannot be understood as anything other than an effort to separate voters” based on race. Shaw, 509 U.S. at 649; see Alabama Legis. Black Caucus v. Alabama, 575 U.S. 254, 262- 63 (2015). Courts conduct “a two-step analysis” to determine whether a legislative district is an unconstitutional racial gerrymander. Harris, 581 U.S. at 291. “First, the plaintiff must prove that ‘race was the predominant factor motivating the legislature's decision to place a significant number of voters within or without a particular district.'” Id. (quoting Miller, 515 U.S. at 916). If “the legislature subordinated other factors - compactness, respect for political subdivision, partisan advantage, [etc.] - to racial considerations,” race is the predominant factor. Id. (quotation marks and citation omitted).
The plaintiff has the burden of proof when asserting a challenged district was enacted with discriminatory intent. Abbott v. Perez, 585 U.S. 579, 603 (2018). Because of the “serious intrusion” courts make into “the most vital of local functions” when analyzing redistricting cases, the “good faith of the state legislature must be presumed.” Id. (quoting Miller, 515 U.S. at 915). This presumption, however, can be overcome if the plaintiff makes a sufficient showing that the legislature subordinated traditional race-neutral districting principles to race. See Miller, 515 U.S. at 915-16. Plaintiffs can use “direct evidence of legislative intent, circumstantial evidence of a district's shape and demographics,” or both to meet this burden and to establish race as the predominant factor. Harris, 581 U.S. at 291 (quotation marks and citation omitted).
The second step in the analysis shifts the burden to the State to withstand strict scrutiny. Id. at 292. If race was the predominant factor, the State must prove that “its race-based sorting of voters” and the district's design serve a “compelling interest” and are “narrowly tailored to that end.” Id. (quotation marks and citation omitted). The Supreme Court recognizes compliance with the Voting Rights Act - including Section 2 - as a compelling interest. Id. Using Section 2 to justify race-based districting, however, requires a State show “it had ‘a strong basis in evidence' for concluding that the statute required its action.” Id. (quoting Alabama Legis. Black Caucus, 575 U.S. at 278). In other words, the legislature must have had “good reasons” to believe such district drawing was necessary for statutory compliance. Alabama Legis. Black Caucus, 575 U.S. at 278.
The racial-predominance inquiry “concerns the actual considerations that provided the essential basis for the lines drawn, not post hoc justifications the legislature in theory could have used but in reality did not.” Bethune-Hill, 580 U.S. at 189-90. It concerns how the legislature conducted the redistricting, not why the legislature conducted it that way. See Harris, 581 U.S. at 308 n.7. The Plaintiffs thus bear the burden under this inquiry “to show, either through circumstantial evidence of a district's shape and demographics or more direct evidence going to legislative purpose, that race was the predominant factor motivating the legislature's [redistricting] decision,” Bethune-Hill, 580 U.S. at 187 (quoting Miller, 515 U.S. at 916), “regardless of [the legislature's] ultimate objective,” Harris, 581 U.S. at 308 n.7.
Unlike a Section 2 claim that may be established by discriminatory results, the Plaintiffs were required to provide sufficient evidence of “race-based decisionmaking” to prove their Equal Protection claim. See Miller, 515 U.S. at 915. In other words, the Plaintiffs' task “is simply to persuade the trial court - without any special evidentiary prerequisite - that race (not politics) was the predominant consideration” of the Mississippi Legislature during the redistricting process when drawing the Enacted Plans. Harris, 581 U.S. at 318 (quotation marks and citation omitted). This is a demanding burden, and “[p]roving racial predominance with circumstantial evidence alone is much more difficult” when partisanship is raised as a defense. Alexander v. S.C. State Conf. of the NAACP, 144 S.Ct. 1221, 1234-35 (2024).
“[P]artisan and racial gerrymanders ‘are capable of yielding similar oddities in a district's boundaries' when there is a high correlation between race and partisan preference.” Id. at 1235 (quoting Harris, 581 U.S. at 308). “When partisanship and race correlate, it naturally follows that a map that has been gerrymandered to achieve a partisan end can look very similar to a racially gerrymandered map.” Id. To prevail on their racial gerrymandering claim, the Plaintiffs “must disentangle race from politics by proving that the former drove a district's lines.” Id. (emphasis in original) (quotation marks and citation omitted).
The Intervenor Defendant insisted throughout trial the Plaintiffs stipulated this constitutional claim away with one of their interrogatory responses:
9. Does the identification of invidiously discriminatory intent on the basis of race constitute technical knowledge which may be the subject of expert testimony under Rule 702? Plaintiffs' Response: Plaintiffs do not assert any claims premised on invidious discriminatory intent on the basis of race.Doc [199], 16 (Pretrial Order). The Plaintiffs, however, attempt to distinguish “invidious” behavior and the necessary predominance under Equal Protection. Footnote 10 of the Plaintiffs' post-trial brief argues it this way:
In contrast to claims involving racial animus or invidious discrimination, “the essence of the [E]qual [P]rotection claim recognized in Shaw is that the State has used race as a basis for separating voters into districts.” [Miller, 515 U.S. at 911.] Racial gerrymandering claims are thus cognizable even when the State undertakes the racial sorting with a “benign” motivation. Shaw, 509 U.S. at 653; see also Bush, 517 U.S. at 984. Consistent with that distinction, racial gerrymandering claims require that racial sorting be the “predominant”
consideration affecting a “significant” number of voters, whereas claims of invidious discrimination are actionable even if discriminatory purpose is merely “a motivating factor” affecting any racial group, regardless of the number of individuals affected. Compare [Harris], 581 U.S. at 291-92 with Vill[age] of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 265 (1977).Doc [221], 25 n.10.
One of the cited precedents discusses whether a statute's explicit racial classification was “benign” and subject to the less-searching review Justice Souter argued for in dissent or was “motivated by illegitimate notions of racial inferiority or simple racial politics.” Shaw, 509 U.S. at 642-43, 653 (quotation marks and citation omitted). The Supreme Court held that either motive, if focused on race, was subject to strict scrutiny under the Equal Protection Clause “because without it, a court cannot determine whether or not the discrimination truly is ‘benign.'” Id. at 653.
Other caselaw indicates racial reasons might be able to survive strict scrutiny if the State believes a racial gerrymander enactment is needed to satisfy the Voting Rights Act. As recently as 2022, the Supreme Court stated that a consciously gerrymandered district could be upheld for that reason:
Under the Equal Protection Clause, districting maps that sort voters on the basis of race “‘are by their very nature odious.'” Shaw[,] 509 U.S. [at] 643[.] Such laws “cannot be upheld unless they are narrowly tailored to achieving a compelling state interest.” Miller[,] 515 U.S. [at] 904[.] We have assumed that complying with the VRA is a compelling interest. [] Harris, 581 U.S. [at 292]. And we have held that if race is the predominant factor motivating the placement of voters in or out of a particular district, the State bears the burden of showing that the design of that district withstands strict scrutiny. Ibid. Thus, our precedents hold that a State can
satisfy strict scrutiny if it proves that its race-based sorting of voters is narrowly tailored to comply with the VRA. Ibid.Wisconsin Legis. v. Wis. Elections Comm'n, 595 U.S. 398, 401 (2022). The word “invidious” was not used in either the 2022 Wisconsin Legislature opinion or the 2017 Harris opinion that it cites. The Supreme Court did use “invidious” in its 1995 Miller opinion, which also dealt with a motive of advantaging minorities. There, the State argued that the Department of Justice made it utilize race-based redistricting:
Our presumptive skepticism of all racial classifications prohibits us as well from accepting on its face the Justice Department's conclusion that racial districting is necessary under the Act. Where a State relies on the Department's determination that race-based districting is necessary to comply with the Act, the judiciary retains an independent obligation in adjudicating consequent equal protection challenges to ensure that the State's actions are narrowly tailored to achieve a compelling interest.Miller, 515 U.S. at 922 (citation omitted).
The Miller Court noted it was uncontested that the parties “practically stipulated” that a particular district was a racial gerrymander and explicitly discussed the “maximizing majority-black” policy that drove the enactment of the “max-black plan.” Id. at 910, 924. The majority never tried to distinguish the advantaging-minorities motive from harming them, simply referring to “invidious discrimination” in its concluding paragraph as the label for what the legislature had done even though it was trying to help the minority population have an equal opportunity to gain public office. Id. at 927. The Court reprimanded any form of “automatic invocation of race stereotypes” as they “retard[] that progress and cause[] continued hurt and injury” to the minority population in election law. Id.
There is little caselaw discussing whether favorable motives are “invidious” even when they fail under strict scrutiny as they did in Miller. We thus give the Plaintiffs some leeway as to the relevance of the label “invidious” considering the Supreme Court recognized racial motives that might be upheld in Shaw, Harris, and Wisconsin Legislature. Still, no precedent has made the point the Plaintiffs urge here - that the word “invidious” does not apply to some unacceptable motives. The only racial motives so far discussed by the Court that might be upheld are used to comply with the Voting Rights Act. Wisconsin Legis., 595 U.S. at 401; Miller, 515 U.S. at 922.
The allegations here are not that. The Plaintiffs allege that the Enacted Plans contain unconstitutional racial gerrymanders; they do not challenge them as failed efforts possibly intended to comply with federal law. Thus, it would take invidious discrimination to make an Equal Protection case here. The Plaintiffs clarified they “are not alleging invidious” discrimination or “animus” as part of their Equal Protection claim. Trial Tr. 1672:6-8. Without that intent, however, the Plaintiffs cannot meet their required burden. See Abbott, 585 U.S. at 603.
Even if we assume the Plaintiffs preserved a benign-motive case, there was insufficient evidence of racial predominance under the Equal Protection Clause. The Plaintiffs rely on the testimony of Dr. Jordan Ragusa, a tenured professor of political science at the College of Charleston, to prove race predominated in the design of the five challenged districts. Trial Tr. 968:16- 971:10. We accepted Dr. Ragusa as an expert in quantitative methods and analysis, the modeling of electoral districting, and American politics, Trial Tr. 980:17-25, and allowed his explanation as to whether he disentangled the effects of race, partisanship, and traditional redistricting principles on the Enacted Plans, Trial Tr. 981:9-18.
Dr. Ragusa conducted a statistical analysis comparing Census data to the Enacted Plans and created multivariate logistic-regression models of the information. Trial Tr. 981:19-982:5. In doing so, the Plaintiffs contend Dr. Ragusa was able to evaluate the Mississippi Legislature's decisions made during the redistricting process and determine what factor predominantly influenced its decisions. Trial Tr. 982:6-15. Dr. Ragusa's regression models showed the effect of the black voting-age population (“BVAP”) was statistically significant in at least one model for all five districts, which Dr. Ragusa testified allowed him to “conclude that race was a significant factor in all five of the challenge[d] districts.” Trial Tr. 982:21-24. The Plaintiffs assert Dr. Ragusa's regression models sufficiently control for partisanship and other factors and show race was the motivating factor in moving voters in and out of districts. Doc [221], 27-28. Thus, according to the Plaintiffs, their Equal Protection claim was clearly established. Id.
The Defendants, however, persuasively discredited those models and articulated how they establish partisanship was just as much a factor in the drawing of district lines as race. Doc [218], 6-7. One problem with Dr. Ragusa's analysis is that he used what he called the “county envelope” method to determine whether geographical units called census blocks, see Trial Tr. 983:23-986:9 (discussing census blocks), were more likely to be moved into a district based on their racial composition, Trial Tr. 995:4- 996:8. The Supreme Court recently concluded that the county-envelope model is flawed because it inadequately accounts for contiguity and compactness. See Alexander, 144 S.Ct. at 1246. Further, even though Dr. Ragusa initially attempted to conclude his findings could not “be dismissed as a simple byproduct of partisan gerrymandering or adherence to th[e] common redistricting principles,” Trial Tr. 983:3-5, he later stated he was “unable to offer definitive proof” of racial gerrymandering because “[t]he analysis can't disentangle intentional from unintentional racial discrimination. Both are reasonable explanations for the results . . . but [he] can't say one way or the other what caused the effect of race. What [he] can say is that race was a factor.” Trial Tr. 1120:12-20.
Thus, at most, the Plaintiffs' expert's opinion was that race was one factor in the Mississippi Legislature's redistricting that merged with partisan protection. That is not enough. The Enacted Plans must have predominantly utilized race such that the only rational way to view the districts is “an effort to separate voters” based solely on their race. Shaw, 509 U.S. at 649. Because Plaintiffs failed to prove that race was the predominant factor, we do not consider the evidence and specific arguments that two enacted senate districts and three enacted house districts were racial gerrymanders. Without invidious intent, the Plaintiffs cannot establish an Equal Protection violation.
IV. SECTION2OFTHEVOTINGRIGHTSACT
A. Private Right of Action Under Section 2
We start with the Defendants' argument that Section 2 of the Voting Rights Act cannot be enforced by private parties. They acknowledge foreclosure of the issue in the Fifth Circuit because of a recent opinion recognizing such a right. See Robinson, 86 F.4th at 588. The Defendants present their no-private-right-of-action arguments to preserve the issue for later review.
Though the Robinson holding is binding on this court, that opinion's explanation was brief, responding in kind to the limited argument made by Louisiana in that appeal. Further review of the private-right-of-action issue in the present case may occur, and we include additional analysis now.
The Robinson court concluded the issue was essentially foreclosed by a Fifth Circuit precedent holding that Section 2 of the Voting Rights Act abrogated state sovereign immunity. Id. (citing OCA-Greater Hous. v. Texas, 867 F.3d 604, 614 (5th Cir. 2017)). The court determined that the abrogation was not without purpose and that private individuals had been granted a right to sue under Section 2. Id.
The Robinson court did not discuss the revamped analysis used by the Supreme Court since 2001 for determining whether a congressional enactment may be enforced by private individuals. See Gonzaga Univ. v. Doe, 536 U.S. 273, 284 (2002); Alexander v. Sandoval, 532 U.S. 275, 286 (2001). That analysis first examines the relevant statute for an explicit grant of a right to sue. See, e.g., Gonzaga Univ., 536 U.S. at 283-84. Absent such welcomed textual clarity, a private cause of action may be implied when a statute (1) contains rights-creating language and (2) displays “an intent ‘to create . . . a private remedy.'” Id. at 284 (quoting Sandoval, 532 U.S. at 286).
An overlapping though separate inquiry is “whether a statutory violation may be enforced through [42 U.S.C.] § 1983.” Id. at 283. “Plaintiffs suing under § 1983 do not have the burden of showing an intent to create a private remedy because § 1983 generally supplies a remedy for the vindication of rights secured by federal statutes.” Id. at 284. Therefore, if “a plaintiff demonstrates that a statute confers an individual right, the right is presumptively enforceable by § 1983.” Id.
In looking for congressionally created private rights that can be protected under Section 1983, the Gonzaga and Sandoval opinions rejected earlier, less-demanding approaches for deciding when Congress has created a private right in a statute. See Id. at 282-83; see also Sandoval, 532 U.S. at 287 (stating the Supreme Court had “sworn off the habit of venturing beyond Congress's intent” on whether a statute created private rights).
We start with the text of Section 2 of the Voting Rights Act:
(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.52 U.S.C. § 10301.
There certainly is an emphasis on rights in subsection (a) when it prohibits states taking actions that “result[] in a denial or abridgement of the right . . . to vote.” § 10301(a). Less obvious is textual language revealing a congressional intent to provide for a private remedy. In considering how to evaluate what Section 2 does and does not state, we find it useful to review an Eighth Circuit opinion issued after the Fifth Circuit's Robinson opinion. See Arkansas State Conf. NAACP v. Ark. Bd. of Apportionment, 86 F.4th 1204 (8th Cir. 2023). There, the majority held Section 2 provided no private right of action. Id. at 1206-07. It also determined that the plaintiffs had neither included in their complaint nor made a meaningful argument that their rights could be enforced using Section 1983. Id. at 1218. The Plaintiffs in the case before us included Section 1983 in their complaint as one of the bases for the suit, thus there is no waiver of the argument. Amended Comp. ¶ 11.
With respect for the Eighth Circuit majority, we find Chief Judge Smith's dissent in that case to express the more persuasive analysis. See Arkansas State Conf. NAACP, 86 F.4th at 1218-24 (Smith, C.J., dissenting). Of course, we are bound by Robinson regardless of which opinion persuades. Still, we explain what we find persuasive about the dissent.
The Eighth Circuit majority and dissent each discuss a 1996 Supreme Court opinion in which a plurality stated that “the existence of the private right of action under Section 2 . . . has been clearly intended by Congress since 1965.” Morse v. Republican Party of Va., 517 U.S. 186, 232 (1996) (plurality opinion) (quoting Voting Rights Extension, S. Rep. No. 97-417 (1982), 30, as reprinted in 1982 U.S.C.C.A.N. 177)). Though we find that a majority of the Morse Court agreed with that part of the plurality opinion, it may have been dicta. See Arkansas State Conf. NAACP, 86 F.4th at 1215.
It was not until a little more than 35 years after the Voting Rights Act was enacted that the Supreme Court articulated a test for determining whether Congress has created a private right of action to enforce a particular enactment. See Sandoval, 532 U.S. at 286.
Even if the Morse statement is dicta, the Fifth Circuit has held that it is generally bound by Supreme Court dicta, especially when it is “recent and detailed.” Gearlds v. Entergy Servs., Inc., 709 F.3d 448, 452 (5th Cir. 2013). The Morse opinion is neither particularly recent nor detailed, but the Fifth Circuit seemingly would tread cautiously before taking a different path. As to the second issue - that the decision might be questionable because it predates Sandoval - we find that the Supreme Court has told inferior courts to remain faithful to its on-point precedent: “[I]f a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to [the Supreme Court] the prerogative of overruling its own decisions.” Agostini v. Felton, 521 U.S. 203, 237 (1997) (citation omitted).
Also weighing against any court's marking its own path is that, before Sandoval and continuing with quite recent opinions, there has been substantial caselaw in which “the Federal Government and individuals have sued to enforce [Section] 2, and injunctive relief [was held to be] available in appropriate cases to block voting laws from going into effect.” Shelby County, 570 U.S. at 537 (citations omitted). The existence of a private right of action in opinions such as Shelby County has simply been assumed by courts. Indeed, for decades, Supreme Court and circuit court opinions have resolved Section 2 cases brought by private parties without addressing whether there was even a right for those parties to sue.
Eighth Circuit Chief Judge Smith reviewed this same history and expressed prudent caution:
Furthermore, since the Court decided Morse, “scores if not hundreds of cases have proceeded under the assumption that Section 2 provides a private right of action. All the while, Congress has consistently reenacted the VRA without making substantive changes, impliedly affirming the previously unanimous interpretation of Section 2 as creating a private right of action.” Coca [v. City of Dodge City, 669] F.Supp.3d [1131, 1140 (D. Kan. 2023)]. . . . Until the Supreme Court instructs otherwise, I would hold that [Section] 2 contains an implied private right of action.
Arkansas State Conf. NAACP, 86 F.4th at 1223-24 (Smith, C.J., dissenting).
Finally, few congressional enactments have had a more profound effect on the country than the Voting Rights Act of 1965, and a large percentage of the enforcement actions under the Voting Rights Act have been brought by private individuals. Though the following does not come from statutory text, it is noteworthy that on the page of the Senate Report immediately following the discussion of what we now call the Senate Factors appears this assertion: “the Committee reiterates the existence of the private right of action under Section 2, as has been clearly intended by Congress since 1965. See Allen v. Board of Elections, 393 U.S. 544 (1969).” S. Rep. No. 97-417, at 30.
If a court now holds, after almost 60 years, that cases filed by private individuals were never properly brought, it should be the Supreme Court, which has the controlling word on so momentous a change. Regardless, we are bound by the Fifth Circuit Robinson opinion that the Plaintiffs may properly bring this suit to enforce their rights under the Voting Rights Act.
B. The Gingles Framework
To succeed in proving a Section 2 vote-dilution claim, plaintiffs must first satisfy the three Gingles preconditions. Milligan, 599 U.S. at 18. “First, the minority group must be sufficiently large and [geographically] compact to constitute a majority in a reasonably configured district.” Id. (alteration in original) (quotation marks and citation omitted). A district is reasonably configured when it complies “with traditional districting criteria, such as being contiguous and reasonably compact.” Id. Second, the minority group must be politically cohesive. Id. Third, the white majority must be shown to vote sufficiently as a bloc to usually defeat the minority-preferred candidate. Id. “The third precondition, focused on racially polarized voting, ‘establish[es] that the challenged districting thwarts a distinctive minority vote' at least plausibly on account of race.” Id. at 19 (alteration in original) (quoting Growe v. Emison, 507 U.S. 25, 40 (1993)).
If a plaintiff fails to establish any one of these three preconditions, a court need not consider the others nor continue the analysis. See League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 425 (2006) [hereinafter LULAC v. Perry].
The first and second Gingles preconditions “are needed to establish that the minority has the potential to elect a representative of its own choice in some single-member district.” Emison, 507 U.S. at 40. The second and third Gingles preconditions “are needed to establish that the challenged districting thwarts a distinctive minority vote by submerging it in a larger white voting population.” Id. “Unless these points are established, there neither has been a wrong nor can be a remedy.” Id. at 40-41.
Although “the Gingles requirements cannot be applied mechanically and without regard to the nature of the claim[,] . . . [i]t remains the rule . . . that a party asserting [Section] 2 liability must show by a preponderance of the evidence” that all three preconditions are met. Bartlett v. Strickland, 556 U.S. 1, 19-20 (2009). The Plaintiffs must meet this burden, and “[a]ny lack of evidence in the record regarding a violation of the Voting Rights Act . . . must be attributed to” the Plaintiffs. League of United Latin Am. Citizens #4552 v. Roscoe Indep. Sch. Dist., 123 F.3d 843, 846 (5th Cir. 1997).
Once a plaintiff sufficiently establishes the three threshold preconditions, he must then “show, under the totality of the circumstances, that the political process is not equally open to minority voters,” which, in turn, establishes a Section 2 violation. Milligan, 509 U.S. at 18 (quotation marks and citation omitted). Courts are guided by factors identified by the Supreme Court in their totality-of-the-circumstances analyses. League of United Latin Am. Citizens, Council No. 4434 v. Clements, 999 F.2d 831, 849 (5th Cir. 1993) (en banc) [hereinafter LULAC v. Clements]; Robinson, 86 F.4th at 589 n.2. We will identify and analyze these factors after reviewing the three preconditions.
1. Gingles Precondition One
“The first Gingles precondition focuses on [the] geographical compactness and numerosity” of the minority population. Robinson, 86 F.4th at 589 (citing Milligan, 599 U.S. at 18). Under this precondition, a plaintiff is required to establish that the minority population forms a sufficient majority in a district to have the potential to elect a minority-preferred candidate if there is political cohesion. Emison, 507 U.S. at 40. To do so, the plaintiff “must show by a preponderance of the evidence that the minority population in the potential election district is greater than 50 percent.” Strickland, 556 U.S. at 19-20. “This percentage is analyzed in terms of the [BVAP] because only eligible voters can affect the Gingles analysis.” Robinson, 86 F.4th at 590. If the BVAP is greater than 50 percent, it must also be sufficiently compact such that a reasonably configured majority-minority district can be drawn. See LULAC v. Perry, 548 U.S. at 433.
This first precondition is satisfied as follows:
Where an election district could be drawn in which minority voters form a majority but such a district is not drawn, or where a majority-minority district is cracked by assigning some voters elsewhere, then - assuming the other Gingles factors are also satisfied - denial of the opportunity to elect a candidate of choice is a present and discernible wrong that is not subject to the high degree of speculation and prediction attendant upon the analysis of crossover claims.Strickland, 556 U.S. at 18-19. It also “requires the possibility of creating more than the existing number of reasonably compact districts with a sufficiently large minority population to elect candidates of its choice.” LULAC v. Perry, 548 U.S. at 430 (quoting Johnson v. De Grandy, 512 U.S. 997, 1008 (1994)). Thus, it must be shown that a large, geographically compact minority population exists that could be part of an additional, reasonably configured majority-minority district that the State did not draw. See Milligan, 599 U.S. at 20.
The Defendants argue that after the Supreme Court's recent Milligan decision, there is another requirement - a precondition to the precondition as it were. They rely on Justice Kavanaugh's concurrence in that opinion to assert that a court must first find that the districts as legislatively drawn combined or divided the black population. Before quoting the part of the concurring opinion on which the Defendants rely, we address a part of the majority opinion they ignore. The majority opinion described the first precondition this way:
First, the “minority group must be sufficiently large and [geographically] compact to constitute a majority in a reasonably configured district.” Wisconsin Leg[.,] 595 U.S. [at 400] (per curiam) (citing Gingles, 478 U.S. at 46-51, 106 S.Ct. 2752). A district will be reasonably configured, our cases explain, if it comports with traditional districting criteria, such as being contiguous and reasonably compact.Milligan, 599 U.S. at 18 (first alteration in original) (citing Alabama Legis. Black Caucus, 575 U.S. at 272). Justice Kavanaugh joined the majority's analysis of the Gingles preconditions. Id. at 42 (Kavanaugh, J., concurring in part and concurring in the judgment). It is difficult to interpret anything else Justice Kavanaugh wrote as being an alteration of what he accepted as the majority's understanding of precondition one.
Instead of revising precondition one, his concurring opinion was responding to the State's argument that Gingles needed to be abandoned because it “inevitably requires a proportional number of majority-minority districts.” Id. at 42-43. That is not what Gingles demands, Justice Kavanaugh wrote, as that would require the combining of “geographically dispersed minority voters into unusually shaped districts, without concern for traditional districting criteria such as county, city, and town lines.” Id. at 43. “Gingles requires the creation of a majority-minority district only when, among other things, (i) a State's redistricting map cracks or packs a large and geographically compact minority population and (ii) a plaintiff's proposed alternative map and proposed majority-minority district are reasonably configured.” Id. (quotation marks and citations omitted).
We view this portion of Justice Kavanaugh's opinion as simply explaining the natural effect of satisfying precondition one. The effect is that if a reasonably configured majority-minority district can be formed that satisfies traditional redistricting principles and does not join “geographically dispersed minority voters into unusually shaped districts,” then minority voters have been cracked from that minority district to form majority districts. Id. Justice Kavanaugh also agreed that analyzing the preconditions “requires in certain circumstances that courts account for the race of voters so as to prevent the cracking or packing . . . of large and geographically compact minority populations.” Id. at 44.
We conclude, then, that Justice Kavanaugh's concurring opinion cannot be read as his abandonment of almost 40 years of jurisprudence applying Gingles. We still, however, need to understand what compactness of a district means. We see it as an imprecise concept, but we know we are to consider traditional districting principles like communities of interest and maintaining traditional boundaries. LULAC v. Perry, 548 U.S. at 433. “Communities of interest vary between states, generally defined by the given state's districting guidelines.” Robinson, 86 F.4th at 590; see Milligan, 599 U.S. at 20-21. The districting guidelines applicable here require communities that share a common interest, are likely to have similar legislative concerns, and might benefit from cohesive representation in the state legislature. PTX-001, 20.
In Milligan, the Supreme Court concluded that “proposed maps [which] split the same number of county lines as (or even fewer county lines than) the State's map” and contained no “tentacles, appendages, bizarre shapes, or any other obvious irregularities” can strongly support the first Gingles precondition. 599 U.S. at 20 (emphasis in original) (quoting Singleton v. Merrill, 582 F.Supp.3d 924, 1011 (N. D. Ala. 2022)). We will look for that in the Plaintiffs' Illustrative Plans.
The Court further identified that equal populations, contiguity, and respect for existing political subdivisions like counties, cities, and towns would allow illustrative maps to satisfy traditional districting criteria and strongly suggest black voters could constitute a majority in a reasonably configured district. Id. In analyzing communities of interest, the Court concluded that even urban and rural communities can be configured into reasonably compact districts if they share cultural, economic, social, and educational ties despite the geographical distance. See Id. at 19-21; LULAC v. Perry, 548 U.S. at 434-35.
Race can also be considered in drawing illustrative districts to satisfy the first Gingles precondition. See Milligan, 599 U.S. at 31. Certainly, to demonstrate that reasonably configured, majority-minority districts could be drawn, race must be considered, but race cannot be “the predominant factor in drawing district lines.” Id. Here, William S. Cooper testified as an expert for the Plaintiffs in redistricting relevant to the first Gingles precondition, and he drew the Illustrative Plans the Plaintiffs presented as part of his overall analysis. Cooper testified that he was aware of and considered race as required by Gingles when drawing his maps, but it did not predominate in his analysis. Trial Tr. 109:2-5; 152:20-21. We will explore the specifics of Cooper's testimony when we examine each of the illustrative districts. For now, it suffices to say that we have found as to some of the districts, but not all, race predominated and traditional factors were not followed.
Besides arguing that race predominated, the Defendants argue that Cooper did not take in account the traditional redistricting principle of “political performance.” As we mentioned earlier, the best explanation of that phrase offered by the Defendants is that it refers to length of service and incumbency. To the extent this can be seen as, among other interests, protecting incumbent-constituent relationships and maintaining hard-earned legislative expertise, that is a valid state interest but should not be deferred to at the cost of “constitutional norms.” League of United Latin Am. Citizens, Council No. 4434 v. Clements, 986 F.2d 728, 763 (5th Cir.), rev'd en banc, 999 F.2d 831 (5th Cir. 1993) (quoting White v. Weiser, 412 U.S. 783, 791 (1973)).
Much of the law on incumbent protection is in the context of equal protection claims. For example, the Supreme Court has held that “avoiding contests between incumbent[s]” is a legitimate state goal. Karcher v. Daggett, 462 U.S. 725, 740 (1983). Similarly, we conclude that political performance, a concept that gives weight to the length of service of incumbents and their relationships with constituents, is a valid consideration under the first Gingles precondition.
Courts must consider each illustrative district independently to “determine if the illustrative districts have similar needs and interests beyond race.” Robinson, 86 F.4th at 590; see LULAC v. Clements, 999 F.2d at 877-94 (analyzing each district in detail). An illustrative map proposing overall compactness improvements does not necessarily meet Gingles one as to each illustrative district within the map. See LULAC v. Clements, 999 F.2d at 877-94. Thus, the key consideration here is that if one of the Illustrative Plans presented by the Plaintiffs identifies a reasonably configured, compact majority-minority district that respects traditional redistricting principles, the first Gingles precondition is met. See Milligan, 599 U.S. at 20. We discuss each of the Plaintiffs' Illustrative Plans and their respective districts to determine whether each satisfies the first Gingles precondition.
a. An Overview of the Plaintiffs' Illustrative Plans
As noted, the Plaintiffs presented the expert testimony of William Cooper, who was tendered and accepted as an expert in redistricting, demographics, and census data, to prove the first Gingles precondition. Trial Tr. 84:11-14, 85:23-86:1. Cooper works with different organizations and states as a private consultant in redistricting work. He has prepared redistricting maps in approximately 750 jurisdictions in 45 states and has been qualified as a redistricting and demographics expert in over 50 voting-rights cases in 20 states since 1987. Trial Tr. 78:21-79:5, 80:10-13, 81:24- 82:5; See PTX-001, 81-91. In Mississippi, Cooper has over 30 years of experience in voting cases and has served as a redistricting and demographics expert in multiple statewide cases. Trial Tr. 82:15-18, 83:10-84:10.
In preparation for this trial, Cooper developed two Illustrative Legislative Plans - the Illustrative Senate Plan and the Illustrative House Plan - to assess whether Mississippi's black population is sufficiently geographically compact to create additional majority-minority districts. Trial Tr. 86:5-17. We found Cooper's extensive experience, particularly in Mississippi cases, qualified him to testify as an expert in redistricting and demographics relevant to this case.
We first offer the two statewide Illustrative Plans. Each shows the location of the Plaintiffs' illustrative additional majority-minority districts: four senate districts and three house districts.
This is the map showing all the illustrative senate districts:
(Image Omitted)
The Enacted Senate Plan has 50.36 percent of black voters living in black-majority districts and 84.33 percent of white voters living in white-majority districts. PTX-001, 50. In Cooper's Illustrative Senate Plan, 58.39 percent of black voters live in black-majority districts and 75.24 percent of white voters live in white-majority districts. Id. By modifying 41 of the 52 enacted senate districts, the Illustrative Senate Plan reduces the alleged representation gap found in the Enacted Senate Plan by 17.13 percent. Trial Tr. 98:1-12; PTX-001, 28, 67-68.
This is the map showing all the illustrative house districts.
(Image Omitted)
The Enacted House Plan has 62.38 percent of black voters living in black-majority districts and 82.92 percent of white voters living in white- majority districts. PTX-001, 74. In Cooper's Illustrative House Plan, 64.78 percent of black voters live in black-majority districts and 80.12 percent of white voters live in white-majority districts. Id. The Illustrative House Plan modifies 33 of the 122 enacted house districts and improves the alleged representation disparity found in the Enacted House Plan by 5.19 percent. Id.
We next consider the testimony explaining the drawing of these maps.
To develop his Illustrative Plans, Cooper's initial analysis began with black-population statistics in Mississippi. Cooper used several data points, including population and geographic data from the 1990 to 2020 Censuses, socioeconomic data published by the Census Bureau, Mississippi precinct boundaries, and incumbent-address information. PTX-001, 93-94. The specific Census population data set Cooper used is the complete population file designed by the Census Bureau for use by states in legislative redistricting. Id. Cooper then used the well-known redistricting software “Maptitude for Redistricting” to compile the data and draw his Illustrative Plans. Id. Maptitude displays various kinds of voter data and precinct lines and permits the map-drawer to see precincts shaded in different colors based on their BVAP. Doc [201], 78, 109-110.
“This software is deployed by many local and state governing bodies across the country for redistricting and other types of demographic analysis.” PTX-1, 92.
According to the 2020 Census data, the non-Hispanic White population comprises 55.35 percent of the total population in Mississippi. See PTX-001, 9. African Americans are the next largest racial/ethnic category, representing 37.94 percent of the total population in 2020, which is the highest proportion of any state. Id. Mississippi's overall population grew by 116,621 persons from 2000 to 2020. Doc [220], 16-17. The African American population alone grew by 81,905 persons, making it the largest element of Mississippi's population growth. PTX-001, 10; Trial Tr. 87:5-13. Mississippi's BVAP also increased from 33.29 percent to 36.14 percent during that same period. PTX-001, 6, 10-11. In contrast, Mississippi's non-Hispanic White population fell by 88,831, and the non-Hispanic White VAP dropped from 64.16 percent to 57.76 percent from 2000 to 2020. PTX-001, 6, 10-11; Trial Tr. 87:5-13.
Cooper used Mississippi's Planning and Development Districts (“PDDs”) to evaluate the state's population change at the regional level. Trial Tr. 90:23-24. These PDDs were designed to “provide a consistent geographic base for the coordination of Federal, State, and local development programs.” PTX-020, 1 (Exec. Order). They were also “organized with boundaries which represent natural, social, and economic relationships.” Id. at 2. Cooper used these PDDs as “a way to organize [Mississippi] in[to] regions . . . that actually matter today.” Trial Tr. 90:23-24. However, the PDD boundaries have remained unchanged since at least 1971. See PTX-020. That by itself suggests they are poorly designed tools for the purposes of evaluating legislative districts.
The Plaintiffs filed a pretrial motion seeking judicial notice of facts related to PDDs, and we deferred ruling. See Pls.' Mot. [196]. Under Federal Rule of Evidence 201(b), a “court may judicially notice a fact that is not subject to reasonable dispute because it: (1) is generally known within the trial court's territorial jurisdiction; or (2) can be accurately and readily determined from sources whose accuracy cannot reasonably be questioned.” Some facts the Plaintiffs mentioned in their motion were introduced into evidence, thus mooting those issues. For example, former Mississippi Governor John Bell Williams's 1971 executive order was introduced as PTX-020. See Mem. [197] at 6. Cooper then testified about the general purpose of PDDs. See, e.g., Trial Tr. at 90-91. Beyond what was admitted into evidence, the Plaintiffs have not established that other potentially relevant facts satisfy Rule 201. For example, they have not shown that we should take judicial notice of facts found on a trade association's website. See Pls.' Mem. [197] at 6. Even if we were to take judicial notice of the remaining facts, it would not change the conclusion that PDDs are ill-suited for legislative districting.
Cooper combined the 2020 Census data and population information provided by the PDDs to determine whether additional majority-minority districts could be drawn in his Illustrative Plans. PTX-001, 16-18. Cooper focused primarily on PDD regions with substantial black populations that experienced double-digit black-population growth or double-digit white-population decline between 2000 and 2020. PTX-001, 18-21; Trial Tr. 156:11-157:9, 159:18-160:3. These were the areas in which Cooper “felt it was likely . . . [to] develop additional majority [b]lack districts.” Trial Tr. 160:4-8.
Cooper's analysis of Mississippi's PDDs showed the black population growth at the regional level was concentrated in four PDD regions: Central Mississippi, North Delta, Southern Mississippi, and Three Rivers. Trial Tr. 92:1-11; PTX-001, 17. The net black-population growth in these four regions was 120,399 persons between 2000 and 2020, and the white population loss was 7,636 persons. Trial Tr. 92:1-11; PTX-001, 17. Cooper testified the net black-population growth in these four regions equates to the drawing of two 100 percent black senate districts and about five 100 percent house districts, thus suggesting “it would be very easy to draw additional majority-[b]lack districts in the state of Mississippi in these specific areas.” Trial Tr. 95:3-8. By focusing on these areas, Cooper further asserted his Illustrative Plans “prove superior or equal to the 2022 Legislative Plans across almost every conceivable race-neutral quantitative measure of community of interest.” PTX-001, 21.
While we credit Cooper's overall analysis and drawing of the Illustrative Plans as viable methods for us to consider when determining compactness, we find the evidence does not support his use of PDDs as communities of interest. The PDDs were created about 60 years ago, PTX-0020, and they are “voluntary nonprofit corporations,” Trial Tr. 156:5-7, that have differing priorities and common interests, DTX-016. Further, there was testimony that PDDs are not public civil divisions for which the Census Bureau reports. Trial Tr. 154:5-15. The Plaintiffs even concede PDDs do not help further analysis of the first Gingles precondition, stating the court “can eliminate the concept of [PDDs] and still come up with . . . similar” districts. See Trial Tr. 156:25-157:9, 158:2-5, 160:2-3. There is also no evidence the Mississippi Legislature considered PDDs in drawing the Enacted Plans. See Miss. Code Ann. § 5-3-101. We find the use of PDDs in Cooper's analysis to be unhelpful.
Of importance, though, is that the 2022 Enacted Plans contained the same number of majority-minority districts as the previously enacted plans despite the admitted black-population growth and white-population decline. PTX-001, 22-25, 45-46, 54-57, 68-70, 177-79, 244-46, 498-501, 590-93; JTX-002; JTX-004; JTX-005. Indeed, both the state senate and state house have only added one additional majority-minority district in their respective legislative plans since 2002. Trial Tr. 89:12-90:1. The additional majority-minority senate district resulted from a court order four years ago. See Thomas, 961 F.3d at 800-01. Trial Tr. 90:2-8.
Cooper testified that he developed the Illustrative Plans here by using traditional redistricting principles, including population equality, compactness, contiguity, communities of interest, traditional political boundaries, non-dilution of minority voting strength, and incumbent pairings. Trial Tr. 105:4-107:12. He further testified that he balanced the traditional redistricting principles such that none predominated over the other when drawing the Illustrative Plans. Trial Tr. 107:13-22, 109:2-6, 123:13-14. We credit this statement as to some of his illustrative districts, but not every illustrative district.
The Defendants challenged whether Cooper properly used these traditional redistricting principles. At trial, Cooper was asked to review the legislative redistricting process and the factors the Standing Joint Committee was required to consider - one person, one vote; contiguity of the districts; political performance; compliance with all state and federal laws such as Section 2 of the Voting Rights Act; compactness; and minimalization of county and precinct splits. JTX-010, at 8-14; Trial Tr. 168:20. The Standing Joint Committee examined the significant population shifts throughout Mississippi where major areas experienced population loss and indicated this necessitated the collapse of two districts to be moved to the areas with the largest increase in population. JTX-010, 11-14.
Cooper acknowledged that he neither examined the many competing interests the Mississippi Legislature examines when drawing maps nor considered political performance while drawing the Illustrative Plans. Trial Tr. 225:21-226:2. On the other hand, the Defendants failed to provide any detail as to how the failure to consider those interests impacted any of Cooper's illustrative districts. The Defendants' evidence was fairly brief videos of each committee chair explaining to his chamber what that committee had tried to do. General statements about political performance were made. Nothing in that evidence addresses how Cooper's failure or inability to consider political performance invalidated any of the illustrative districts. Even if “political performance” is a trump card, which it is not, the Defendants must at least place it face up on the table.
The Defendants offered the testimony of Dr. Thomas Brunell to prove that the Plaintiffs did not meet their burden under Gingles precondition one. See generally Trial Tr. 1242-1495. Dr. Brunell is a professor with a doctorate in political science at the University of Texas at Dallas, where he has been employed since 2005. DX-003, 22. He is an appointed member of the Census Scientific Advisory Committee for the Census Bureau where he provided guidance and assistance related to redistricting development procedures. Trial Tr. 1250:18-1251:13. Dr. Brunell was accepted at trial as an expert in redistricting, elections, the Voting Rights Act, and representation and statistics. Trial Tr. 1253:24-1254:2. Although the Defendants offered Dr. Brunell's testimony specifically to challenge Cooper's methods, Dr. Brunell agreed Cooper was not maximizing the number of majority-minority districts. Trial Tr. 1336:1-9.
Regarding compactness and split precincts, Dr. Brunell opined that there were only marginal differences between the compactness scores of the Enacted Plans and Cooper's Illustrative Plans. DX-003, 12-14. Cooper considered three mathematical compactness scores generated by the accepted Maptitude software and asserted that, on average, the Illustrative Senate Plan is slightly more compact than the Enacted Senate Plan and the compactness of the Illustrative and Enacted House Plans are comparable. Trial Tr. 106:3-13, 111:7-112:24. That is generally true.
The only geographical boundaries Mississippi law requires be followed “as nearly as possible” are county, municipal, and precinct lines. Miss. Code Ann. § 5-3-101. Both the Enacted Plans and Illustrative Plans satisfied this requirement. The Illustrative House Plan has the same number of split counties and modestly fewer split municipalities and precincts compared to the Enacted House Plan. DX-003, 15. The Illustrative Senate Plan splits fewer counties but has almost the same number of total county splits as the Enacted Senate Plan. Dr. Brunell created a chart to show the comparisons.
Split Counties
Total County Splits
2020 VTD Splits
Municipalities not Split
Total Municipal Splits
2022 House
&7
179
255
216
225
Cooper House
67
167
228
218
221
2022 Senate
43
58
41
244
128
Cooper Senate
34
52
38
253
no
Dr. Brunell's chart supports Cooper's statement that “[t]here's very little difference” between the Enacted Plans and the Illustrative Plans. Trial Tr. 112:22-23. The Illustrative Plans perform only a “little better[, n]ot a lot but a little,” better than the Enacted Plans, such that “[t]hey're almost the same.” Trial Tr. 112:10-11, 22-23.
As to incumbent pairings, Dr. Brunell noted the Enacted Plans have only one incumbent pairing in the senate and three pairings in the house. DX-003, 14. Separating incumbents into different districts was an important factor the Mississippi Legislature considered when drawing its plans. JTX-10, 12:3-15, 20:19-24; JTX-011, 8:5-10. That is a valid consideration for the legislature. See Vera, 517 U.S. at 963-64; League of United Latin Am. Citizens, Council No. 4434, 986 F.2d at 763. The Illustrative Plans, however, have more incumbent pairings than the Enacted Plans, despite incumbency protection being an important consideration. DX-003, 14. Cooper testified he was unable “to obtain complete information about all of the incumbents,” and it was sometimes impossible “to keep all incumbents in separate districts.” Trial Tr. 106:23-107:12. Nonetheless, the Enacted Plans are more favorable as to incumbency protection.
We found at trial that Dr. Brunell is qualified to testify as an expert in redistricting and demographics, but the following discussion of the individual illustrative districts will reveal some of his testimony that we do not credit.
b. Illustrative Senate District 2
(Image Omitted)
Illustrative S.D. 2 is an additional majority-minority district with a BVAP of 50.91 percent that Cooper asserts can be drawn in Tunica and DeSoto Counties. PTX-001, 29-34, 324. We reproduce the western-most part of the map. DeSoto County contains the fastest growing black population in Mississippi. Trial Tr. 130:2-3. Cooper combined some of the population from three enacted senate districts throughout DeSoto County, to create Illustrative S.D. 2 in addition to maintaining a current majority-black district. PTX-001, 29-34. Cooper divided the City of Horn Lake and its substantial black population once, keeping most of it in a single district, and eliminated the splitting of Tate and Panola Counties under the Enacted Senate Plan. PTX-001, 29-34; Trial Tr. 131:23-132:8. Cooper testified that he followed natural boundaries by tracking the Highway 61 transportation and economic corridor, allowed for population growth by including precincts with low BVAP, and kept the population deviation on the lower end of the required 5-percent deviation. Trial Tr. 131:14-20, 190:15-191:5.
Pamela Hamner also testified for the Plaintiffs to explain how Illustrative S.D. 2 respects communities of interests. Hamner has lived in DeSoto County for 25 years as a reporter covering northern Mississippi. She recently ran for political office, which caused her to contact voters and residents in the areas contained in Illustrative S.D. 2. Trial Tr. 704:10-14, 705:15-706:4, 710:1-4, 720:5-11. Hamner testified that she received some threats while campaigning in white communities, though few details were given. She testified that Highway 61 connects the communities and allows the residents to travel between towns in DeSoto County for church, healthcare, entertainment, and other activities. Trial Tr. 723:13-20. Hamner also explained in detail the economic, education, healthcare, and numerous other issues the residents are similarly concerned with. Trial Tr. 723:7-12, 714:14-715:6, 727:16-22.
In creating Enacted S.D. 2, the Mississippi Legislature effectively cracked the majority-black community in Horn Lake across three districts and split Horn Lake and the historically black town of Jago rather than keeping them together like in Illustrative S.D. 2. Trial Tr. 713:15-714:6, 716:5- 25, 718:7-20. Hamner testified this effectively took away the power of the black communities to seek representation when the Enacted Plans became law. Trial Tr. 714:7-13, 715:10-17. We credit that testimony.
The Defendants disputed that Cooper's methods followed traditional redistricting principles for this illustrative district. Defendants' expert Dr. Brunell opined that Cooper specifically targeted the majority-black precincts to include in Illustrative S.D. 2 while excluding the majority-white precincts to increase the number of black-majority districts. DX-003, 11; Trial Tr. 1336:1-9. On that point, we know that in preparing illustrative districts, the Supreme Court recognizes their purpose is to show what can be done, i.e., that a reasonably compact majority-minority district can be created that respects traditional districting principles. Milligan, 599 U.S. at 30. There “is a difference ‘between being aware of racial considerations and being motivated by them'” in preparing the maps. Id. (quoting Miller, 515 U.S. at 916). Cooper's identifying those majority-minority areas was a necessary part of the exercise.
This district is reasonably compact and satisfies traditional redistricting factors. Cooper testified that he was not maximizing the number of majority-minority districts, and Defendants' expert Dr. Brunell opined the same. We credit both witnesses' testimony as it applies to this illustrative district. Further, there was no evidence offered about how considerations of political performance affected the districting here. We thus find that Illustrative S.D. 2 satisfies the first Gingles precondition.
c. Illustrative Senate District 9
(Image Omitted)
Illustrative S.D. 9 is an additional majority-minority district with a BVAP of 50.95 percent that Cooper asserts can be drawn in Forrest and Lamar Counties and adjacent to the City of Hattiesburg. PTX-001, 38-41, 324. It includes Rawls Springs, Glendale, and substantial portions of Arnold Line and West Hattiesburg, and it is drawn primarily around Hattiesburg's municipal borders. See PTX-001, 394. Cooper testified this illustrative district better respects traditional redistricting principles than the Enacted Senate Plan. Trial Tr. 134:1-14; PTX-001 at 38-41, 247. The Enacted Senate Plan split Hattiesburg across four districts and went far north “to pick up pieces of Jones County and Laurel and then on into Jasper County.” PTX-001, 302. Illustrative S.D. 9 instead keeps the city almost entirely whole, follows municipal boundaries extending from Hattiesburg, avoids pairing an incumbent, and evenly includes and excludes various BVAP precincts. Trial Tr. 134:19-20, 134:23-135:10, 204:17-25, 205:7-15.
Lay testimony supports the reasonableness of keeping Hattiesburg together. Dr. Joseph Wesley, a Hattiesburg resident since 1977, explained that Hattiesburg is the “Hub City” for people from the surrounding areas near Illustrative S.D. 9 to come for education, culinary, and economic needs. Trial Tr. 665:7-9, 676:2-12. He testified Illustrative S.D. 9 better respects the geographic boundaries of highways that serve the surrounding areas and allows easier access to education and healthcare for the communities. Trial Tr. 679:4-684:9. The common histories and traditions for the black communities are also centered in Hattiesburg, where the communities are kept whole through Illustrative S.D. 9. Trial Tr. 676:13-677:8, 678:3-5. We credit his unrebutted testimony.
The Defendants presented evidence that the boundary lines of Illustrative S.D. 9 followed along the racial lines of larger concentrations of black populations and split two precincts between Enacted S.D. 9 and S.D. 42. PTX-001, 38-41, 324; Trial Tr. 205:7-15. The Defendants asserted there was no race-neutral explanation for Cooper's redrawing of Illustrative S.D. 9 in this way because altering either of the split precincts to make them whole would undermine the 50.95 percent BVAP of Illustrative S.D. 9 and preclude Cooper's racial objective. DX-003, 46; Doc [219], 36. However, the Defendants' overlay of Illustrative S.D. 9's borders on a racially shaded precinct map does not demonstrate such clear-cut cherry-picking between majority-black and majority-white precincts, as several majority-white precincts just outside Hattiesburg are also within Illustrative S.D. 9's borders. See DX-003, 046.
Here, too, we find that Cooper has created a reasonably compact district that follows traditional redistricting principles. His purpose, as Gingles precondition one requires, was to identify whether a reasonably compact majority-minority district could be formed that respected traditional redistricting factors. We find that the City of Hattiesburg itself is less split in the illustrative district. Although Illustrative S.D. 9 does exclude what we accept are some majority-white areas, we find the design still leaves the district reasonably compact. We find this district satisfies Gingles precondition one.
d. Illustrative Senate District 17
(Image Omitted)
Illustrative S.D. 17 is an additional majority-minority district with a BVAP of 53.54 percent that Cooper asserts can be drawn in the Golden Triangle and Three Rivers areas. PTX-001, 323-25. S.D. 17 had an original BVAP of 29.48 percent in the Enacted Senate Plan. Id.; JTX-049, 4. To create Illustrative S.D. 17, Cooper took portions of the black populations from Enacted S.D. 6, 7, 8, and 16. PTX-001, 249-50, 386; Trial Tr. 136:7-10. Cooper testified this better respected black communities along Highway 45 that share socioeconomic and historical interests, followed whole precincts, and tracked natural boundaries. Trial Tr. 136:7-21, 199:24-200:14, 223:25- 224:3.
Mamie Cunningham, a retired schoolteacher and lifelong resident of Chickasaw County, testified as to the common interests that Illustrative S.D. 17 brings together. Trial Tr. 233:2-234:2, 244:17-245:1. Residents in the area share socioeconomic, civic, economic, educational, and other interests across county lines, and the municipalities are connected by natural boundaries Highway 45 and Alternate Highway 45, which are frequently traveled by residents to access retail outlets and services. Trial Tr. 245:2- 246:13, 247:14-248:2. Although she resides in Chickasaw County, Cunningham herself worked in neighboring Monroe and Clay Counties, all of which are within Illustrative S.D. 17. Trial Tr. 244:13-245:13. We credit her testimony in general, but we do not credit her reliance on the fact that a particular highway goes through the entire district to say a community of interest exists.
According to the Defendants, however, Cooper “dismantles” Enacted S.D. 7 and S.D. 16 by reducing their BVAPs from 40.08 percent and 63.06 percent to 17.83 percent and 53.54 percent, respectively, to draw Illustrative S.D. 17. Doc [219], 27; PTX-001, 324; JTX-049, 3-4. They assert Cooper crossed PDD boundaries, violating his own redistricting criteria, and used only public-school-district athletics in his consideration for Illustrative S.D. 17 because he did not think black children would be attending private schools. Trial Tr. 197:1-6, 200:15-204:2. We have already found that the PDDs are not useful in the analysis, and Cooper's deviation from them helps support our earlier finding. On cross, Cooper explained the precinct splits he created protected incumbents and better preserved other precinct lines, resulting in fewer splits than the Enacted Plans. Trial Tr. 196:25-198:3.
The Defendants maintain that it is clear Illustrative S.D. 17 was created for no reason other than race. They assert Cooper had to dissolve Enacted S.D. 7, which elected a minority-preferred white Democratic candidate in 2023, and the dismantling of an existing minority-performing district is not required under Section 2. Doc [219], 34-35.
This illustrative district splits three major municipalities in the area, crossing multiple relevant boundaries. It splits Tupelo and captures only black-majority precincts but excludes white-majority precincts. DX-003, 42. It further splits West Point and Amory by crossing PDD and geographic boundaries that split communities of interest. See PTX-001 at 338; Trial Tr. 136, 197-198. These facts support that Cooper had a racial objective and that race predominated in his drawing of this district to achieve said objective. Although he gave his opinion that there was a community of interest because Illustrative S.D. 17 follows Highway 45 and the Tennessee-Tombigbee Waterway from north to south, we do not accept that following those two transportation corridors supports that the communities they encounter share interests. Rather, the evidence indicates Cooper combined parts of distinct communities in an attempt to force one community of interest to exist.
We find that Cooper's claimed considerations of communities of interest and lack of a racial objective are not credible. Illustrative S.D. 17 does not satisfy the first Gingles precondition.
e. Illustrative Senate District 35
(Image Omitted)
Illustrative S.D. 35 is an additional majority-minority district with a BVAP of 52.12 percent that Cooper asserts can be drawn in Copiah, Simpson, Lincoln, Lawrence, and Jefferson Davis Counties. PTX-001, 324. Enacted S.D. 35 had an original BVAP of 39.38 percent in the Enacted Senate Plan. Id.; JTX-049, 4. To achieve his illustrative district, Cooper splits Lincoln County, crossing the county boundary to cut out a portion of the City of Brookhaven and to incorporate the area between it and the City of Wesson, which has a “significant Black population.” Trial Tr. 206:15-18, 139:4-7. Cooper testified Illustrative S.D. 35 also keeps Copiah and Jefferson Davis Counties whole; splits only one precinct to keep the City of D'Lo whole; respects the geographical, transportation, and educational connections along Highway 51; and respects high-school-sports leagues in the counties. Trial Tr. 139:1-14, 224:9-225:10; PTX-001, 341, 367.
In redrawing Illustrative S.D. 35, Cooper split counties and municipalities to achieve his apparent racially predominant objective. See Trial Tr. 206:24-207:22; DX-003, 44; PTX-001, 398. By running a narrow extension south into a portion of Brookhaven, Cooper splits the city between two districts and includes only predominantly black precincts within Illustrative S.D. 35 to reach the desired BVAP. See Trial Tr. 206:24-207:22; DX-003, 44; PTX-001, 398. This extension into the black precincts of Brookhaven, not into the whole city, certainly falls into the category of a “tentacle” that shows the illustrative district does not have a reasonably compact majority-minority voting-age population.
In defending Illustrative S.D. 35, Cooper testified that Highway 51 and Interstate Highway 55 run north from Brookhaven through Hazelhurst and to Crystal Springs as common boundaries and that a significant number of Mississippians commute to work via these highways. See Trial Tr. 138-39, 224-25; PTX-001, 398. Nonetheless, the tentacle into Brookhaven excludes Interstate 55 and hugs only the east side of Highway 51 to capture the higher BVAP population of the area. See DX-003, 44; PTX-001, 398.
Plaintiffs' witness Ashley Wilson, a lifelong resident of the City of Crystal Springs, testified there were significant community ties between Brookhaven and Copiah County. See Trial Tr. 873-85. Specifically, many people in Copiah County and Brookhaven share economic, shopping, work, hospital, and travel interests that are all connected along Highway 51 and Interstate 55. Id. These ties, however, are not enough to overcome the significant county and municipality splits in Illustrative S.D. 35, nor do they explain how one tentacle of Brookhaven that is majority black would have any more in common with Copiah County municipalities than other precincts.
Illustrative S.D. 35 does not satisfy the first Gingles precondition.
f. Illustrative House District 22
(Image Omitted)
Illustrative HD 22 is an additional majority-minority district with a BVAP of 55.41 percent that Cooper asserts can be drawn in Chickasaw and Monroe Counties by splitting small areas from the Cities of Aberdeen and Houston. PTX-001, 716, 897. HD 22 had an initial BVAP of 29.86 percent in the Enacted House Plan. PTX-001, 716; JTX-051, 4. Cooper testified Illustrative HD 22 is more compact than the Enacted House Plan by containing only two counties and following Highway 45 rather than splitting predominantly black communities across three districts like in the Enacted House Plan. Trial Tr. 142:5-143:8; PTX-001, 61-64. He explained Illustrative HD 22 encompasses whole precincts and follows natural boundaries such as waterways and county borders. Trial Tr. 144:2-6. The Enacted House Plan instead “crack[s the b]lack population in the midsection of Chickasaw and Monroe Counties” three ways and connects them via a narrow land bridge. Trial Tr. 142:21-143:8.
Mamie Cunningham again testified about the communities of interest for Illustrative HD 22, and like for Illustrative S.D. 17, residents in the Illustrative HD 22 area share socioeconomic, civic, economic, educational, and other similar interests across county lines, and the municipalities are connected by the frequently traveled natural boundaries Highway 45 and Alternate Highway 45. Trial Tr. 245:2-246:13, 247:14-248:2. We credit her testimony.
The Defendants do not so much challenge the illustrative district as they explain why a different district was drawn. They rely on the state house committee chairman's summary statement explaining the plan and how there was population loss in the state. JTX-010, 6. The Defendants interpret the statement as indicating that there was a population loss in this specific district, but that is not what the chairman said. Regardless, all such a statement means is that a district that lost population must add population to comply with the required population-deviation standards, nothing more.
The Defendants also assert that Enacted HD 22 was designed to protect an incumbent. That fact does not affect the legitimacy of the illustrative district. Further, the Defendants argue Illustrative HD 22 is drawn along racial lines such that majority-black population precincts are included and majority-white precincts are excluded solely to achieve a racial objective. DX-003, 50. We explained before that so long as race is not the predominant factor, the Supreme Court has approved the consideration of race to show that a reasonably compact majority-minority district could be drawn. Milligan, 599 U.S. at 30-31. We distinguish our earlier discussion of Illustrative S.D. 17, where we found that Cooper allowed race to predominate when he split the black populations from three different cities, and the Plaintiffs offered little credible evidence that the resulting illustrative district joined communities of interest. There is no significant splitting of cities in Illustrative HD 22, and we find the evidence credible that the resulting district shares relevant interests.
Illustrative HD 22 was not drawn with race as the predominant factor, and it satisfies traditional redistricting criteria. We find this district satisfies the first Gingles precondition.
g. Illustrative House District 56
(Image Omitted)
Illustrative HD 56 is an additional majority-minority district with a BVAP of 58.99 percent that Cooper asserts can be drawn in Hinds and Madison Counties by splitting the City of Clinton. PTX-001, 717; Trial Tr. 143:724, 21513-16. Enacted HD 56 had an original BVAP of 22.97 percent in the Enacted House Plan. PTX-001, 716; JTX-051, 6. Cooper testified Illustrative HD 56 is “an extremely compact district” anchored in Clinton that does not stretch into Madison County like the Enacted House Plan and is only 15 miles long. Trial Tr. 144:11-24, 217:2-4. Illustrative HD 56 instead tracks the Interstate 20 border and contains whole precincts, while removing allegedly unnecessary county splits. Trial Tr. 145:14-23.
While this illustrative district is entirely contained in Hinds County and the Jackson municipal area, it does not appear Cooper gave the needed consideration as to whether there is an actual community of interest between Clinton and west Jackson. See PTX-001, 821. Cooper identified that because “[t]hey're right next door to one another,” “it's an urban area [where] there could be all sorts of common side relating roads and highways,” and “they're so close to one another,” Illustrative HD 56 could contain a potential community of interest. Trial Tr. 216:8-18. He could not identify any of the “other factors that [he] considered in determining whether or not there's a community of interest.” See Trial Tr. 216:21-217:4. Cooper only discussed Clinton's proximity to west Jackson and its “significant increase in the Black population.” Trial Tr. 217:14-15; see Id. This does not explain why the parts of Clinton included in Illustrative HD 56 have more in common with west Jackson than the rest of the city other than race.
Sharon Moman, a real-estate broker in and lifelong resident of Clinton and Jackson, testified about the strong connections between the two areas that make Illustrative HD 56 a community of interest. Trial Tr. 644:4-10, 649:17-651:2. Though Moman identified various connections, she did not testify that the specific portion of west Jackson that was included in the illustrative district shared such interests with Clinton. Instead, most of her testimony centered on Jackson amenities that were not in the part of Jackson included in the illustrative district. While Moman credibly answered the questions she was asked, her testimony does not establish a community of interest across Illustrative HD 56.
In addition, the Defendants argue that political performance justified Enacted HD 56 and undermined the illustrative district. We would describe the argument as being that the illustrative district did not follow traditional redistricting criteria because it did not take political performance into account. Here, the criterion is better understood as being incumbent protection. Though counsel's argument to the court mentioned the connection, no evidence was introduced that Speaker of the House Philip Gunn represented the enacted district. Trial Tr. 214:17-24. Nonetheless, we can take judicial notice that Speaker Gunn represented that district, but we must also take notice that he did not run for re-election. Nothing in the record, or that we can take judicial notice of, indicates whether it was already known by the Mississippi Legislature that he would not run again when the redistricting maps were adopted. In light of that uncertainty, we give only slight weight to the argument that the illustrative district did not follow the traditional redistricting criteria of political performance.
Considering all the evidence, we find that the Plaintiffs did not prove that Illustrative HD 56 was reasonably configured using traditional redistricting factors, particularly as to a community of interest. Thus, we find Illustrative HD 56 does not satisfy the first Gingles precondition.
h. Illustrative House District 84
(Image Omitted)
Illustrative HD 84 is an additional majority-minority district with a BVAP of 53.05 percent that Cooper asserts can be drawn in Newton, Jasper, and Clark Counties by splitting the City of Quitman. PTX-001, 717. The Enacted House Plan left Quitman whole and had a BVAP of 37.28 percent. Id.; JTX-051, 6. Cooper's Illustrative HD 84 is underpopulated by 2.76 percent while Illustrative HD 81, located east of Illustrative HD 84, is overpopulated by 4.67 percent. See DX-003, 50. Cooper nonetheless testified that Illustrative HD 84 reduces the number of county splits from eight to two, keeps the City of Newton mostly whole, and avoids splitting Newton's only majority-black precinct, all while the Enacted House Plan splits multiple counties and multiple precincts. PTX-001, 857-62, 870, 881- 91; see Trial Tr. 697:1-16.
Deacon Kenneth Harris and Terry Rogers both testified as to the community of interest found within Illustrative HD 84. Deacon Harris is a lifelong Newton County resident who previously served as a Newton County Supervisor and was a member of the East Central Planning and Development Board that worked on the redistricting of the district. Trial Tr. 685:17- 687:16. Rogers is a nineteen-year-old lifelong City of Quitman resident and current college student who ran for Mississippi Commissioner of Agriculture and Commerce. Trial Tr. 933:11-19, 936:17-937:18. Both Harris and Rogers credibly testified as to the common, rural, low-income nature of the communities combined in the illustrative district and how these communities share traditions, festivals, retail and church venues, and sports rivalries. Trial Tr. 691:3-693:21, 938:23-940:16. They also explained the employment and healthcare opportunities shared among Newton, Jasper, and Clark County residents. Id. According to Harris and Rogers, Illustrative HD 84 does a better job maintaining communities than the Enacted House Plan and “binds” the area “together instead of splitting it like it is now.” Id.; Trial Tr. 693:20-21.
However, this testimony does not provide enough to overcome the significance of splitting three counties and the City of Quitman. Neighboring-county residents attending major festivals in the area; preferring to attend church, shop, and work in Newton County; and being served by Newton County's one hospital speaks to their geographic proximity but is not enough to support that this illustrative district comprises a distinct community of interest. See Trial Tr. 690:16-693:20; 938:20- 942:24. That the one hospital does not appear to be within Illustrative HD 84 and an identified community of interest - Quitman - is split undermines the assertion that this illustrative district contains a community of interest. See Trial Tr. 693:9-10; PTX-001, 694. Illustrative HD 84 is thus not reasonably drawn and does not satisfy Gingles one.
Under the first Gingles precondition, the Plaintiffs were required to prove the geographical compactness and numerosity of Mississippi's minority population. Robinson, 86 F.4th at 589. The evidence presented must prove Mississippi's minority population in a potential election district is greater than 50 percent and is compact enough to create another black-majority district that the State did not draw, and the potential district must respect traditional districting criteria. See Strickland, 556 U.S. at 19-20; Milligan, 599 U.S. at 20.
We find that three of the illustrative senate and house districts reflect minority groups that are sufficiently large and geographically compact enough to constitute a majority in a reasonably configured district. Milligan, 599 U.S. at 18. These districts are Illustrative S.D. 2, S.D. 9, and HD 22.
The remaining four illustrative districts - Illustrative S.D. 17, S.D. 35, HD 56, and HD 84 - do not satisfy this Gingles precondition.
2. Gingles Preconditions Two and Three
The Supreme Court requires that the evidence allow a court to make findings as to all three preconditions as they specifically relate to an illustrative district. “Those three showings, [the Court] explained, are needed to establish that ‘the minority [group] has the potential to elect a representative of its own choice' in a possible district, but that racially polarized voting prevents it from doing so in the district as actually drawn because it is ‘submerg[ed] in a larger white voting population.'” Harris, 581 U.S. at 302 (second and third alteration in original) (quoting Emison, 507 U.S. at 40). Thus, once an acceptable illustrative district for purposes of precondition one has been identified, the Plaintiffs must establish that the second and third preconditions are satisfied as to the enacted districts in the geographical area of the illustrative district. In addition, evidence covering other geographical areas, if it can be shown those areas share relevant characteristics, could be shown to be relevant.
The second Gingles precondition concerns the voting behavior of black voters. See Gingles, 478 U.S. at 51, 56. It asks whether a significant number of black voters “usually vote for the same candidates,” id. at 56, such that they would elect their representative of choice in a majority-minority district, see Milligan, 599 U.S. at 18-19.
The third Gingles precondition focuses on the electoral outcomes resulting from racially polarized behavior. See Gingles, 478 U.S. at 51, 56. Concentrating on racially polarized voting, “the plausibility that the challenged legislative districting thwarts minority voting on account of race” must be established. Robinson, 86 F.4th at 595. In other words, a plaintiff must provide proof that “whites vote sufficiently as a bloc usually to defeat the minority's preferred candidates.” Gingles, 478 U.S. at 56. “Thus, the question whether a given district experiences legally significant racially polarized voting requires discrete inquiries into minority and white voting practices.” Id.
The second and third Gingles preconditions are often analyzed together. Emison, 507 U.S. at 40; see Milligan, 599 U.S. at 22. Courts determine first if the black voters are politically cohesive. Milligan, 599 U.S. at 18-19. Then, “[t]he question is not whether white bloc voting is present, but whether such bloc voting in a given district amounts to legally significant racially polarized voting.” Robinson, 86 F.4th at 595; see also LULAC v. Clements, 999 F.2d at 850. The Supreme Court recognizes a difference between legally significant and statistically significant racially polarized voting. See Gingles, 478 U.S. at 53, 55. Statistics may be used to determine the voting percentages and support for candidates in a given election, but what amounts to statistical importance may not rise to legal significance given the relevant facts. See Id. at 79 (clarifying the determination of “whether the political process is equally open to minority voters” is “peculiarly dependent upon the facts of each case” (quoting Rogers v. Lodge, 458 U.S. 613, 621)); see also LULAC v. Clements, 999 F.2d at 850-51.
Thus, “[t]he proper question to ask is this: If the [S]tate's districting plan takes effect, will the voting behavior of the white majority cause the relevant minority group's preferred candidate usually to be defeated?” Robinson, 86 F.4th at 597 (quotation marks and citation omitted). In other words, do cohesion among the minority group and bloc voting among the majority population both exist. We now examine the evidence regarding the second and third preconditions as they relate to the enacted districts in the areas where we found the three illustrative districts satisfied precondition one.
The Plaintiffs presented the expert testimony of Dr. Lisa Handley as their evidence of the second and third Gingles preconditions. Dr. Handley received her PhD in political science. She then started her own company where she provides redistricting expertise to various districts throughout the United States. Trial Tr. 260:20-25. She has aided state and local jurisdictions, the Department of Justice, and independent-redistricting and civil-rights organizations by providing racial bloc-voting analysis and expertise. Trial Tr. 255:18-21, 261:1-7. With over 35 years of quantitative- voting-analysis experience, she has testified as an expert in numerous redistricting cases, including in Louisiana, Georgia, and Texas, where she conducted analyses of voting patterns by race and redistricting plans to determine whether the black population is politically cohesive and the voting is polarized. PTX-004, 2-3, 70; Trial Tr. 255:18-21, 257:23-260:16. At trial, we accepted Dr. Handley as an expert in racially polarized voting and the statistical analysis of minority vote dilution and redistricting. Trial Tr. 262:15-22.
To challenge Dr. Handley's analyses, the Defendants presented expert testimony from Rice University professor Dr. John Alford. DX-001, 2. Dr. Alford has been teaching political science at Rice for over 35 years and is well-versed in redistricting, elections, voting behavior, and statistical methods. Id. Dr. Alford has served as an expert witness in numerous redistricting cases, including in Louisiana, Georgia, and Texas, where he presented statistical methodology and analyses related to racially polarized voting pursuant to the Gingles second and third preconditions. Trial Tr. 1410:22-1411:23; DX-001, 30-31. At trial, we accepted Dr. Alford as an expert on assessing the second and third Gingles preconditions and Senate Factor 2. Trial Tr. 1413:12-19.
Dr. Handley analyzed racial voting patterns in the county clusters where the seven illustrative districts were drawn to evaluate the extent of racially polarized voting. She focused on areas where the Illustrative and Enacted Plans overlapped. PTX-004, 6-8; Trial Tr. 262:24-263:7. She classified the senate districts as Area One - North West and North Central; Area Two - Greater Golden Triangle; Area Three - South Central; and Area Four - South East. The house districts were classified as Area Five - Western Jackson; Area Six - Golden Triangle; and Area Seven - East Central. PTX-004, 7-8 (Table One). The two illustrative senate districts we concluded satisfy the first Gingles precondition, Illustrative S.D. 2 and S.D. 9, are in Areas One and Four, and the valid illustrative house district, Illustrative HD 22, is in Area Two, with some overlap in Area Six. See id.
Dr. Handley utilized homogeneous precinct analysis, ecological regression, and ecological inference, three well-accepted statistical methods employed in Gingles analyses. PTX-004, 3-5; Trial Tr. 270:6-12. These three methods calculate the percentage of black voters and white voters who voted for specific candidates in certain elections. Trial Tr. 273:25-274:7. Dr. Handley and Dr. Alford agreed that, of the three methods used, the ecological inference (“EI RxC”) method is the most accurate and reliable for determining credible intervals of racial bloc voting. Trial Tr. 270:13-273:11, 1425:2-9, 1426:8-12. We accept this characterization. Further, EI RxC was the principal method Dr. Handley employed when analyzing the recent statewide general elections, state legislative general elections, statewide Democratic primaries, and nonpartisan judicial contests in the seven areas of interest. PTX-004, 8-11; Trial Tr. 270:13-273:11.
Dr. Handley primarily focused on contests that included both black and white candidates because courts have found these biracial elections to be more probative than contests with only white candidates for a Gingles polarized-voting analysis. Trial Tr. 274:22-275:6. For a comparison, she also analyzed several elections that did not include black candidates. See PTX-004, 9. Dr. Handley found that all seven areas she examined demonstrated consistently high levels of racially polarized voting, with black voters cohesively supporting their preferred candidates and white voters cohesively bloc-voting against black-preferred candidates. Though we must be alert to how the evidence and Dr. Handley's findings relate to the areas encompassing the three illustrative districts that satisfy the first Gingles precondition, we need not disentangle all of Dr. Handley's analyses.
We examine significant parts of Dr. Handley's evidence in what follows, but our findings as to the second and third preconditions rely on evidence relevant to the enacted districts that are in the geographical areas where the illustrative districts that we already found satisfied precondition one are located. As the Fifth Circuit stated, “plaintiffs must show that such bloc voting would be present in the challenged districting plan. And that conclusion must be true for voters in a particular location.” Robinson v. Ardoin, 37 F.4th 208, 224 (5th Cir. 2022) (emphasis in original) (citations omitted).
a. State Legislative Elections
Dr. Handley analyzed 19 biracial state legislative elections in each of the seven illustrative district areas under the 2012 enacted district boundaries. PTX-004, 9-10, 12. Dr. Handley classified these state legislative contests as endogenous elections, meaning the elections are for the same offices as the ones involved in the litigation. Trial Tr. 1447:2-16. She testified that endogenous elections are the most probative when analyzing the second and third preconditions. The Fifth Circuit once stated that when considering whether there is racial bloc voting under Gingles, “elections involving the particular office at issue will be more relevant than elections involving other offices.” Magnolia Bar Ass'n, Inc. v. Lee, 994 F.2d 1143, 1149 (5th Cir. 1993). We agree that legislative elections are the most probative here, but we also acknowledge that evidence as to other elections is potentially relevant.
The Defendants' expert Dr. Alford testified that the legislative elections Dr. Handley considered are better classified as semi-endogenous. Trial Tr. 1447:17-1448:13. His clarification was because the 19 elections concern the right office but not the right districts since Dr. Handley used the 2012 district lines rather than the 2022 enacted district lines. Trial Tr. 1448:11-13; PTX-004, 9. We do not resolve the definitional dispute but simply accept that the 19 legislative elections selected by Dr. Handley are the most analogous ones.
The 2022 district lines and the 2023 state legislative election returns for all county precincts became available as of November 2023, just a few months before trial, but neither Dr. Handley nor any other expert analyzed those returns. Trial Tr. 1202:21-1203:4; Doc [219], 44. Instead, under the previously enacted plans, Dr. Handley considered any election within a district if the district was either wholly contained in any of her seven areas or overlapped with either the illustrative or enacted districts. Trial Tr. 1202:21- 1203:4; Doc [219], 44. She then performed an effectiveness analysis. PTX-004, 58-60; Trial Tr. 322:19-22. Whatever arguments could be made that later and better data existed, we find that at most they affect, slightly, the weight to be given to Dr. Handley's findings.
Dr. Handley used these results to find that black citizens voted cohesively for their candidates while white voters cohesively opposed the black-preferred candidates in all state legislative elections. PTX-004, 12. None of the following findings were limited to just the three geographical areas of the relevant illustrative districts, but there also is no suggestion of significant regional polarization variations in Mississippi. Here, too, we find that the absence of evidence focused just on the three areas is a matter of the weight of the evidence.
Dr. Handley found that on average, 83.3 percent of black voters supported black-preferred candidates compared to 18.3 percent of white voters. Id. Additionally, black candidates were successful in the state legislative elections only in majority-minority districts. Trial Tr. 294:17-21. The black-preferred candidate is sometimes a white Democrat, ones like Senator Hob Bryan, Senator David Blount, and Representative Shandra Yates (first elected as a Democrat, then re-elected as an Independent). Stipulations [199] App. A at 17-19. The Defendants showed that in Enacted S.D. 2, which contains only a 33 percent BVAP, black Democrat (and Plaintiff) Pamela Hamner obtained 43 percent of the vote. PTX-004, 16. Enacted S.D. 7 has a 40.08 percent BVAP and 43.9 percent effectiveness score from Dr. Handley, but white Democrat Hob Bryan obtained 54.89 percent of the vote. Trial Tr. 347:23-348:25.
Dr. Handley's effectiveness analysis is not altered by these references. Under Senate Factor 2, “proof that some minority candidates have been elected does not foreclose a § 2 claim.” Gingles, 478 U.S. at 75. “‘[T]he election of a few minority candidates does not necessarily foreclose the possibility of dilution of the black vote.'” Clark v. Calhoun County, 88 F.3d 1393, 1397 (5th Cir. 1996) (citation omitted). At most these examples show an occasional breakdown of racial polarization for reasons not explained in the record. Quality of candidates and of opposition cannot always be irrelevant. One of the Plaintiffs' experts, Dr. Marvin King of the University of Mississippi, testified as to those considerations:
Q. Dr. King, you also refer to candidate viability. What do you mean by that?
A. Sure. So - so Black voters are strategic, right? So, you know, there are instances when Blacks may not support the Black Democrat candidate if they are not viable. And what we mean by that is they haven't raised money, they might be a perennial candidate that puts their name on the ballot election after election.Trial Tr. 774:22-775:4.
Further, racial polarization in voting is not disproved by evidence that black voters supported a white candidate. We are concerned with racial polarization of voters, i.e., are white voters consistently preventing the election of the candidates that black voters would choose? It is the race of the voters, not of the candidates, that matter.
Dr. Handley testified to the presence of racially polarized voting in all 19 elections, but to the extent possible we focus on her testimony related to the areas of Illustrative S.D. 2, S.D. 9, and HD 22. Of course, none of the elections occurred in the precise districts at issue. Trial Tr. 284:8-18. We find that, nonetheless, the relevant areas comprise Enacted S.D. 2, 10, 19, 34, 42, and 45 and Enacted HD 36 and 39. See PTX-004, 58-60 (App. B).
Of these districts, Dr. Handley specifically testified as to Enacted HD 36 and 39 and Enacted S.D. 42 and 45. Trial Tr. 289-93. Enacted HD 36 was a black-majority district in which the black-preferred candidate won 77.7 percent of the votes and the white-preferred candidate received essentially no black support. Trial Tr. 290; PTX-004, 59 (App. B). Enacted HD 39 was not a black-majority district, and the white candidate received 75 percent of the vote to the black-preferred candidate's 25 percent. Trial Tr. 293:2-9; PTX-004, 60 (App. B). Dr. Handley concluded that, in Enacted S.D. 42, had there been more black voters in the district, the black-preferred candidate would have carried the election. Trial Tr. 291:22-292:2. Instead, the majority-white district gave 85.8 percent of the votes to the white-preferred candidate. PTX-004, 59 (App. B). In Enacted S.D. 45, the white-preferred candidate received 86.8 percent of the vote, while the black-preferred candidate received 13.2 percent. PTX-004, 59 (App. B). The white candidate received over 95 percent of the white vote, while the majority of black voters supported the black-preferred candidate. Trial Tr. 292:3-19. From this, Dr. Handley concluded there was clear racial polarization in each applicable district. Trial Tr. 289-93.
Although Dr. Handley testified all 19 elections were racially polarized, the Defendants argue the black-preferred candidate won in eight of the 19 elections and seven elections were only able to be “best guess estimate[s]” of the racially polarized voting, indicating reliability issues in Dr. Handley's analysis. Doc [219], 45. In most of the state legislative contests (14 out of 19), Dr. Handley found that cohesion among black voters was at least 75 percent. Pls. FOF ¶¶ 162-163. The average level of black support for preferred candidates across the seven areas was 94.3 percent in biracial contests. Trial Tr. 283:6-10. The Defendants' expert did not dispute that black voters are cohesively supporting preferred candidates in those areas.
In Dr. Alford's opinion, Dr. Handley's definition of cohesion is flawed. Dr. Handley defines cohesive voting as when “a substantial number of minority voters consistently vote for the same candidates,” but she does not identify a specific numerical threshold to determine cohesion. Trial Tr. 267:8-19. Dr. Handley testified that courts do not set a bright-line rule on cohesive voting and instead provide a range. Trial Tr. 267:14-19. Dr. Alford generally agreed with that but then criticized Dr. Handley's analysis that found “every preferred candidate gets cohesive support [b]ecause they couldn't be the preferred candidate” without cohesive support. Trial Tr. 1455:1-10. Dr. Alford explained the candidates would not be considered preferred “unless they [had gotten] more votes than the other candidate” in the election. Id. In his opinion, no finding can be made that the second and third Gingles preconditions are met to any degree of scientific certainty because Dr. Handley's cohesion standard lacks any indication of the needed level of black support to meet that standard. Id.; DX-001, 11.
We credit Dr. Handley's testimony that, even absent a precise definition for cohesiveness, black and white voters are voting for separate candidates at a sufficiently high percentage to satisfy the Gingles requirements for racial polarization. In majority-black districts, the evidence shows that white voters do not prevent the election of candidates that black voters prefer, but that fact supports and does not undermine that preconditions two and three are satisfied. We thus reject Dr. Alford's criticisms of Dr. Handley's methodology or her findings.
b. Statewide General Elections
Dr. Handley testified that legislative elections were the most probative, but she also found useful data from other elections. She analyzed 17 statewide general elections in each of the seven illustrative district areas. PTX-004, 8-9. Eleven elections had a black Democrat running against a white Republican, and the remaining five had no black candidates. Id. Dr. Handley concluded that voting was consistently and starkly racially polarized in all 17 elections. PTX-004, 11-12. An average of 94.3 percent of black voters cohesively supported their preferred candidate, while an average of 6.9 percent of white voters voted for the black-preferred candidate in the 11 biracial elections. Id. at 11. In the five elections with no black candidates, the average white support for black-preferred candidates increased to 9.1 percent for white candidates. Id. at 11-12. Specifically in these statewide general elections, Dr. Handley calculated what she qualified as “quite stark” voting behavior. Trial Tr. 266:7. We provide Dr. Handley's calculations, then address Dr. Alford's responses.
Dr. Handley analyzed the 2020 presidential race in which a white man, Joseph Biden, headed the Democratic ticket and a black woman, Kamala Harris, was his running mate. PTX-004, 8. Neither Dr. Handley nor the Plaintiffs presented further statistics or analysis on the 2020 Presidential election other than to state it “was also starkly polarized in all seven areas of interest.” PTX-004, 11 n.16. Thus, we discuss only the 16 statewide elections where meaningful data was presented to the court.
First, the areas encompassing the three illustrative districts that satisfy Gingles precondition one. For Illustrative S.D. 2 and 9, those are Areas One and Four. In Area One, black-voter support for black-preferred candidates ranged from 96.6 percent to 73.9 percent, with an average of 92.29 percent. PTX-004, 37-39; Trial Tr. 281:22-282:11. White-voter support for the black-preferred candidate ranged from 20.5 percent to 6.3 percent, with an average of 9.67 percent. PTX-004, 37-39; Trial Tr. 282:12-20. In Area Four, black-voter support for the black-preferred candidate ranged from 96.2 percent to 83.2 percent, with an average of 93.73 percent. PTX-004, 46-48; see generally Trial Tr. 281:22-283:5. White-voter support for the black-preferred candidate ranged from 13.1 percent to 2.8 percent, with an average of 5.02. Id.
For Illustrative HD 22, we consider Dr. Handley's Areas Two and Six. In Area Two, black-voter support for the black-preferred candidate ranged from 97.4 percent to 85.3 percent, with an average of 95.23 percent. PTX-004, 40-42; see generally Trial Tr. 281:22-283:5. White-voter support for the black-preferred candidate ranged from 14.7 percent to 2.9 percent, with an average of 5.84 percent. Id. In Area Six, black-voter support for the black-preferred candidate ranged from 96.9 percent to 85.2 percent, with an average of 94.98 percent. PTX-004 at 52-54; see generally Trial Tr. 281:22- 283:5. White-voter support for the black-preferred candidate ranged from 14.8 percent to 3.1 percent, with an average of 5.35 percent. Id.
Now, the statistics for the remaining three areas. In Area Three, black voter support for the black-preferred candidate ranged from 98.4 percent to 89.3 percent, with an average of 96.61 percent. PTX-004, 43-45; see generally Trial Tr. 281:22-283:5. White-voter support for the black-preferred candidate ranged from 17.5 percent to 4.2 percent, with an average of 7.49 percent. Id. In Area Five, black-voter support for the black-preferred candidate ranged from 98.3 percent to 89.0 percent, with an average of 96.31 percent. PTX-004, 49-51; see generally Trial Tr. 281:22-283:5. White-voter support for the black-preferred candidate ranged from 32 percent to 5.3 percent, with an average of 15.68 percent. Id. In Area Seven, black-voter support for the black-preferred candidate ranged from 96.6 percent to 82.9 percent, with an average 94.26 percent. PTX-004, 55-57; see generally Trial Tr. 281:22-283:5. White-voter support for the black-preferred candidate ranged from 11.2 percent to 2.2 percent, with an average of 3.86 percent. Id.
Dr. Handley explained only the elections results in Area One at trial. See Trial Tr. 276-283. She discussed her determinations with respect to two of the elections in this area of interest, finding clear black-voter cohesion and racial polarization. See Trial Tr. 278-281. Using her specific results from Area One and the above voter-support ranges for the remaining six areas, Dr. Handley opined that the 17 statewide general elections portrayed black voters as being “very cohesive” and that “[y]ou couldn't get much more cohesive than this.” Trial Tr. 279:21-22; see generally Trial Tr. 276-283. She further articulated that “[v]oting is quite polarized in th[e] seven areas” of interest. Trial Tr. 283:18.
Dr. Alford used Dr. Handley's data to formulate his own opinions about the 17 statewide elections. He insisted that the data does not demonstrate racially polarized voting but, instead, voting that is polarized around party. Trial Tr. 1443:1-:13. Dr. Alford explained that the data showed a gap in voter preference in these statewide general elections among black and white voters in a contest between two white candidates with a mean difference of 86.5 percent based solely on party. DX-001, 10. In an election between a white Republican and a black Democrat, however, the mean difference in voting support among black and white voters was 87.6 percent. Id. Removing the racial cue, the mean difference between the white Republican and black Democrat is 86.3 percent. Id. Democratic candidates had a 90 percent range of cohesive support by black voters, and Republican candidates had an 80-90 percent range of cohesive support by white voters. Id. at 7-8. In Dr. Alford's opinion, whether a candidate is a Democrat or a Republican makes an almost 90 percent difference to black and white voters in these elections, but whether a candidate is black or white barely registers. Id. at 10.
Dr. Alford asserted this indeed shows partisan polarized voting because “when we vary the race of candidates, it simply doesn't [] make a difference.” Trial Tr. 1442:10-11. Dr. Alford conceded, however, that it is “not empirically untrue of” Dr. Handley's data that “Black voters overwhelmingly prefer Black candidates[, and] White voters overwhelmingly prefer White candidates.” Trial Tr. 1443:1-6. He clarified this “is not actually something the [data] could disconfirm” and opined the better conclusion is that partisanship is what is polarizing the voters because “the data [in its entirety] doesn't support” race as the polarizing factor. Trial Tr. 1443:5-12.
We find Dr. Alford's concession - that race versus party as the motivating factor is quite difficult to distinguish from the data - may be true but does not prevent us from making findings here. We see in the 16 statewide contests almost always a range of no less than 95 percent support by black voters for the Democratic candidate, and, in those same elections, less than 10 percent support from white voters. There are aberrant elections, as both experts testified. The most significant is when the last of the Democratic statewide officials, Attorney General Jim Hood, ran for Governor in 2019. He received almost 18 percent of the white vote and 96 percent of the black vote. We contrast that with the evidence in Dr. Handley's report that when black Hattiesburg mayor Johnny DuPree was the Democratic nominee for Governor in 2011, his white crossovers were only 8.4 percent, less than half of what Attorney General Hood received. Race matters.
We do not find that, if an occasional candidate supported overwhelmingly by black voters can break the color barrier just a bit among white voters, this proves anything other than that there will be outliers in statistical analysis.
Similarly, at times the black Democratic candidate did not have a significant campaign. One of these was the 2015 Governor's race, where the black Democratic nominee Robert Gray received only 84 percent of the black vote. No evidence was introduced about him, but we accept that election as an example of a Democratic candidate who may not have appeared credible to many black voters. We find that this data demonstrates racial bloc voting exists in the areas of the three illustrative districts that satisfied the first precondition.
c. Statewide Democratic Primaries and Judicial-District Elections
Dr. Handley next analyzed eight statewide biracial Democratic primary elections. PTX-004, 10-11. She found reliable statistical estimates impossible to generate solely in the seven areas and thus reported statewide estimates. Trial Tr. 296:21-297:4. Dr. Handley focused solely on Democratic primaries because black voters generally prefer Democrats over Republicans, meaning there would be no black-preferred candidates in the Republican primaries. Moreover, so few black voters participate in Republican primaries such that reliable voting behavior estimates cannot be produced. Trial Tr. 296:9-18. Just these preliminary observations reveal that these elections are not useful. Accordingly, we do not discuss the evidence with the exception of the following.
Dr. Handley examined three recent Mississippi non-partisan elections for the state supreme court in which there was at least one black and one white candidate. She found sharp racial polarization in the two contests in which the black candidate won in a district with a voting-age population fairly evenly divided between black citizens and white citizens. She found no racial polarization in the third election, won by the white candidate in a district with a substantial white-majority electorate. PTX-004, 13. The two polarized elections had over 90 percent black-voter support for black candidates and white-voter support for white candidates. Id. Dr. Handley conducted this analysis solely “to rebut the contention that it could be party not race” causing these results, and she concluded polarization was present “regardless of the fact that party was not on the ballot.” Trial Tr. 300:7-8, 21-22.
Dr. Alford's rebuttal was to acknowledge that while Mississippi judicial elections are nominally nonpartisan, it is a widely acknowledged fact that the candidates and their campaigns are anything but nonpartisan. DX-001, 15. Partisan advertisements, appeals, candidate endorsements, and donations are common in Mississippi, even resulting in court battles over party alignment for candidates. See generally James v. Westbrooks, 275 So.3d 62 (Miss. 2019). We agree with the Plaintiffs that not all voters would be aware of the partisan alliances behind individual supreme court candidates. Nonetheless, a high-enough percentage of voters know which party supports which judicial candidate for us to reject Dr. Handley's factual claims as to these elections.
We find racial polarization among voters in Mississippi is quite high. Trial Tr. 267:4-6. Black-preferred candidates are consistently unable to win elections unless running in a majority-minority district. Trial Tr. 268:1-9. White voters are also cohesive in voting for candidates that usually defeat the black-preferred candidates. Though there is no standard set by courts on the level of cohesion needed to support the analysis under Gingles, we find that Dr. Handley is correct that the level of cohesion here is sufficient, particularly for the areas encompassing the three valid illustrative districts.
Having met their burden under the second and third Gingles preconditions, the Plaintiffs have no duty “to disprove that factors other than race affected voting patterns” in the areas of interest. Teague v. Attala County, 92 F.3d 283, 290 (5th Cir. 1996). The Defendants, however, could “rebut the plaintiffs' evidence by showing that no such [racial] bias exists in the relevant voting community.” Id.
We have already mentioned that the Defendants attempt to rebut by saying this is all partisanship, not race. Whether that argument should be addressed here or under the totality-of-the-circumstances analysis is unclear. The Defendants at least acknowledged that it could fall under Senate Factor 2, see Doc [218], 15, but precedent sometimes blurs that line.
For example, the en banc court in LULAC v. Clements held that the evidence failed to show race, not partisanship, best explained voting in Dallas County. 999 F.2d at 877. As such, there was no “threshold showing required by Gingles.” Id. That implicates the preconditions. That same opinion, however, also noted when discussing partisanship that the question is “whether the political processes are equally open,” which rests “upon a searching practical evaluation of the past and present reality.” Id. at 860 (quotation marks and citations omitted). That is a totality inquiry. See Milligan, 599 U.S. at 19.
Just where the discussion best fits will not matter in this case, as ultimately the evidence does not support that racial polarization has become partisan divisions. For detailed reasons we set out in our analysis of Senate Factor 2, we find that, although there certainly was evidence and argument presented on that possibility, the Defendants have not rebutted the Plaintiffs' showing under Gingles preconditions two and three.
The Plaintiffs have thus satisfied the three Gingles preconditions for Illustrative S.D. 2, S.D. 9, and HD 22.
3. Totality of the Circumstances
Having found that the Plaintiffs satisfied all three preconditions for a vote-dilution claim, we must now engage in a “searching practical evaluation of the past and present reality” of the political process in Mississippi. Gingles, 478 U.S. at 79 (quotation marks and citation omitted). That evaluation considers the totality of the circumstances, described as “an intensely local appraisal of the design and impact of the contested electoral mechanisms.” Id. (quotation marks and citation omitted).
In 1986, the Supreme Court adopted “typical factors” to be considered in this portion of the Section 2 vote-dilution analysis. Id. at 36 & n.4. The United States Senate referred to these same factors in its report on the 1982 amendments to the Voting Rights Act. See S. Rep. No. 97-417, at 23, 28-29. Most courts refer to them as the Senate Factors, and so will we.
The following is the usual enumeration:
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; [and]
7. the extent to which members of the minority group have been elected to public office in the jurisdiction. . . .
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; [and]
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.Teague, 92 F.3d at 292-93 (alterations in original) (quoting Gingles, 478 U.S. at 36-37).
Gingles invited consideration of other factors: “While the enumerated factors will often be pertinent to certain types of [Section] 2 violations, particularly to vote dilution claims, other factors may also be relevant and may be considered.” Gingles, 478 U.S. at 45 (footnote omitted). Rarely do courts consider others, though. We do not consider Factor 4, as no evidence was offered relevant to the slating of candidates.
Importantly, the totality-of-the-circumstances inquiry recognizes that a court's application of these factors “is peculiarly dependent upon the facts of each case.” Id. at 79. Defendants may attempt “to rebut plaintiffs' claim[s] of vote dilution via evidence of objective, nonracial factors” like partisan politics; as we stated before, it is not the Plaintiffs' burden to negate “all nonracial reasons possibly explaining” voting patterns. Teague, 92 F.3d at 292, 295 (quotation marks and citation omitted).
Indeed, “[i]t will be only the very unusual case in which” the Gingles preconditions are established and liability does not follow, and, in such a case, the court “must explain with particularity why it has concluded” there is no Section 2 violation. Id. at 293 (quoting Clark v. Calhoun County, 21 F.3d 92, 97 (5th Cir. 1994)).
We now review the factors.
a. Senate Factors 1 and 3
The first factor in examining the totality of the circumstances is “the extent of any history of official discrimination in the state” that diluted or denied “the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process.” Gingles, 478 U.S. at 36-37. The third is “the extent to which the state . . . has used . . . voting practices or procedures that may enhance the opportunity for discrimination against the minority group.” Id. at 37. We consider the two factors together because both look at the history of discrimination. The evidence as to each is at times the evidence as to both.
How much history to consider is one question. To answer it, we distinguish between judicial precedents that examined history relevant to an Equal Protection analysis, which requires showing discriminatory or invidious intent, and those that examined history relevant to Section 2 of the Voting Rights Act.
The Defendants argue that distant history of discrimination in Mississippi is all but irrelevant. Among their authorities is Shelby County, 570 U.S. 529. There, the Supreme Court concluded the factual circumstances in 1965 that allowed Congress to impose preclearance obligations on changes to election and voting rules in those states with a history of racial discrimination were constitutionally insufficient to support the continued application of the preclearance obligations. Id. at 556-57. As the Court phrased it, the “extraordinary measure[]” of requiring certain “States to obtain federal permission before enacting any law related to voting” was “a drastic departure from basic principles of federalism.” Id. at 534-35. Making that requirement applicable only to some states was “an equally dramatic departure from the principle that all States enjoy equal sovereignty.” Id. at 535. The Court held that current conditions must justify so wrenching a change to constitutional norms as to require states to preclear their laws. Id. at 536.
Obviously, Shelby County was not a Section 2 case and did not concern Senate Factor 1. Moreover, we are involved in a less revolutionary task than distorting the constitutional norms of equal sovereignty. We are considering a range of circumstances, including history, that the Supreme Court has identified to help us evaluate a State's recent redistricting decision. The Shelby County decision thus does not provide us with relevant guidance.
Further, the Defendants rely on a recent Fifth Circuit opinion which held that the circumstances surrounding the adoption of Mississippi's 1890 Constitution did not invalidate voting qualifications as currently applied in the state. Harness v. Watson, 47 F.4th 296, 306-307, 311 (5th Cir. 2022) (en banc), cert. denied, 143 S.Ct. 2426 (2023). That was an Equal Protection claim for which the Fifth Circuit had to decide whether the invidious intent behind the initial measure was relevant after later amendments. Id. at 299, 303. There was no analysis of Senate Factor 1. See generally id.
The Defendants urge us to interpret a 2001 Fifth Circuit opinion applying the Voting Rights Act as closing the door to any review of older history. Doc [219], 60 ¶ 206. The opinion concerned Mississippi's being divided into three districts for election of state supreme court justices and members of two state commissions. NAACP v. Fordice, 252 F.3d 361, 364 (5th Cir. 2001). The court remarked that “the abysmal reality of Mississippi's history of official discrimination regarding the right of African- Americans to register and to vote is evident in the record,” and “that African-Americans in Mississippi are less educated, suffer from higher unemployment, earn lower incomes, and live in disparate conditions as compared to Mississippi's white citizens.” Id. at 367. The court then stated that unless plaintiffs show “these facts ‘actually hamper the ability of minorities to participate,'” the evidence does not “support a finding that minorities suffer from unequal access to Mississippi's political process.” Id. (quoting LULAC v. Clements, 999 F.2d at 866). We consider the Fifth Circuit's holding to focus on what is not enough by itself, not to limit what we are to consider as to the totality.
We find no bar in Shelby County, Fordice, or Harness as to when older history becomes too attenuated or diluted by later events to be relevant in a Voting Rights Act claim. There is, however, a clear statement in a Fifth Circuit Voting Rights Act opinion: “[T]he most relevant historical evidence is relatively recent history, not long-past history,” but “even long-ago acts of official discrimination give context to the [Gingles] analysis.” Veasey v. Abbott, 830 F.3d 216, 232, 257 (5th Cir. 2016) (quotation marks and citation omitted). At least for context, then, early history of discrimination has relevance.
For all that, there is no reason to summarize the evidence concerning the earlier history and determine its precise relevance. The Defendants have accepted the accuracy of the statement in one precedent “[t]hat Mississippi has a long and dubious history of discriminating against blacks is indisputable.” Teague, 92 F.3d at 293-94. We find, based on the evidence introduced in this case, that the long and dubious history, with significant acts of violence still occurring, lasted at least through the 1960s. We examine what has occurred since then.
Though we just disclaimed full consideration of early history, we will start with the Plaintiffs' arguments concerning how certain provisions in the 1890 Mississippi Constitution continue to have discriminatory effects. Two of the Plaintiffs' expert witnesses, the previously mentioned Dr. Marvin King and Jackson State University Professor Robert Luckett, elaborated on the effects of the 1890 constitution. The constitution contained both a poll tax and literacy test, allowed for the creation of all-white primaries and segregated education, and created a means to disenfranchise black Mississippians, though it also could have operated to prevent poor and poorly educated whites from voting. Trial Tr. 375:4-377:22, 379:2-17, 421:10- 422:16. Of these provisions, the only identified provisions still in effect in Mississippi are a lifetime disenfranchisement of those convicted of certain crimes and a residency requirement. Trial Tr. 451:13-452:2, 454:15-456:16. There was evidence that only the disenfranchisement provision continues to disproportionately disenfranchise black citizens. The Fifth Circuit concluded that this current, somewhat-amended list of disenfranchising crimes has been shorn of its original discriminatory purpose. Harness, 47 F.4th at 306-307, 311.
Under Section 2 of the Voting Rights Act, though, we are concerned with whether there are discriminatory effects. Plaintiffs' expert Dr. Byron D'Andra Orey testified that “formerly incarcerated individuals are less likely to participate in politics” in Mississippi and that “African-Americans are disproportionately incarcerated.” Trial Tr. 519:22-25. Dr. Orey opined that Mississippi's disenfranchising laws cause a huge racial disparity in whether black Mississippians are able to participate in voting. Id.; see Trial Tr. 569:19-571:13. Dr. Orey further identified that Mississippi's severe lifetime-disenfranchisement law and black Mississippians' overrepresentation in the criminal-justice system lead to black Mississippians being 12 percent less likely to vote while white Mississippians are one percent less likely to do so. PTX-008, 19-20; see Trial Tr. 569:19-571:13. Dr. Luckett, another Plaintiffs' expert, presented further evidence that black Mississippians are incarcerated at disproportionately higher rates than whites. Trial Tr. 416:6-13. We are not concerned here with policy justifications, only whether the disenfranchisement provision is disproportionate in its effects. We accept the testimony as credible that it does leave a higher percentage of black than white Mississippians ineligible to vote.
The two Plaintiffs' experts also found racial discrimination in the State's rules for nominating candidates. Dr. King found that in 1972, the legislature first adopted the requirement that party primaries include a runoff if no candidate receives a majority, and that the adoption was “an effort to minimize nascent Black voting strength.” PTX-013, 16-17. That is factually incorrect. The statute Dr. King cites as creating the requirement of runoff primaries was the recodification of a prior statute into a new Mississippi Code adopted in 1972. See Miss. Code Ann. § 23-3-69 (1972) (recodifying Miss. Code § 3194 (1942)). A still earlier statute was the first to mandate runoff primaries. See Miss. Code § 3701 (1906).
The Plaintiffs do not argue that party primaries are discriminatory. Their arguments solely apply to the runoff requirement. Because the only testimony to support a discriminatory purpose to runoff primaries was inaccurate, we find no racial motive behind adoption of the requirement.
Regardless of motive, though, runoff primaries qualify as potential evidence relevant to Senate Factor 3 as a majority-vote requirement that could provide an opportunity for discrimination. Any discriminatory effect of the runoffs applies only to party nominations, not to general elections. In fact, the report by Plaintiffs' expert Dr. Lisa Handley stated that “candidates supported by black voters usually managed to win the Democratic nomination.” PTX-004, 13. Her trial testimony was arguably even stronger: “[E]ven if voting is [racially] polarized in the Democratic primary, it's relatively easy, because so few whites participate in the Democratic primary for the candidate of choice of Black voters to succeed in the Democratic primary.” Trial Tr. 296:4-8. We find that the Plaintiffs did not present evidence to support that runoff primaries enhance the opportunities to discriminate against black voters in Mississippi.
The Plaintiffs also criticize a 2023 statute governing the removal of names, or purging, from the voting rolls. See H.B. 1310, 2023 Leg., Reg. Sess., 2023 Miss. Laws, ch. 534 (codified as Miss. Code Ann. § 23-15-153). It provides that, along with other reasons, names of registered voters may be removed from the voter rolls if they have “failed to comply with the provisions of Section 23-15-152.” Miss. Code Ann. § 23-15-153(1). Both the referenced Section 23-15-152 and a federal statute allow removal of voters' names when they have not voted for two consecutive federal elections. Compare Miss. Code Ann. § 23-15-152(4), with 52 U.S.C. § 20507(d)(1)(B)(ii). Federal law also requires that state policies be nondiscriminatory and not result in the removal of a registered voter who has voted or appeared to vote in one of the past two general federal elections. 52 U.S.C. § 20507(b)(1)-(2). We have not been shown how the Mississippi statutes on removing names of those who have not voted violate any federal law.
Another of the Plaintiffs' claims concerns Mississippi's absentee-voting rules. Mississippi restricts absentee voting to voters with one of eight acceptable excuses. Miss. Code Ann. § 23-15-713(a)-(h). Requesting an absentee ballot by mail may also require notarization. See § 23-15-627. The Plaintiffs do not claim, and we do not hold, that the State is violating the legal rights of voters in having more restrictive absentee-voting practices and not allowing early voting. Those are policy decisions on the proper procedures for conducting elections. A more traditional state might not embrace all the current options for voting other than in person on election day. The factual issue under Senate Factor 3, however, is whether identified voting procedures tend to enhance the opportunity for discrimination against the minority group. Dr. Luckett testified that Mississippi's restrictions on absentee voting disproportionately impact black voters who rely on absentee voting in greater proportions than white voters due to financial hardships, work requirements, and health disparities. Trial Tr. 403:18-408:9. We credit that testimony. The main factor allegedly driving the disproportionate impact is the nature of black poverty. See Trial Tr. 404:19-406:20.
The Plaintiffs also claim a recent statutorily implemented voting practice disproportionately affects black voters. In 2023, the legislature enacted House Bill 1020, which expanded the already existing Capitol Complex Improvement District of Jackson (“CCID”). See H.B. 1020, 138th Leg., Reg. Sess., 2023 Miss. Laws, ch. 546. Dr. Luckett asserted the expansion of the police force and court system in the CCID now encompassed “what constitutes 100 percent of the measurable white population in the city of Jackson.” Trial Tr. 411:4-6. Dr. Luckett claimed this enactment was “part of a long history and a continuum of attempts to diminish and disfranchise African-Americans in the state of Mississippi from 1868 to the present.” Trial Tr. 413:2-4.
Dr. Luckett's testimony was based on a provision of the enactment that created four temporary judgeships to be appointed by the State's chief justice. See H.B. 1020, supra, at sec. 1. Because black residents compose 83 percent of the Jackson population, Dr. Luckett asserted those unelected judges would disproportionately impact black voters. See Trial Tr. 412.
Events have overtaken the issue. In September 2023, the Mississippi Supreme Court held that creating four temporary appointed judgeships with terms just short of four years violated the state constitution. Saunders v. State, 371 So.3d 604, 608 (Miss. 2023). On the other hand, the court recognized the chief justice's long-existing and frequently utilized statutory authority to appoint special judges to assist a “judicial district in Mississippi facing exigent circumstances.” Id. (emphasis in original); see also Miss. Code Ann. § 9-1-105(2). Such judgeships would end when the exigencies ended. See Saunders, 371 So.3d at 608. We thus find that the allegedly disenfranchising aspects of H.B. 1020 are gone. Moreover, the state court's analysis revealed that the same power to appoint temporary judges had already been granted to the chief justice. The part of H.B. 1020 that was invalidated was essentially a duplication of existing authority, though the prior statute did not mandate its immediate use in a specific locale as did H.B. 1020. The differences between the chief justice's authority under the two statutes is an insufficient basis to find a dilution of black voting rights.
The Defendants urge that under Senate Factor 1 we consider the history of legislative redistricting. The Defendants suggest we start with redistricting after the 2000 census. We instead will begin one decade earlier, when a three-judge district court rejected the Mississippi Legislature's plan adopted after the 1990 census and which had been used in the 1991 election; the court ordered a new election in 1992 under a court-ordered plan. Watkins v. Mabus, 771 F.Supp. 789, 797-98, 807 (S.D.Miss. 1991), aff'd in part, vacated in part, 502 U.S. 954 (1991). The Defendants are correct, though, that the next two redistricting plans for the legislature, adopted after the 2000 and 2010 censuses, were precleared by the Department of Justice under the then-operative requirements of the Voting Rights Act.
Less favorable evidence concerns congressional redistricting. The Plaintiffs' only argument regarding congressional redistricting was in one sentence detailing Mississippi's need to draw new Section 5-compliant congressional districts in 2002. See generally Smith v. Clark, 189 F.Supp.2d 503 (S.D.Miss. 2002). The additional court opinions that have often found those districts to violate either the Constitution or the Voting Rights Act are available to us for consideration as well. We consider them.
The legislature drew new congressional districts in 1981, but the United States Attorney General objected to the districts under the preclearance requirement of Section 5 of the Voting Rights Act. See Jordan v. Winter, 541 F.Supp. 1135, 1138 (N.D. Miss. 1982). The 1982 primaries and general elections were held under an interim redistricting plan designed by a three-judge district court. Id. at 1144-45. That court's 1982 decision was later vacated and the case remanded by the Supreme Court “for further consideration in light of Section 2 of the Voting Rights Act of 1965, 42 U.S.C. [§] 1973, as amended in 1982.” Brooks v. Winter, 461 U.S. 921, 921 (1983). That amendment allowed a finding of a Section 2 violation if the results of a measure relating to voting, regardless of intent, denied or abridged voting rights. See Voting Rights Act Amendments of 1982, Pub. L. No. 97-205, § 3, 96 Stat. 131, 134 (codified as amended 52 U.S.C. § 10301). On remand, the district court imposed a new map that increased the black voting-age population in the Second District. Jordan v. Winter, 604 F.Supp. 807, 810, 814 (N.D. Miss.), aff'd sub nom. Mississippi Republican Exec. Comm. v. Brooks, 469 U.S. 1002 (1984).
There was no litigation over the Mississippi Legislature's plan for congressional districts after the 1990 Census. That is the most recent congressional redistricting that has not been challenged.
After the 2000 Census, which led to Mississippi's loss of one congressional seat, the legislature failed to adopt a redistricting plan in time for the 2002 election filing deadlines. See Smith, 189 F.Supp.2d at 504-05. Therefore, a three-judge district court devised a redistricting plan and ordered that it be used for the 2002 elections and every succeeding election until the State produced an acceptable plan. Smith v. Clark, 189 F.Supp.2d 548, 559 (S.D.Miss. 2002). The Supreme Court affirmed. Branch v. Smith, 538 U.S. 254, 265 (2003).
After the 2010 Census, the same three-judge district court found the new census rendered the court's 2002 plan malapportioned, but the legislature had yet to produce a viable new plan for use in the 2012 elections. Smith v. Hosemann, 852 F.Supp.2d 757, 760-61 (S.D.Miss. 2011). The court developed a plan to which no party objected, and the court in 2011 ordered its use until the State produced an acceptable plan. Id. at 765-67.
After the 2020 Census, the legislature adopted its own congressional-redistricting plan. Smith v. Hosemann, No. 3:01-cv-855, 2022 WL 2168960, *1 (S.D.Miss. May 23, 2022). The defendants moved to vacate the 2011 order, while the plaintiffs insisted the legislature's plan was invalid. Id. at *1- 2. The district court vacated the 2011 injunction, allowing the legislature's plan to go into effect, but it also held that the plaintiffs were not barred from seeking relief under the Voting Rights Act. Id. at *8. The United States Supreme Court dismissed for lack of jurisdiction. Buck v. Watson, 143 S.Ct. 770 (2023). There have been no further proceedings in that case.
Thus, unlike for legislative redistricting after the 2000 and 2010 censuses when there were no challenges to those plans, the legislature has been held to violate the Voting Rights Act in drawing new congressional districts in 1980, then in not drawing any in 2000 and 2010. This is some evidence of a continuation of official discrimination.
To summarize, this factor is about history. When the United States Supreme Court upheld a district court's finding in Milligan that the evidence supported Senate Factor 1, the Court quoted the district court finding that “Alabama's extensive history of repugnant racial and voting-related discrimination is undeniable and well documented.” Milligan, 599 U.S. at 22 (quoting Singleton, 582 F.Supp.3d at 1020). We examine the Alabama district court's opinion to determine what the Supreme Court found sufficient. The district court included these events:
(1) A successful constitutional challenge to legislative redistricting after the 2010 census. Singleton, 582 F.Supp.3d at 1020. The Alabama suit concerned congressional redistricting yet considered constitutional defects in prior legislative redistricting. Our facts are a mix, such that in this challenge to the legislature's redistricting we have considered challenges in Mississippi both for Congress and for the legislature. We find that record to be of at least comparable weight as the record in Alabama.
(2) A successful challenge to “local at-large voting systems with numbered post created by the State Legislature.” Id. at 1021. This is another example of a successful challenge to a legislative measure, and we have identified relatively recent and successful litigation in Mississippi both for congressional and legislative redistricting.
(3) “[T]he Justice Department has sent election observers to Alabama nearly 200 different times, and . . . between 1965 and 2013, more than 100 voting changes proposed by the State or its local jurisdictions were blocked or altered under Section 5 of the Voting Rights Act.” Id. Here, the Justice Department issued 81 voting-determination letters to Mississippi counties between 1985 and 2012, ordering districts be redrawn and voting practices be amended (PTX-063; Trial Tr. 396:16-400:12); Mississippians have successfully challenged at-large election procedures (PTX-007, 29-30; Trial Tr. 389:3-390:25); and more restrictive voting changes proposed by Mississippi were enjoined by federal courts (PTX-007, 38; Trial Tr. 405:24- 406:9; 407:1-408:9; 1231:25-1232:12).
We consider the evidence here to be comparable to what the Milligan Court found satisfied the first and third Senate Factors for Alabama. To be sure, Mississippi is a much different state today than during the period of its “long and dubious history of discriminating against blacks.” Teague, 92 F.3d at 293-94. Nonetheless, we find more recent events still cause the first and third Senate Factors to weigh in favor of the Plaintiffs.
b. Senate Factor 2
Senate Factor 2 measures “the extent to which voting in the elections of the state or political subdivision is racially polarized.” S. Rep. No. 97-417, at 29. Said differently, the factor considers “racial bloc voting.” Gingles, 478 U.S. at 52 n.18; see also S. Rep. No. 97-417, at 55.
As explained before, Senate Factor 2 overlaps with Gingles precondition three, which is also routinely described as concentrating on racial polarization. See, e.g., Milligan, 599 U.S. at 19 (identifying Gingles precondition three as “focused on racially polarized voting”). We have concluded under Gingles precondition three that the Plaintiffs met their burden and the Defendants failed to rebut it. We now consider the issue under the totality of the circumstances.
When viewed under the totality of the circumstances, racial bloc voting is one of the two most important factors, the other being minority candidates' success or failure in elections. Gingles, 478 U.S. at 48 n.15 (citing S. Rep. No. 97-417, at 28-29). If both factors are “present, the other factors . . . are supportive of, but not essential to, a minority voter's claim.” Id. (emphasis omitted).
To begin, the scope of relevant evidence is broader at this stage. While we focused our review of the Gingles preconditions on the districts at issue, we may consider more under the totality analysis. See LULAC v. Perry, 548 U.S. at 438 (citing Gingles, 478 U.S. at 44-45) (noting that statewide evidence has been used under other Senate Factors and finding it “[p]articularly” appropriate under the proportionality factor “given the presence of racially polarized voting . . . throughout Texas”); see also Milligan, 599 U.S. at 22 (relying on statewide evidence under totality of circumstances); Fordice, 252 F.3d at 370 (noting that exogeneous elections are probative under Senate Factor 2).
Turning to the facts, the parties agree that Mississippians vote along racial lines. But under the totality-of-the-circumstances analysis, “[a] defendant may try to rebut plaintiffs' claim of vote dilution via evidence of ‘objective, nonracial factors.'” Teague, 92 F.3d at 292 (quoting Nipper v. Smith, 39 F.3d 1494, 1513 (11th Cir. 1994)). The Defendants take that approach, arguing that the existing polarization is not “on account of race” because Mississippi voters are “driven by partisan polarization rather than racial polarization.” Doc [219] ¶ 233.
The Supreme Court recently considered a race-versus-partisanship issue in Alexander, 144 S.Ct. 1221. The plaintiffs' claim, though, was of racial gerrymandering violative of the Equal Protection Clause, and the question before the court was whether “race predominated in the drawing of a district.” Id. at 1252. The Alexander plaintiffs did “not rely on the Voting Rights Act of 1965,” nor did the defendants. Id. (Thomas, J., concurring in part). Though the Alexander Court considered a similar issue to the one before us, its decision is inapplicable to effects-based review under Section 2 of the Voting Rights Act. See Milligan, 599 U.S. at 13 (discussing the effects standard).
The parties say the Court must determine whether partisan affiliation or race “best explains the divergent voting patterns among minority and white citizens.” Doc [219] ¶ 183 (citation omitted); see Doc [220] ¶ 677; see also Lopez v. Abbott, 339 F.Supp.3d 589, 602-03 (S.D. Tex. 2018) (applying similar standard). That test comes from LULAC v. Clements, a Section 2 case filed by black and Hispanic residents challenging the way Texas elected its trial judges. 999 F.2d at 838.
The LULAC v. Clements district court held that “plaintiffs need only demonstrate that whites and blacks generally support different candidates to establish legally significant white bloc voting.” Id. at 850. The Fifth Circuit reversed, agreeing with the defendants that “the record indisputably prove[d] that partisan affiliation, not race, best explain[ed] the divergent voting patterns among minority and white citizens in the contested counties.” Id. (emphasis added). The appellate court also observed that “[a]bsent evidence that minorities have been excluded from the political process, a ‘lack of success at the polls' is not sufficient to trigger judicial intervention.” Id. at 853 (quoting City of Mobile v. Bolden, 446 U.S. 55, 109 (1980) (Marshall, J., dissenting)).
The Supreme Court has never applied a best-explains test, but it has not overruled that binding precedent.
The LULAC v. Clements court found no evidence that minorities had been excluded from the political process on account of race. Id. at 861. Instead, it found facts “unmistakably” proving “partisan affiliation,” not race, defeated the minority-backed judicial candidates in Texas. Id. “First, white voters constitute[d] the majority of not only the Republican Party, but also the Democratic Party, even in several of the counties in which the former dominate[d].” Id. Second, both parties, “especially the Republicans, aggressively” nominated minority candidates who were then supported “without fail” by white voters just as much as those voters supported white candidates. Id. Finally, the court found no evidence that white elected officials were unresponsive to minority constituents, something the court considered a hallmark of racial bloc voting. Id. at 858-59.
This case is different. While we recognize that our review requires a “‘functional' and ‘practical'” examination rather than a mechanical checklist of the facts that buttressed the LULAC v. Clements opinion, id. at 861, none of those facts exist here. There is no proof that whites constitute a majority of the Democratic Party, that Republicans aggressively recruit and unfailingly support black Republican candidates, or that elected white officials respond to black constituents as they did in Texas.
On the responsiveness question, the parties addressed various public issues and whether the Republican-dominated legislature has been receptive to the stated desires of black voters. We are more persuaded, though, by witnesses Joseph Wesley, Kenneth Harris, Pamela Hamner, Gary Fredericks, and Terry Rogers who all testified - without contradiction - that their elected officials ignore the black community. Trial Tr. 669:3-670:8 (Wesley); 698:15-699:16, 702:9-14 (Harris); 726:10-727:2 (Hamner); 906:20-907:18 (Fredericks); 943:22-944:3 (Rogers). According to LULAC v. Clements, these facts indicate racial bloc voting and not partisan bloc voting. 999 F.2d at 859; see also Rogers v. Lodge, 458 U.S. 613, 623 (1982) (noting that racial bloc voting “allows those elected to ignore [minority] interests without fear of political consequences”).
Not only are the salient LULAC v. Clements facts missing here, but we also find no analogous facts exist in this record. Instead, the Plaintiffs presented credible evidence that the “political process is not ‘equally open' to minority voters.” Milligan, 599 U.S. at 18 (quoting Gingles, 478 U.S. at 45-46) (summarizing a plaintiff's burden under the totality-of-the-circumstances test). As noted above, the Enacted Plans split black communities like Horn Lake and Jago. See Section IV(B)(1)(b), supra. As a result, there is no dispute that black voters are disproportionally located in districts in which they are the minority. Under the Enacted Senate Plan, 84.33 percent of white voters live in white-majority districts while only 50.36 percent of black voters live in black districts. PTX-001, 50. A similar disparity exists under the Enacted House Plan - 82.92 percent of whites live in majority districts compared to just 62.38 percent of black voters. PTX-001, 74.
Once in those minority districts, it is almost impossible for a black-preferred candidate to prevail because crossover voting is nearly nonexistent. We credit Dr. Handley's testimony regarding “stark” polarization across varied local and statewide elections. Trial Tr. 266:7; see also Section IV(B)(2)(a)-(c), supra. When Dr. Handley considered votes in the areas of interest during statewide elections, black voters supported black Democrats 94.3 percent of the time while white voters crossed over at only a 6.9 percent rate. That crossover number rose to 9.1 percent when the Democrat was white. PTX-004, 11-12.
Other courts have found that crossover percentages like these support a finding of racial bloc voting. Most notably in Milligan, the district court found racial bloc voting under Gingles two and three where “on average, Black voters supported their candidates of choice with 92.3 [percent] of the vote” while “white voters supported Black-preferred candidates with 15.4 [percent] of the vote.” 599 U.S. at 22 (quoting Singleton, 582 F.Supp.3d at 1017). The Supreme Court affirmed. Id. at 23, 42.
A similar result followed in Robinson v. Ardoin, when white crossover voting was around 11.7 percent in statewide elections (per Dr. Handley) and about 20.8 percent in “a different set of elections” (per another expert witness). 605 F.Supp.3d 759, 842 (M.D. La. 2022), vacated and remanded, 86 F.4th 574 (5th Cir. 2023) (affirming finding of Section 2 violation but vacating injunction as legislature prepared new map). Black voters' support for their chosen candidates averaged 83.8 percent and rose to 93.5 percent in two-person races. Id. at 801. The Fifth Circuit addressed “racial polarization” under the third Gingles precondition and affirmed the finding that the plaintiffs proved its existence. Robinson, 86 F.4th at 597. The averages for Mississippi resemble - if not surpass - those in Milligan and Robinson.
The extent of that polarization further distinguishes LULAC v. Clements, which found no legal significance to “blacks generally support[ing] different candidates” than whites. 999 F.2d at 850. That case did not consider polarization of this magnitude or the effect it has on open participation when coupled with cracking and packing. We find as a factual matter the extent of the polarization between races across Mississippi provides at least circumstantial evidence that the divide is based on race.
This is not, however, the only evidence; other facts make this case much like Milligan. There, the Supreme Court explained that the plaintiffs had carried their burden at the totality-of-the-circumstances stage because “elections in Alabama were racially polarized; . . . ‘Black Alabamians enjoy[ed] virtually zero success in statewide elections'; . . . political campaigns in Alabama had been ‘characterized by overt or subtle racial appeals'; and . . . ‘Alabama's extensive history of repugnant racial and voting-related discrimination [was] undeniable and well documented.'” 599 U.S. at 22 (quoting Singleton, 582 F.Supp.3d at 1018-24). Though Alabama did not appeal the totality holding, the Supreme Court still found no reason to dispute the district court's “careful factual findings,” id. at 23, including the finding that Senate Factor 2 “weigh[ed] heavily in favor of the” plaintiffs' Section 2 claim because the “pattern of racially polarized voting [was] clear, stark, and intense,” Singleton, 582 F.Supp.3d at 1018.
These facts are equally true here in Mississippi. The two states share a similar history, and neither has elected a black candidate in a statewide election since Reconstruction. Trial Tr. 752:19-25, 788:14-18. Also, black legislative candidates in Mississippi have had virtually no success in federal or state elections outside majority-black districts. Trial Tr. 310:6-9, 388:10- 389:2, 789:1-7. We find that the Plaintiffs have shown racially polarized voting under the totality of the circumstances.
The Defendants attempt to rebut all this, however, and argue Mississippians vote based on party, not race. To begin, they say we should consider two Equal Protection cases that addressed the politics-versus-race issue: White v. Regester, 412 U.S. 755 (1973), and Whitcomb v. Chavis, 403 U.S. 124 (1971). “Congress codified the ‘results' test [the Supreme Court] had employed, as an interpretation of the Fourteenth Amendment, in White and Whitcomb.” LULAC v. Clements, 999 F.2d at 851 (quoting Gingles, 478 U.S. at 97 (O'Connor, J., concurring in the judgment)). To the extent they are relevant, our case is like White, not Whitcomb.
In White, the Court found Equal Protection violations in some Texas districts based on evidence of lack of minority political success since Reconstruction, a history of discrimination, voting requirements that depressed minority votes, racial campaign tactics, and a lack of responsiveness from elected officials in minority communities. 412 U.S. at 765-69; see also S. Rep. No. 97-417, at 21-22. Our facts and record are similar.
By contrast, the Court found in Whitcomb that politics, not race, motivated the disputed election practices because there were no impediments to black participation and black-preferred and minority candidates enjoyed some success at the polls with support of white voters. 403 U.S. at 149-50. According to LULAC v. Clements, Whitcomb
established a clean divide between actionable vote dilution and “political defeat at the polls”; the 1982 amendments [were] enacted to restore a remedy in cases “where a combination of public activity and private discrimination have joined to make it virtually impossible for minorities to play a meaningful role in the electoral process.”999 F.2d at 850-51 (emphasis omitted) (quoting Hearings on the Voting Rights Act Before the Subcomm. on the Constitution of the S. Comm. of the Judiciary, 97th Cong., 2d Sess. 1367-68 (1982) (statement of Professor Drew Days)). Again, the Plaintiffs have shown that impediments do exist in Mississippi - for example being disproportionately placed in minority districts - and that they have had almost no success electing their candidates of choice outside majority-minority districts.
The Defendants also rely on Dr. Alford's expert opinion that partisanship explains the polarization Dr. Handley's data reveals. Although we have considered all opinions Dr. Alford offered, none convince us that the Defendants have overcome the Plaintiffs' showing of racially polarized voting. We will, however, address a few examples.
Dr. Alford identifies select races in predominantly white house districts where Dr. Handley's statistics have massive confidence intervals, indicating uncertainty. See DX-001, 11 (Alford Report). It may be true uncertainty exists, but Dr. Handley explained that districts with small minority populations are harder to evaluate. Trial Tr. 278:5-11, 288:2-18. That statistical reality does not outweigh the breadth of Dr. Handley's opinions and supporting evidence about stark racial bloc voting, nor does it offer rebuttal evidence to meet the Defendants' burden.
Another fact Dr. Alford relies on is the recent success of a lone black Republican candidate in a legislative election. Trial Tr. 1470:18-23. The same thing happened in Milligan, but the district court held under Senate Factor 2 that “[o]ne election of one Black Republican is hardly a sufficient basis . . . to ignore . . . the veritable mountain of undisputed evidence that in all the districts at issue in this case, and in all statewide elections, voting in Alabama is polarized along racial lines.” Singleton, 582 F.Supp.3d at 1019. The Supreme Court took no issue with these Senate Factor 2 findings. 599 U.S. at 23. In any event, “proof that some minority candidates have been elected does not foreclose a [Section] 2 claim.” Gingles, 478 U.S. at 75; see Clark, 88 F.3d at 1397 (“‘[T]he election of a few minority candidates does not necessarily foreclose the possibility of dilution of the black vote.'” (citation omitted)).
Dr. Alford generally says black voters support white Democrats in equal measure with black Democrats, thus demonstrating that partisanship is the answer. Trial Tr. 1521:4-15. That statement is superficially true but fails to consider the record evidence explaining why black voters might choose white Democrats - like support for issues important to black citizens. And even Dr. Alford acknowledges that “it's possible for political affiliation to be motivated by race.” Trial Tr. 1537:3-6. Yet he never examined the political positions of the two state parties - or any candidates - to determine whether race factored into partisan voting. Trial Tr. 1504:9-25. Another defense expert, Dr. Brunell, has written that “the split between the political parties rests on a racial division.” Trial Tr. 1319:19:19-22. He also testified that there is “[n]o question” “that racial division . . . is still part of what's going on in politics today.” Trial Tr. 1320:5-8.
Dr. Alford also cites the 2012, 2016, and 2020 presidential elections, when voting percentages remained roughly the same for candidates Barack Obama, Hillary Clinton, and Joe Biden despite their differing races. DX-001 at 6-7. We mention that, though the 2016 Democratic Presidential ticket had no black nominee, Joe Biden's 2020 running mate was Kamala Harris, who was a black candidate. Additionally, Dr. Alford's analysis does not consider how race may factor into partisan voting. As we have explained in our discussion of Gingles preconditions two and three, the baseline white opposition to black-preferred candidates is so high - between 92 percent and 93.6 percent in these three elections, DX-1 at 6-7 - there is little room for additional white opposition to a black Democratic candidate.
LULAC v. Clements anticipated a similar issue, remaining “sensitive to the reality that political positions can be proxies for racial prejudice.” 999 F.2d at 879. While acknowledging this possibility, the Fifth Circuit saw no such concern “where white voters support black candidates of a particular party in larger percentage than they support white candidates of the same party.” Id. In addition, as noted, Senate Factor 2 is part of a “searching practical evaluation of the ‘past and present reality.'” Milligan, 599 U.S. at 19 (quoting Gingles, 478 U.S. at 79). Given the undisputed history of polarized voting in Mississippi - including recent history - Dr. Alford has not convincingly separated race from politics.
Context also matters. The opinions from the Plaintiffs' expert historian Dr. King need not be taken as correct in every detail, but in recent decades the Democratic Party has certainly been the one more closely associated with civil rights. The Defendants' expert Dr. Brunell agrees. He testified that racial division and the Democratic Party's association with civil rights “definitely played a role” in party realignment. Trial Tr. 1320:4.
That Mississippi voters have been separated by race even when most black voters were Republicans and white voters were Democrats adds weight to our finding that racially polarized voting best explains the divide - at least on this record. Dr. King testified “that the racial polarization throughout Mississippi political history precedes the partisan polarization. So the foundation of any polarization people see is the racial polarization that comes first, right. Race has always been the preeminent political issue in Mississippi and so that defines the dividing lines for the parties. So race comes first. That's the foundation for polarization.” Trial Tr. 764:21-765:3.
We do not hold - nor need we - that realignment was solely race-based or that every Republican is motivated by it. But the evidence is undisputed that race plays a role.
To conclude on Dr. Alford, he offered similar testimony in Robinson, where the court found his “opinions are unsupported by meaningful substantive analysis” and “border on ipse dixit.” 605 F.Supp.3d at 840. We share those concerns. While we accepted Dr. Alford as an expert and find that some of his opinions are plausible, he has not overcome Dr. Handley's testimony.
In short, we find that the Plaintiffs established racially polarized voting “on account of race or color” as those words are interpreted by the Senate Report and the Supreme Court. See Milligan, 599 U.S. at 25 (explaining that “it is patently clear that Congress has used the words ‘on account of race or color' in the Act to mean ‘with respect to' race or color, and not to connote any required purpose of racial discrimination” (quoting Gingles, 478 U.S. at 71 n.34). The Defendants' evidence fails to show that “partisan affiliation, not race, best explains the divergent voting patterns among minority and white citizens.” LULAC v. Clements, 999 F.2d at 850.
As in Singleton, when “we look deeper” into the race-versus-politics issue “we are looking at very little evidence.” 582 F.Supp.3d at 1018-19. Although we acknowledge that partisanship plays a role, we agree with Dr. Brunell that there is “[n]o question” “that racial division . . . is still part of what's going on in politics today.” Trial Tr. 1320:5-8. The Defendants' proof does not outweigh the Plaintiffs' evidence on this factor, and it certainly would not make this one of those “very unusual case[s]” where no violation is found despite establishing the Gingles preconditions. Clark, 21 F.3d at 97 (citation omitted). Senate Factor 2 weighs in favor of the Plaintiffs.
c. Senate Factor 5
This factor considers the extent to which members of the minority group still bear the effects of discrimination in areas such as education, employment, and health, and whether as a result they are hindered in their ability to participate effectively in the political process.
Much of the relevant evidence was offered by Dr. Orey, a professor in the Department of Political Science at Jackson State University. Trial Tr. 492:13-19. Dr. Orey has both teaching and publication experience in areas directly linked with race and its effect on political behavior. Trial Tr. 495:6- 16. He teaches multiple undergraduate- and graduate-level classes regarding minority politics, black voters in the political system, and applicable research methods. Trial Tr. 498:1-16. Dr. Orey also has experience conducting voter-turnout analyses using the EI RxC methodology, conducting descriptive statistical analyses for Senate Factor 5, and testifying as an expert in numerous court cases. Trial Tr. 495:25-496:3, 503:4-17, 504:2-16. Based on these qualifications, we accepted Dr. Orey as an expert in political science, political participation and behavior, and race and politics. Trial Tr. 506:8- 15.
The Plaintiffs offered Dr. Orey's testimony specifically to explain the “social and economic indicators [that] had an impact on voting amongst African-Americans” and how those “indicators negatively impacted turnout amongst African-Americans.” Trial Tr. 506:21-24. To conduct this Senate Factor 5 analysis, Dr. Orey used descriptive statistics like poverty, education access, turnout data, and EI RxC data to determine whether black citizens had the ability to effectively participate in Mississippi politics. Trial Tr. 504:23-505:13. These methodologies and data sources Dr. Orey used in his analysis are ones commonly relied upon among experts in political science. Trial Tr. 505:14-24.
Dr. Orey started his analysis with a theoretical framework regarding the resources required for Mississippians to participate in the political process. PTX-008, 4. We find persuasive his determination that voting and political participation has economic costs such that whether individual voters participate is influenced by financial resources, leisure time, and education. Dr. Orey further concluded black voters face larger disparities than white voters in areas such as education and income, affecting both the relative likelihood that they have the resources that are needed to vote and the ability to participate in the voting process. PTX-008, 3-11.
We find Dr. Orey is correct that black Mississippians suffer socioeconomic disparities that impair their ability to participate in the political process. Black Mississippians are significantly worse off in terms of income, poverty, unemployment, educational attainment, internet access, vehicle ownership, and health-insurance coverage. We accept as accurate the evidence that about 31 percent of black Mississippians live below the poverty line as compared to about 11.5 percent of white Mississippians. PTX-008, 5. Dr. Orey testified, based on political-science literature and his own regression analysis, that there is a strong correlation between financial status and voter turnout. Trial Tr. 512:10-15, 544:17-546:23. The analysis he conducted supports that income and poverty have been significant factors influencing voter participation, generally and specifically in Mississippi. PTX-008, 27-28.
Dr. Orey quantified racial disparities in Mississippi regarding the level of education as establishing black citizens have long been segregated from attaining the needed education for coherent political participation. PTX-008, 9-12. About 10.3 percent of white Mississippians did not complete high school, compared to 17.9 percent of black Mississippians. PTX-008, 8. As to college degrees, 28.5 percent of white Mississippians have a bachelor's degree or higher, compared to 18.2 percent of black Mississippians. PTX-008, 8.
Dr. Orey testified that these educational disparities can be traced to the long history of both de jure and de facto racial segregation in Mississippi. Residential patterns and the quality of education located in certain Mississippi locales, among other things, have resulted in many school systems being as segregated today as they were decades ago. PTX-008, 9; see Trial Tr. 512-514. Numerous negative effects of segregation in housing and education were also identified by Dr. Orey. For example, because of the racial economic disparities and resulting variations in local tax bases, residential segregation results in lower-funded black schools compared to predominantly white ones. PTX-008, 9.
Dr. Orey described a very strong, “well-established” correlation between educational attainment and voter turnout such that political-science literature renders education as a requirement in “virtually” any “analysis of voting behavior . . . model.” Trial Tr. 515:20-516:1. Education is “one of the resources that individuals need to vote” in order to “understand[] the issues” and “navigate the registration process[,]” for instance. Trial Tr. 515:7-19. We accept as fact that there has been a historical pattern, based on all these factors and perhaps others, that black citizens of Mississippi have participated in the political process in lower percentages than white citizens.
The factual issue for us is the current effect of these conditions on black-voter turnout. The disparities continue to exist, but are black Mississippians currently voting in lower percentages than whites because of the effects of discrimination? If not, then any past hindrance caused by those conditions to black citizens' ability to participate effectively in the political process has been overcome. Dr. Orey addressed this factual question by examining voter turnout only in the 2020 general election. Trial Tr. 579:3- 5. The Defendants' expert Dr. Brunell said it was important to examine trends over time and one election was inadequate. Trial Tr. 1258:15-1259:5. We find that examining multiple elections would give a clearer picture, but we acknowledge that Dr. Orey's analysis is only weakened, not discredited, by its consideration of solely 2020.
In his 2020 election analysis, Dr. Orey used three different, generally accepted methods: (1) ecological-inference analysis based on precinct-level election results and United States Census racial demographic data; (2) Bayesian Improved Surname Geocoding (“BISG”) of the Mississippi Secretary of State's full voter database (which includes voter history); and (3) reviewing estimates from the Cooperative Election Study (“CES”), a survey where voter turnout behavior is independently validated to eliminate the known problem of overreporting voting behavior in polls and surveys. PTX-008, 22-25. Dr. Orey testified that each of these methods showed a significant gap in turnout between black and white Mississippians. Trial Tr. 642:4-17.
Dr. Orey's EI RxC estimate was that white turnout in 2020 was 65.84 percent, while black turnout was 56.03 percent. PTX-008, 25. The Defendants' response does not significantly challenge the validity of the EI RxC analysis itself but argues this estimate is unreliable because it is based on data from one federal election for offices that are not the subject of this suit. Doc [219], 73. The Defendants' arguments are legitimate, but we find the estimates are still sufficiently reliable for our consideration.
Dr. Orey then used BISG to estimate the racial composition of the Mississippi voter file. BISG is an algorithmic method used to predict a person's race or ethnicity based on their last name and where they live on a map. PTX-008, 24. Applying BISG to a copy of the Mississippi voter file, which contains voter names, addresses, and voting history from 2020 and prior elections, Dr. Orey estimated a 69.7 percent turnout rate for white Mississippi voters, compared to 57.3 percent for black voters. PTX-008, 25.
The Defendants argue Dr. Orey's BISG analysis uses an incomplete data set, in part because the contents of the state voter file are so far removed from the 2020 election. Doc [219], 73-74. Dr. Orey used information from the file as it existed in June 2022 to infer individuals' turnout in the November 2020 general election. The file is continuously updated, and as voters move and new information is recorded, the file reflects each change. Dr. Orey acknowledged he had not previously performed a BISG analysis but stood behind the results despite there being a 5 percent loss in data. Trial Tr. 583:23-585:10.
Here, too, we do not refuse to consider the analysis. We find the questions about the accuracy of the data Dr. Orey used for his analysis to be legitimate but not dispositive.
Finally, Dr. Orey examined turnout by race estimates using the CES, a 50,000-plus person national stratified sample survey administered by the polling firm YouGov. PTX-008, 24-25. The CES dataset includes “validated vote” information, representing an independent authentication of whether a survey respondent voted, which Dr. Orey used for his analysis. Trial Tr. 532:24-533:15. Among registered voters, the CES's validated vote-turnout estimate was 72.5 percent for black Mississippians in 2020, compared to 86.8 percent turnout for white Mississippians. PTX-008, 26. Among all adults, the validated vote-turnout rates estimated by the CES for 2020 were 46.1 percent for black Mississippians and 59.6 percent turnout for white Mississippians. PTX-008, 26 n.59. Dr. Orey used the weighting variables corresponding to turnout among registered voters and among all adults to confirm his observation that a racial gap in turnout existed across both measures. Trial Tr. 535:22-536:24.
The Defendants identified some errors in the computations, including voters that should not have been included. The central criticism of these last calculations by the Defendants' expert Dr. Brunell is in evaluating the statistical significance of the figures. Dr. Brunell testified that 0.05 p-value (probability value) - or a 95 percent confidence level - is typically the lowest level of statistical significance that is used to determine the reliability of the estimates. Dr. Orey admitted that a p-value of less than 0.05 means that one can be 95 percent confident in the estimates. He testified that this level is the “more stringently and more commonly” used threshold for tests of statistical significance. Dr. Orey himself utilized the 95 percent threshold in some of his analysis.
Dr. Orey acknowledged, however, that his analysis did not quite reach a p-value of less than 0.05. By his calculations, the statistical significance level of the racial-turnout gap among all adults in Mississippi (excluding noncitizens) was 0.058, meaning that one can be “94 percent sure that this is the correct estimate,” which is a high level of confidence. Trial Tr. 540:2-9. The confidence level looking at the racial-turnout gap among registered voters was similar at 0.055. Trial Tr. 543:8-15.
Dr. Orey explained “there's nothing definitive about [0].05,” yet Dr. Brunell said there is. Dr. Brunell nonetheless agreed that choosing any particular threshold as a cutoff was “arbitrary” or “artificial,” and that social scientists may choose different thresholds. Another of the Plaintiffs' experts, Dr. Ragusa, supported this statement and explained that for statistical significance, “a p-value of less than .1, but not less than .05, is one of [the] options” and “more people, in my judgment, are using a p-value of less than [0].1 as a minimally significant result.”
Disputes among the experts about statistical significance have arisen in other cases. We accept the view of a Ninth Circuit opinion that insofar as admissibility is concerned, “[a]s a general matter, so long as the evidence is relevant and the methods employed are sound, neither the usefulness nor the strength of statistical proof determines admissibility under Rule 702.” Obrey v. Johnson, 400 F.3d 691, 696 (9th Cir. 2005). This evidence was properly admitted. Once admitted, how we as factfinders consider the evidence is the issue. The Seventh Circuit was particularly dismissive of the 95 percent standard:
Litigation generally is not fussy about evidence; much eyewitness and other nonquantitative evidence is subject to
significant possibility of error, yet no effort is made to exclude it if it doesn't satisfy some counterpart to the 5 percent significance test.Kadas v. MCI Systemhouse Corp., 255 F.3d 359, 362 (7th Cir. 2001).
We need not be nearly so loose in our standards as that quote, but we should consider evidence that the expert acknowledges fall short of the 95 percent confidence level, but not too far, and that the expert explains as statistically significant. We find questions about the accuracy of the data Dr. Orey used for his analysis to be legitimate but not dispositive.
To respond to Dr. Orey's analysis, the Defendants presented a recently released March 2024 BISG study from the Brennan Center for Justice at NYU Law School, entitled “Growing Racial Disparities in Voter Turnout, 2008-2022.” Trial Tr. 1272:10-15. No party had been aware of the study when the eight-day trial began, and it was introduced only for identification. Trial Tr. 1282:16-17, 1283:12-15. In fact, the study states it was published on March 2, 2024, the Saturday at the end of the first week of trial.
The study showed a narrow gap in voter turnout between black and white Mississippians. Most importantly, the study found that in 2020 there was a slightly higher-percentage turnout among blacks than among whites. Trial Tr. 1286:13-19. Dr. Brunell explained that the premise of the study was to show that the voter-turnout gap had widened post-Shelby County, particularly in states previously covered by Section 5. Trial Tr. 1285:1-9. The study found that the voter-turnout gap reversed only in Mississippi, with blacks turning out in a higher percentage than whites in 2020 and slightly higher in 2022. Trial Tr. 1286:21-1287:6.
The Plaintiffs objected to the survey, claiming that it was not disclosed by the expert-designation deadline and was not sufficiently reliable under Federal Rule of Evidence 702. We took those objections under advisement and now overrule them. Starting with the disclosure, we consider the missed deadline under the standard set forth in Hamburger v. State Farm Mutual Automobile Insurance, 361 F.3d 875, 883 (5th Cir. 2004). Under that standard, good cause existed for the late disclosure because the survey was first published after the expert-disclosure deadline and just four days before the Defendants offered it. The Defendants also provided a copy to the Plaintiffs once they discovered the survey and before mentioning it in court. Any prejudice caused by the late disclosure was addressed when the court recessed the case to allow the Plaintiffs to depose Dr. Brunell about the report. That approach also eliminated the need for a longer continuance. Although the importance of the testimony is minimal, this factor does not outweigh the others, which favor admission.
As for Rule 702, Dr. Brunell testified that the Brennan Center is a respected institution and that surveys like this are the kind he would rely on in his normal work. Trial Tr. 1277:11-25. “As a general rule, questions relating to the bases and sources of an expert's opinion affect the weight to be assigned that opinion rather than its admissibility and should be left for the jury's consideration.” United States v. 14.38 Acres of Land, More or Less Situated in Leflore County, 80 F.3d 1074, 1077 (5th Cir. 1996) (quoting Viterbo v. Dow Chem. Co., 826 F.2d 420, 422 (5th Cir. 1987)). We find that Dr. Brunell's testimony regarding the Brennan Center survey was admissible, and the reliability of the survey is a question of weight.
We accept that survey into evidence. Nonetheless, the parties had almost no time to explore the details of the survey nor to consider through their own experts its possible flaws. As we understand the survey report, it was not peer reviewed. Trial Tr. 1387:23-1388:7. Dr. Brunell stated he had not previously heard of its authors, though we do not know the significance of such lack of fame. Trial Tr. 1388:11-20. We accept that the Brennan Center is a respected institution and that its reports are not amateurish. Nonetheless, we know too little about this particular report to have it outweigh other evidence in the case.
The Defendants submitted evidence from the U.S. Census Bureau and the U.S. Department of Labor's Current Population Survey (“CPS”). The Defendants emphasize the Supreme Court's explicit reliance on CPS data in Shelby County as evidence of its trustworthiness. The Court held that conditions that once supported requiring only some states but not others to preclear all changes to voting practices with the Department of Justice no longer existed. See Shelby County, 570 U.S. at 554. Starting with the adoption of the Voting Rights Act in 1965, “Census Bureau data indicate that African- American voter turnout has come to exceed white voter turnout in five of the six States originally covered by [Section] 5.” Id. at 535. As the Court mentioned, a chart using the CPS data was placed in both the Senate and House reports on the bill that reauthorized the Voting Rights Act in 2006 and was also reproduced in the opinion. Id. at 548 (citing S. Rep. No. 109-295, at 11 (2006); H.R. Rep. No. 109-478, at 12 (2006)).
The relevant part of the CPS survey is conducted every two years. Voters are asked questions about the election, specifically whether they voted. The CPS shows insignificant differences between black and white turnout either among registered voters or among the voting-age population in Mississippi. Trial Tr. 560:19-562:11. Dr. Orey disagreed, testifying that black turnout remains lower than white turnout. Dr. Orey's criticism of the CPS survey focuses on how the data is collected. He testified that “the literature shows that African-Americans are more likely to overreport” voting “because of racial identity” and a sense that “what happens to other Blacks impacts one's individual life,” which leads to increased pressure to over-report. Trial Tr. 531:15-23.
While Dr. Orey raises facially plausible problems with CPS data, we are not necessarily persuaded by this testimony. We acknowledge the Supreme Court's Shelby County opinions contain no discussion of arguments that the CPS data is unreliable. If the argument was not raised, there was no need for the Court to consider possible inaccuracies. The Supreme Court found the data sufficient to discuss, but it is unnecessary to resolve any potential statistical issue to decide this case. Dr. Orey's arguments are reasonable, but we find the evidence supports, at the very least, that whatever gap existed in the turnout in Mississippi in 1965 is greatly reduced today.
In addition to his analysis of voter turnout by race, Dr. Orey also conducted a regression analysis to examine the extent to which black turnout in Mississippi was in fact driven by the socioeconomic markers discussed previously. PTX-008, 26-27. Dr. Orey used “data from the voter file” and “aggregated the data” along with “census data at the [bloc] level” and, in doing so, he was “able to examine whether or not the turnout amongst Blacks was a function of some of these indicators.” Trial Tr. 544:12-546:23.
This factor required us to consider the extent past discrimination affected black voters' ability to participate in Mississippi's political process today. Teague, 92 F.3d at 292. We credit Dr. Orey's conclusion that voting and political participation is influenced by financial resources, education, income, and unemployment. We find that black Mississippians are worse off than white voters in terms of these factors. The percentage of black voters who live below the poverty line is more than twice that of white voters, and they are less likely to graduate high school and college where they would have access to the required education needed to navigate the political process.
Irrespective of the size or even existence of turnout discrepancies, we find that the record establishes black Mississippians' ability to participate effectively in Mississippi politics is hindered by racial gaps in education access, financial status, and health. Senate Factor 5 thus weighs in favor of the Plaintiffs.
d. Senate Factor 6
This factor considers the use of “overt or subtle racial appeals” in political campaigns. Gingles, 478 U.S. at 37. The Plaintiffs' evidence identified other elections, but we will discuss only those elections the Plaintiffs identified in their proposed findings of fact regarding this factor.
In a Republican primary contest, one candidate's 30-second video advertisement briefly showed the candidate speaking on a stage next to a furled Confederate flag. That was an isolated incident, but we accept it was seeking white-voter support.
In a general election between a white Republican senator and a black challenger, the state Republican Party issued a campaign flyer focusing on the fact that the Democratic candidate “was forced to resign after being indicted in Washington, D.C. as U.S. Agriculture Secretary.” The flyer did not acknowledge that he had been acquitted following a trial. Though perhaps unfair, a political party's highlighting that the other party's candidate was indicted for a crime, regardless of the candidate's race and even if there had been an acquittal or no trial as of yet, is neither surprising nor a racial appeal.
Two other examples in evidence are of a white congressional candidate in 1982 and a white supreme court candidate in 2004 who each identified himself as “one of us” when running against a black man. Another three-judge district court categorized the congressional candidate's slogan as a racial appeal. See Jordan, 604 F.Supp. at 813 & n.8. Similarly appealing to white voters, a sitting state representative in 2015 urged “voters to vote against a ballot initiative because, if it passed, it would allow a Black judge to decide what happens with public schools.” Trial Tr. 781:14-19.
The Plaintiffs also introduced evidence of how some black candidates were exposed to the racial animosities of some voters. That evidence did not reveal official discrimination under Senate Factor 1, nor racial appeals by opposing candidates under Senate Factor 6. Nonetheless, if white voters are insulting or even threatening black candidates, that is relevant to understanding the totality of the circumstances. Because a consideration of this evidence is supported by the Supreme Court's observation in Gingles that the “list of typical factors is neither comprehensive nor exclusive,” see Gingles, 478 U.S. at 45 (citing S. Rep. No. 97-417, at 29-30), we conclude that, if there have been threats to black candidates because of their race, it is a relevant consideration.
One relevant witness on this additional consideration was Pamela Hamner. She was questioned about her campaign in DeSoto County, having been a candidate both in 2021 for the Board of Aldermen and in 2023 for the state senate.
Q. I know you mentioned earlier that you campaigned in Horn Lake, did you campaign in Hernando?
A. No -- very little. Hernando, I -- so when I ran in '21 I had the police called on me, and people in my party, some of the men, they told me don't go out by myself, which I'm glad I didn't, so I didn't spend a lot of time in Hernando, I -- and my canvassers had the police called on them this time too.Trial Tr. 720:15-22.
Hamner described Hernando as “the county seat [that is] an older historic area, too, but it's . . . more rural and it's more White. . . . Hernando is predominantly White.” Trial Tr. 719:10-19. Hamner's trial testimony implies she and her party members were treated this way because she was a black candidate campaigning in a predominantly white area.
Another witness was Terry Rogers. At age 18, he ran for the statewide office of Agriculture Commissioner in 2023. He testified that while campaigning at a county fair that is the state's largest political forum, he saw Confederate flags. While he was giving a speech there, someone in the audience held up a cell phone so that Rogers could see a picture of a dog with a noose around its neck.
The evidence supports that some candidates continue to make racial appeals. We find that Senate Factor 6 weighs in favor of the Plaintiffs. Other candidates occasionally encounter offensive or threatening actions by voters due to their race. We also find that the limited evidence of such conduct to add slight weight in favor of the Plaintiffs.
e. Senate Factor 7
This factor considers “the extent to which members of the minority group have been elected to public office in the jurisdiction.” Gingles, 478 U.S. at 37. Senate Factor 7 is a particularly important factor in the totality-of-the-circumstances analysis. The success of black candidates is far from negligible. As stated by Plaintiffs' expert Dr. Orey, Mississippi is among the states with the highest number of black elected officials.
The evidence shows, though, that in order for black-preferred candidates to be elected for the legislature or for Congress, they almost always have to be running in majority-black districts. The evidence further showed there has not been a black candidate elected to statewide office since the end of Reconstruction 150 years ago. In recent elections, the Democratic Party has had some black nominees for statewide office, including for the United States Senate and Governor, but all have been defeated.
Because a black-majority district is a virtual prerequisite for black candidates' success in Mississippi politics, it is relevant that the Mississippi Legislature left the number of majority-black districts unchanged following the recent Census, despite substantial increases in the black population and corresponding losses in the white population. There was testimony that the number of black legislators today is disproportionate to their percentage of the state's population. We do not overlook that the Voting Rights Act states “[t]hat nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.” 52 U.S.C. § 10301(b). Indeed, the Supreme Court concluded in its recent decision regarding Alabama congressional districts that proportional representation can rarely be achieved because of diffusion of the relevant population. Milligan, 599 U.S. at 28-29. Nonetheless, proportionality “is a relevant fact in the totality of the circumstances to be analyzed when determining whether members of a minority group have ‘less opportunity . . . to participate in the political process.'” De Grandy, 512 U.S. at 1000 (citation omitted). The Court later restated that relevance in relation to majority-minority districts: “Another relevant consideration is whether the number of [black-majority] districts . . . is roughly proportional to [the black] population in the relevant area.” LULAC v. Perry, 548 U.S. at 426.
Balancing the substantial success for black-preferred candidates with its currently unbreachable limits, we conclude this factor, if not neutral, slightly favors the Plaintiffs.
f. Senate Factor 8
This factor considers whether “elected officials are unresponsive to the [minority's] particularized needs.” Gingles, 478 U.S. at 45. The Plaintiffs seek to show a lack of responsiveness on the part of the white Republican Mississippi Legislature with evidence of such matters as a failure to expand Medicaid, not adequately funding education, and the decision to enact the legislative-redistricting plan being challenged in this case.
We find, based on testimony, that Medicaid expansion is important to the black community. Trial Tr. 701:12-24, 726:14-727:2. The unrebutted testimony is that the failure to expand Medicaid disproportionately harms black communities, particularly those in the Mississippi Delta where regional hospitals are in financial struggles. Trial Tr. 519:4-15. Dr. Orey testified that black Mississippians “will more likely depend on Medicaid relative to whites” and face “vast differences” in health coverage. Trial Tr. 518:4-19.
Regarding education funding, the Plaintiffs provided two expert opinions, and we find both credible. Dr. Orey and Dr. Luckett concluded that Mississippi's successes with respect to education have been unevenly distributed, with predominantly black schools being underfunded and underperforming as compared to predominantly white schools. PTX-008, 8-12; PTX-007, 48-50.
Other witnesses, such as Dr. Joseph Wesley, Deacon Kenneth Harris, and Terry Rogers, testified that white legislators do not campaign in black communities or attend events hosted by black civic organizations. Dr. Wesley testified that, during his role as Political Action Chair for the Forrest County Branch of the NAACP, the Branch has held numerous community forums - including for statewide candidates - open to all candidates, and yet his elected senator has never attended or otherwise engaged with the black community in the district. Finally, Pamela Hamner testified that her senator was not responsive to the needs and concerns of black voters in her area. The testimony of these witnesses was unrebutted, and we credit it.
We mention that funding is a component of the decision that the Plaintiffs say should be made. Funding based on a set budget is a zero-sum calculation. Whether there is money for full funding of education and to expand Medicaid, even though Medicaid expansion has significant federal contribution, and also pay for the other significant work authorized by the Mississippi Legislature is not for this court to decide. We mention as well that the Defendants did not offer any testimony that could have added context for the decisions.
Under this factor, we analyze whether there are circumstances in which Mississippi has failed to respond to the needs of black voters. Teague, 92 F.3d at 292. The Defendants are quite correct that the state government is not running roughshod over the interests of those in the minority. Although both chambers of the Mississippi Legislature are controlled by Republicans, a significant number of Democrats chair committees in both chambers, having been appointed by Republican leadership. Doc [199], 17- 18. Even its redistricting decision served the minority's interests when the majority agreed to unpair two incumbent Democrats such that the new plan would not require them to run against each other.
In sum, the Plaintiffs' examples of a lack of responsiveness include shortfalls in funding for education and the failure to expand Medicaid, the redistricting plan itself, and the failure of some white legislators to participate in black-community events. We find Senate Factor 8 favors the Plaintiffs.
g. Senate Factor 9
This factor considers “whether the policy underlying the state or political subdivision's use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.” Gingles, 478 U.S. at 37. This factor requires a consideration of the policies used by Mississippi to justify its districting decisions in this case. Teague, 92 F.3d at 292. The Plaintiffs do not contest that the legislature followed its policy of creating districts that were sufficiently equal in population for each chamber and consisted of contiguous tracts. The legislature's policies also require that the districts comply with all state and federal laws relating to redistricting, that the districts be compact and minimize county and precinct splits, and that the design of the districts consider “political performance.”
The Defendants described the basic criteria the Standing Joint Committee was required to follow to avoid unconstitutional race-based redistricting and to retain the cores of the existing districts: One Person, One Vote; contiguity of the districts; and compliance with all state and federal laws. JTX-010, 8-14. Those are legitimate, non-tenuous policies, but how they were applied needs to be explained. We accept that some limited packing or cracking of minority populations in order to protect a white incumbent is a non-tenuous reason. All we have in that regard are quite general pronouncements made by the chair of the committee in each chamber when explaining the relevant plan that assisting incumbents was a factor. There is no evidence of when such considerations led to the creation of a particular district and resulted in the failure to create a black-majority district in that same area. We needed more than generalized statements for this defense in light of the Plaintiffs' evidence of a Section 2 violation.
We find the legislature did not enact an egregiously flawed plan, the equivalent of a political gerrymander of squeezing the minority into as few districts as possible. Yet, we do not find any specific, non-tenuous justifications for why black-majority districts were not created in the three identified areas. We find, therefore, that this factor weighs in favor of the Plaintiffs.
h. Summary of the Senate Factors
We have detailed the law, evidence and our findings on all the Senate Factors except for the inapplicable Senate Factor 4. We found all clearly to favor the Plaintiffs with two exceptions. One exception was Senate Factor 6, concerning overt or subtle racial appeals election campaigns. We found it only slightly favors the Plaintiffs. We also found that Senate Factor 7, which considers the success of minority candidates in elections, either is neutral or only slightly favors the Plaintiffs.
Senate Factors 2 (racially polarized voting) and 7 are considered “the most important” factors in a totality-of-the-circumstances analysis. Gingles, 478 U.S. at 48 n.15. One of these most important factors clearly favors the Plaintiffs, while the other is perhaps almost neutral. As indicated, though, all other factors in the totality of circumstances clearly favor the Plaintiffs.
After engaging in the required “intense[] local appraisal” and “searching practical evaluation of the past and present reality” of Mississippi's political process, see Gingles, 478 U.S. at 79 (quotation marks and citation omitted), we conclude this is not one of the “only very unusual case[s]” where the Gingles preconditions are established, but there is no liability. Teague, 92 F.3d at 293 (quoting Clark, 21 F.3d at 97). The Plaintiffs instead met their burden under both the Gingles preconditions and the totality of the circumstances. We therefore find Mississippi's 2022 Enacted Plans violate Section 2 of the Voting Rights Act.
V. REMEDY
We have concluded that the Plaintiffs met their burden under the Gingles framework to establish that Mississippi's 2022 Enacted Plans violate Section 2 of the Voting Rights Act. The task now is to establish a remedy.
Once a Section 2 violation is found, the court faces “the difficult question of the proper remedial devices which federal courts should utilize in state legislative apportionment cases.” Reynolds v. Sims, 377 U.S. 533, 585 (1964) (Fourteenth Amendment case). One thing is clear though: the court's “first and foremost obligation” must be “to correct the Section 2 violation.” United States v. Brown, 561 F.3d 420, 435 (5th Cir. 2009) (citation omitted). “In doing so, the district court ‘should exercise its traditional equitable powers to fashion the relief so that it completely remedies the prior dilution of minority voting strength.'” Id. (quoting S. Rep. No. 97-417, at 31).
Addressing similar equities in the equal-protection context, the Supreme Court has noted that courts must “tak[e] account of what is necessary, what is fair, and what is workable.” North Carolina v. Covington, 581 U.S. 486, 488 (2017) (per curiam) (quoting Reynolds, 377 U.S. at 585). Though neither Covington nor Reynolds are Section 2 cases, the Fifth Circuit has relied on Reynolds and other equal-protection cases when discussing Voting Rights Act remedies. See Veasey v. Abbott, 830 F.3d 216, 270 (5th Cir. 2016). The Fifth Circuit also holds that “any remedy ‘should be sufficiently tailored to the circumstances giving rise to the Section 2 violation,' . . . and to the extent possible, courts should respect a legislature's policy objectives when crafting a remedy.” Id. at 269 (quoting Brown, 561 F.3d at 435) (other citations omitted).
While the proper remedy depends on the specific circumstances of each case, “it would be the unusual case in which a court would be justified in not taking appropriate action to [e]nsure that no further elections are conducted under the invalid plan.” Reynolds, 377 U.S. at 585; see also Veasey, 830 F.3d at 270 (applying Reynolds under Section 2). That is particularly true when elections are imminent, but here the next legislative election would not happen until 2027. Therefore, the question is whether to order a special election to remedy the Section 2 violation found.
The Defendants urge us not to order any special elections, relying on the equitable considerations found in Covington: “[1] the severity and nature of the particular . . . violation, [2] the extent of the likely disruption to the ordinary processes of governance if early elections are imposed, and [3] the need to act with proper judicial restraint when intruding on state sovereignty.” Def. Findings ¶ 421 (quoting Covington, 581 U.S. at 488). We accept the Defendants' invitation to use this test and find that these same equities are relevant under Section 2.
Starting with the first consideration, the Defendants argue the violations are not severe because the Plaintiffs “have not proved any violation.” Def. Findings ¶ 422. We disagree with the premise of this argument - that there are no violations - and the Defendants give us nothing further in deciding how to determine severity. Even so, the severity here is greater than in Covington. There, the Supreme Court unanimously vacated a district-court decision that ordered special elections in 28 North Carolina legislative districts in 2017 that would allow the winners to serve only one year until the next regular legislative elections. Covington, 581 U.S. at 487. The Supreme Court concluded that the district court had given only cursory consideration to the balance of equities before ordering the elections. Id. at 488-89.
The violations in Covington were more numerous than those proved here, but the length of the remaining Mississippi terms is three times as long. The legislators elected under the Enacted Plans have served, at this point, a little more than six months of their four-year terms. Thus, if left as is, black voters in each affected district will be served for a full term by a legislator chosen in an election that diluted black votes. The harm is localized, but it is severe to the affected voters. This is the exact kind of injury that warrants a remedy.
The extent of any likely disruption to the ordinary governmental processes is the next factor to consider. In Section 2 litigation, requiring a special session of a legislature to redraw district lines is a common remedy. See, e.g., Robinson, 86 F.4th at 600-01. Nonetheless, such special sessions are a disruption of the ordinary legislative process and should be ordered only if it is equitable to do so.
One equitable consideration in deciding the proper timing of a remedy is whether the Plaintiffs acted quickly to assert their rights. By the end of March 2022, Mississippi approved redistricting plans for both houses of the legislature. The Plaintiffs did not sue until December 20, 2022. As they state, however, other Section 2 cases were stayed during this period while the Supreme Court considered Milligan. See Nairne v. Ardoin, No. CV 22-178-SDD-SDJ, 2022 WL 3756195, at *1 (M.D. La. Aug. 30, 2022) (staying proceedings until June 2023). Further, despite some initial delay in challenging the Enacted Plans, the Plaintiffs sought an expedited pretrial schedule - which the Defendants opposed - and required no extensions under that schedule. We find that any initial delays are insufficient to make a special election inequitable, nor would they justify not having a special session, which would be unfortunate but likely limited in time.
The final factor is that the court must exercise restraint before intruding onto state sovereignty. In addition to our finding violations, we also find that allowing the violations to go unaddressed for the entire four-year term of affected legislators, when only a little more than six months have been served at this date, is not an equitable result. We therefore need to order some redistricting to rectify the Section 2 violations.
In doing so, we know that “reapportionment is primarily the duty and responsibility of the State through its legislature or other body, rather than of a federal court.” Voinovich v. Quilter, 507 U.S. 146, 156 (1993) (citations omitted). Thus, “in redistricting cases, district courts must offer governing bodies the first pass at devising a remedy.” Brown, 561 F.3d at 435 (citation omitted).
We will do that here with this guidance. Three of the illustrative districts satisfy all three Gingles preconditions: Illustrative Senate Districts 2 and 9 and Illustrative House District 22. Our determination that these specific illustrative districts support a Section 2 violation does not require the State to draw districts as proposed in the Plaintiffs' remedial plans. See Vera, 517 U.S. at 978. The State has discretion in determining how best to remove the violation, subject to further judicial review. See Id. The Plaintiffs' proposed remedy is to require districts be drawn “in which Black voters have an opportunity to elect candidates of their choice in the areas in and around” the relevant illustrative districts. Pls.' Proposed Findings ¶ 781. We agree that would satisfy the State's obligations, but the State has discretion on how to proceed with the remedy.
Because we have concluded that only three of the illustrative districts identified by the Plaintiffs satisfy the three Gingles preconditions, there is no obligation to redraw districts in the areas of the other four illustrative districts. If the Mississippi Legislature creates three new majority-minority districts, however, there will likely be a need to revise other districts to make that possible. We find that the equitable factors identified in Covington allow the State to limit the ripple effect of creating new majority-minority districts as much as reasonably possible. Special elections will need to be called for all revised districts, and minimizing the number of legislative seats subject to election for a briefer-than-usual term is a significant State interest.
So, how much time is needed for a special election? As mentioned before, we are to “afford a reasonable opportunity for the legislature” to create and adopt its own constitutional plan. Wise v. Lipscomb, 437 U.S. 535, 540 (1978). Because we will require elections for any districts the Mississippi Legislature alters in response to this court's ruling, relevant dates are these: (1) for the Mississippi Legislature to adopt new maps; (2) for this court to be presented either with objections to new maps drawn by the Mississippi Legislature or with maps proposed by the parties if the Mississippi Legislature does not act; and (3) for elections to be held.
After trial, we asked the parties to submit their arguments about the necessary timing of new elections for some legislative districts if we found a need to order elections. We have found the need, but the parties' submissions are now outdated. It is the desire of this court to have new legislators elected before the 2025 legislative session convenes, but the parties can make whatever arguments about timing they conclude are valid. We state now that the parties should be prepared to present their own respective maps five days after the deadline for the Mississippi Legislature to adopt its own plan, thereby being able to present alternatives almost immediately should the Mississippi Legislature not act.
In order for the court to receive the arguments of counsel on the issue of timing for the steps that must be taken, we will conduct a video conferencing hearing on Monday, July 8, at 2:00 p.m.; Tuesday, July 9, at 2:00 p.m.; or Thursday, July 11, at 1:00 p.m. Counsel will be contacted by the court to determine the most appropriate date.
It is SO ORDERED.