Bethune-Hillv.Va. State Bd. of Elections

UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF VIRGINIA Richmond DivisionOct 22, 2015
141 F. Supp. 3d 505 (E.D. Va. 2015)

Civil Action No. 3:14cv852.

10-22-2015

Golden BETHUNE–HILL, Plaintiffs, v. VIRGINIA STATE BOARD OF ELECTIONS, et al., Defendants.

John Kuropatkin Roche, Aria Christine Branch, Bruce Van Spiva, Pro Hac, Vice, Elisabeth Carmel Frost, Pro Hac, Vice, Marc Erik Elias, Perkins Coie LLP, Washington, DC, Abha Khanna, Pro Hac, Vice, Kevin John Hamilton, Ryan Spear, Pro Hac, Vice, William Benjamin Stafford, Pro Hac, Vice, Perkins Coie LLP, Seattle, WA, for Plaintiffs. Jeffrey P. Brundage, Daniel Ari Glass, Pro Hac, Vice, Kathleen Angell Gallagher, Pro Hac, Vice, Eckert Seamans Cherin & Mellott LLC, Washington, DC, Anthony F. Troy, Eckert Seamans Cherin & Mellott LLC, Godfrey Thadeus Pinn, Jr., Harrell & Chambiss LLP, Richmond, VA, for Defendants.


John Kuropatkin Roche, Aria Christine Branch, Bruce Van Spiva, Pro Hac, Vice, Elisabeth Carmel Frost, Pro Hac, Vice, Marc Erik Elias, Perkins Coie LLP, Washington, DC, Abha Khanna, Pro Hac, Vice, Kevin John Hamilton, Ryan Spear, Pro Hac, Vice, William Benjamin Stafford, Pro Hac, Vice, Perkins Coie LLP, Seattle, WA, for Plaintiffs.

Jeffrey P. Brundage, Daniel Ari Glass, Pro Hac, Vice, Kathleen Angell Gallagher, Pro Hac, Vice, Eckert Seamans Cherin & Mellott LLC, Washington, DC, Anthony F. Troy, Eckert Seamans Cherin & Mellott LLC, Godfrey Thadeus Pinn, Jr., Harrell & Chambiss LLP, Richmond, VA, for Defendants.

MEMORANDUM OPINION

ROBERT E. PAYNE, Senior District Judge:

This case challenges the constitutionality of twelve Virginia House of Delegates districts (the “Challenged Districts”) as racial gerrymanders in violation of the Equal Protection Clause of the Fourteenth Amendment to the Constitution of the United States. The case is ripe for decision following a four-day bench trial at which the parties presented oral testimony and offered numerous exhibits. Our findings of fact are based on our assessment of the record and are grounded in our determinations respecting the credibility of the witnesses.

Our conclusions of law address the several legal issues presented by the parties. In particular, we have determined that it is the burden of the Plaintiffs to prove by a preponderance of the evidence that race was the predominate factor motivating the decision to place a significant number of voters within or without a particular district in that, as to each of those districts, Virginia's General Assembly subordinated race-neutral districting principles to racial considerations when forming the district. Based on this legal standard and the record, we have concluded that, except as to House District 75, the Plaintiffs have not carried that burden and that race was not shown to have been the predominant factor in the creation of eleven of the twelve Challenged Districts.

We are satisfied that race was the predominant factor in the creation of House District 75. However, we have also concluded that, in using race, the General Assembly was pursuing a compelling state interest, namely, actual compliance with federal antidiscrimination law, and that, in the process, the General Assembly used race in a manner narrowly tailored to achieve that interest.

In the Memorandum Opinion that follows, the Court will review the procedural background of the case in Section I; provide a brief overview of the law relating to racial gerrymandering claims in Section II; and set out its findings on the factual background of the case in Section III. In Section IV, the Court will articulate its understanding of the relevant legal framework for evaluating racial gerrymandering (or “racial sorting”) claims, set out additional factual findings of general applicability, and conduct a district-by-district analysis with district-specific factual findings and district-specific application of the relevant legal framework.

I. PROCEDURAL BACKGROUND

In the wake of the 2010 census, the Virginia General Assembly sought to redraw the legislative districts for the Virginia House of Delegates (“House”) and the Senate of Virginia (“Senate”). The task of redistricting is one that carries great political and legal consequence. In a representative democracy, such legislation shapes more than the abstract boundaries of electoral districts; it shapes the character, conduct, and culture of the representatives themselves. On its face, the legislation recites a singularly tedious list of precincts and counties. But in application, few pieces of legislation have a more profound impact on the function of government and whether it acts as “the faithful echo of the voices of the people.” Justice James Wilson, The Works of the Honourable James Wilson, L.L.D. 433 (Bird Wilson, ed., The Lorenzo Press 1804).

The political significance of redistricting is matched only by its legal complexity. Those shepherding redistricting legislation must traverse a precarious path between constitutional and statutory demands that are often in tension with one another and provide opaque interpretive standards rather than clear rules.

As to the 2011 redistricting, Delegate Chris Jones led this effort in the House. Delegate Jones played an instrumental role in the 2001 redistricting process and drew upon that experience to lead the 2011 redistricting efforts. Pls.' Ex. 35 at 46:18–48:21; Trial Tr. 272:24–274:7 (Jones). Because Virginia was a covered jurisdiction under Section 4 of the Voting Rights Act of 1965 (“VRA”) at the time the redistricting legislation was prepared, and was therefore subject to the requirements of Section 5 of the VRA, (Docket No. 83), it was necessary to ensure that the plan did not result in a “retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise.” Beer v. United States, 425 U.S. 130, 141, 96 S.Ct. 1357, 47 L.Ed.2d 629 (1976). In an attempt to comply with this statutory command, Delegate Jones crafted a plan containing twelve majority-minority House Districts (“HDs” or “Districts”). These are the Challenged Districts: HDs 63, 69, 70, 71, 74, 75, 77, 80, 89, 90, 92, and 95. On December 22, 2014, Plaintiffs filed a Complaint against the Virginia State Board of Elections, the Virginia Department of Elections, and various members thereof in their official capacities (“Defendants”), alleging that the Challenged Districts were racial gerrymanders in violation of the Equal Protection Clause of the Fourteenth Amendment and seeking declaratory and injunctive relief prohibiting Defendants from implementing or conducting further elections based on the Challenged Districts. (Docket No. 1.) The Plaintiffs are twelve citizens of the United States and the Commonwealth of Virginia who are lawfully registered voters in the Commonwealth and each of whom resides in one of the twelve Challenged Districts. (Docket No. 83.) The Plaintiffs requested that the case be heard by a three-judge district court pursuant to 28 U.S.C. § 2284(a) on the grounds that the action “challeng[es] the constitutionality of the apportionment of ... [a] statewide legislative body.” (Docket No. 1.) That request was granted by the Chief Judge of the United States Court of Appeals for the Fourth Circuit. (Docket No 11.)

See 52 U.S.C. § 10303(b) (formerly cited as 42 U.S.C. § 1973b(b)).

“Majority-minority” districts are those with a racial or ethnic minority population above 50% of the district's total population.

Plaintiffs filed a Corrected Amended Complaint on June 16, 2015 after one of the original plaintiffs changed residences. (Docket Nos. 66 & 71.)

The Virginia House of Delegates and the Virginia House of Delegates Speaker William Howell (“Intervenors”) moved to intervene in the case. (Docket No. 12.) That motion was granted. (Docket No. 26.)

A four-day bench trial began on July 7, 2015. (Docket Nos. 99–102.) Because the Defendants are “administrative agencies that implement elections” but “do not draw the districts,” Trial Tr. 12:14–25 (Defendants), the Defendants allowed the Intervenors to carry the burden of litigation but joined the Intervenors' arguments at the close of the case, id. at 830:2–3. For ease of reference, the Defendants and Intervenors will be referred to as the Intervenors.

II. BASIC OVERVIEW OF RACIAL GERRYMANDERING CLAIMS

Before proceeding to the facts of the case and the substance of this litigation, a brief overview of the constitutional and statutory requirements pertinent to racial gerrymandering claims is appropriate. As noted above, these commands often cut counter to each other and require legislators to balance competing considerations. Tracing their evolution is therefore useful as a predicate for the decision that follows.

The Supreme Court has long observed that the right to vote is “fundamental” because it is “preservative of all rights.” Yick Wo v. Hopkins, 118 U.S. 356, 370, 6 S.Ct. 1064, 30 L.Ed. 220 (1886). In Reynolds v. Sims, the Court recognized that “the right of suffrage can be denied by a debasement or dilution of the weight of a citizen's vote just as effectively as by wholly prohibiting the free exercise of the franchise” and held that the malapportionment of state legislative bodies in derogation of the “one person, one vote” principle violates the Equal Protection Clause. 377 U.S. 533, 555, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Because legislation affecting the right to vote “strike[s] at the heart of representative government,” id., the “Constitution leaves no room for classification of people in a way that unnecessarily abridges this right,” id. at 560, 84 S.Ct. 1362, and grants every citizen “an inalienable right to full and effective participation in the political processes of his State's legislative bodies,” id. at 564, 84 S.Ct. 1362.

The decision in Reynolds only required state legislatures to comply with the equal population standard, but its language would come to stand for something more. The next year, in Fortson v. Dorsey, the Court suggested that a “constituency apportionment scheme” may not “comport with the dictates of the Equal Protection Clause” if it “would operate to minimize or cancel out the voting strength of racial or political elements of the voting population.” 379 U.S. 433, 438–39, 85 S.Ct. 498, 13 L.Ed.2d 401 (1965). With Fortson, the Supreme Court first recognized that redistricting legislation may offend Equal Protection Clause principles when it distinguishes between voters on a racial basis.

Over time, the Supreme Court has come to recognize two types of racial gerrymandering claims under the Fourteenth Amendment: (1) claims of racial vote dilution, where the redistricting legislation is “conceived or operated as [a] purposeful devic[e] to further racial discrimination by minimizing, canceling out or diluting the voting strength of racial elements in the voting population,” Rogers v. Lodge, 458 U.S. 613, 617, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1982) (internal quotation marks omitted); and (2) claims of racial sorting, where the redistricting legislation, “though race neutral on its face, rationally cannot be understood as anything other than an effort to separate voters into different districts on the basis of race, and that the separation lacks sufficient justification,” Shaw v. Reno (Shaw I), 509 U.S. 630, 649, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993).

A. Racial Vote Dilution and the Fourteenth Amendment

The Supreme Court first struck down a districting scheme for unconstitutional racial vote dilution in White v. Regester, 412 U.S. 755, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973). There, the Court stated:

The plaintiffs' burden is to produce evidence to support findings that the political processes leading to nomination and election were not equally open to participation by the group in question—that its members had less opportunity than did other residents in the district to participate in the political processes and to elect legislators of their choice.

412 U.S. at 765–66, 93 S.Ct. 2332. At the time, it was unclear whether such a claim required a showing of discriminatory intent or could be maintained based solely on discriminatory effect.

Several years later, in City of Mobile v. Bolden, the Court suggested in a plurality opinion that both discriminatory intent and discriminatory effect were required to establish a claim of unconstitutional racial vote dilution. 446 U.S. 55, 66, 100 S.Ct. 1490, 64 L.Ed.2d 47 (1980). That holding was reaffirmed by a majority of the Court in Rogers v. Lodge, 458 U.S. 613, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1982). Writing for the majority, Justice White confirmed that “a showing of discriminatory intent has long been required in all types of equal protection cases charging racial discrimination.” Rogers, 458 U.S. at 617, 102 S.Ct. 3272.

Therefore, in a constitutional racial vote dilution case, the plaintiff must show that the State has placed a burden upon the right to vote by intentionally establishing or maintaining devices or procedures that cause minority citizens to have less opportunity than other citizens to participate in the political processes and to elect legislators of their choice. This dilutes the minority voter's ability to exercise the “full and effective” right to vote.

B. Racial Sorting and the Fourteenth Amendment

The other strand of “racial gerrymandering”—a racial sorting claim such as the one presented in this case—is “analytically distinct” from a vote dilution claim. Miller v. Johnson, 515 U.S. 900, 911, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995). “Whereas a vote dilution claim alleges that the State has enacted a ... purposeful device ‘to minimize or cancel out the voting potential of racial or ethnic minorities,’ ... the essence of [a racial sorting claim] is that the State has used race as a basis for separating voters into districts.” Id. (internal citations omitted).

In Shaw I, the Supreme Court faced two patently bizarre legislative districts. 509 U.S. at 635, 113 S.Ct. 2816. One resembled a “Rorshach ink-blot test” or a “bug splattered on a windshield,” while the other was “even more unusually shaped”:

[The district] is approximately 160 miles long and, for much of its length, no wider than the I–85 corridor. It winds in snakelike fashion through tobacco country, financial centers, and manufacturing areas until it gobbles in enough enclaves of black neighborhoods. Northbound and southbound drivers on I–85 sometimes find themselves in separate districts in one county, only to “trade” districts when they enter the next county. Of the 10 counties through which District 12 passes, 5 are cut into 3 different districts; even towns are divided. At one point the district remains contiguous only because it intersects at a single point with two other districts before crossing over them. One state legislator has remarked that “if you drove down the interstate with both car doors open, you'd kill most of the people in the district.”

Id. at 635–36, 113 S.Ct. 2816 (citations and some internal quotation marks omitted). Although the text of the legislation was facially neutral, the Court found that “it rationally can be viewed only as an effort to segregate the races for purposes of voting, without regard for traditional districting principles.” Id. at 642, 113 S.Ct. 2816.

For that reason, rather than requiring the plaintiffs to present evidence of discriminatory purpose and discriminatory effect, the Supreme Court treated the legislation as tantamount to a suspect facial classification and employed strict scrutiny. Id. at 642–43, 113 S.Ct. 2816 (“Express racial classifications are immediately suspect because, absent searching judicial inquiry, there is simply no way of determining what classifications are ‘benign’ or ‘remedial’ and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics.... Accordingly, we have held that the Fourteenth Amendment requires state legislation that expressly distinguishes among citizens because of their race to be narrowly tailored to further a compelling governmental interest. These principles apply not only to legislation that contains explicit racial distinctions, but also to those ‘rare’ statutes that, although race neutral, are, on their face, ‘unexplainable on grounds other than race.’ ”) (quoting Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977)).

In order to prove a racial sorting claim, a plaintiff must show that the legislature “subordinated” traditional race-neutral districting principles in crafting the district's boundaries:

The plaintiff's burden is 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 decision to place a significant number of voters within or without a particular district. To make this showing, a plaintiff must prove that the legislature subordinated traditional race-neutral districting principles, including but not limited to compactness, contiguity, and respect for political subdivisions or communities defined by actual shared interests, to racial considerations.
Miller, 515 U.S. at 916, 115 S.Ct. 2475 (emphasis added). This threshold standard is “a demanding one.” Indeed, the Plaintiffs must overcome a presumption that the legislature acted correctly and in good faith. Id. Thus, the plaintiff “must show that the State has relied on race in substantial disregard of customary and traditional districting practices.” Id. at 928, 115 S.Ct. 2475 (O'Connor, J., concurring).

If the plaintiff makes the requisite showing, the State must demonstrate that the redistricting legislation is narrowly tailored to advance a compelling state interest. In redistricting cases where the State claims a compelling interest in compliance with the VRA, the legislature must show that it had a “strong basis in evidence” to support its use of race-based districting. Alabama Legislative Black Caucus v. Alabama, ––– U.S. ––––, 135 S.Ct. 1257, 1274, 191 L.Ed.2d 314 (2015). In other words, the legislature must have “good reasons to believe” that its use of racial classifications was “required” by the VRA, “even if a court does not find that the actions were necessary for statutory compliance” after the fact. Id. at 1274.

C. The Voting Rights Act

In addition to these constitutional imperatives, redistricting legislation must also comply with the VRA. “The Voting Rights Act was designed by Congress to banish the blight of racial discrimination in voting [.]” South Carolina v. Katzenbach, 383 U.S. 301, 308, 86 S.Ct. 803, 15 L.Ed.2d 769 (1966) abrogated by Shelby Cnty., Ala. v. Holder, ––– U.S. ––––, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013). Enacted pursuant to Congress' enforcement powers under the Fifteenth Amendment, see Shelby Cnty., 133 S.Ct. at 2619–21, the VRA prohibits states from adopting plans that would result in vote dilution under Section 2 or—in covered jurisdictions—retrogression under Section 5.

In Shelby County, the Supreme Court struck down the coverage formula in Section 4, thereby drawing into question the status of covered jurisdictions' Section 5 compliance obligations until such time that Congress enacts a new coverage formula. 133 S.Ct. at 2631. At the time the redistricting plan at issue was developed and enacted, however, compliance with Section 5 was still a necessary consideration in Virginia's districting process. See Alabama, 135 S.Ct. at 1263.

Section 2 of the VRA prohibits the imposition of any electoral practice or procedure that “results in a denial or abridgement of the right of any citizen ... to vote on account of race or color....” 52 U.S.C. § 10301(a). A § 2 violation occurs when, based on the totality of circumstances, the political process results in minority “members hav[ing] less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 52 U.S.C. § 10301(b). By adopting the “discriminatory effect” language from Regester and omitting any requirement to prove discriminatory intent as required by Lodge, Congress created a statutory “results test” that could be brought by plaintiffs who might be otherwise unable to bring a claim of racial vote dilution under the Equal Protection Clause. See Shaw I, 509 U.S. at 641, 113 S.Ct. 2816 (“In 1982, [Congress] amended § 2 of the Voting Rights Act to prohibit legislation that results in the dilution of a minority group's voting strength, regardless of the legislature's intent.”).

In order to prove a § 2 violation, a plaintiff must satisfy three prerequisites: compactness, political cohesiveness, and bloc voting. “First, the minority group must be able to demonstrate that it is sufficiently large and geographically compact to constitute a majority in a single-member r district.” Thornburg v. Gingles, 478 U.S. 30, 50, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986). “Second, the minority group must be able to show that it is politically cohesive.” Id. at 51, 106 S.Ct. 2752. “Third, the minority must be able to demonstrate that the white majority votes sufficiently as a bloc to enable it—in the absence of special circumstances, such as the minority candidate running unopposed—usually to defeat the minority's preferred candidate.” Id. These final two factors are often referred to collectively as “racial polarization.” Once these prerequisites have been satisfied, the court evaluates the plaintiff's evidence based on the totality of the circumstances. The totality of circumstances must be considered with a focus on whether the minority group in question was denied “equal political opportunity.” Johnson v. De Grandy, 512 U.S. 997, 1014, 114 S.Ct. 2647, 129 L.Ed.2d 775 (1994).

With respect to redistricting legislation, § 2 establishes a “natural floor” based on the State's demographics for the number of districts wherein members of a minority group must maintain an “equal political opportunity” to “elect representatives of their choice.” Where a minority group is sufficiently large and geographically compact to constitute a numerical majority in a hypothetical district, § 2 requires the creation of a district wherein members of that group maintain the equal ability to elect representatives of their choice. See Bartlett v. Strickland, 556 U.S. 1, 13, 129 S.Ct. 1231, 173 L.Ed.2d 173 (2009). Proving this hypothetical requires the plaintiffs to present an alternative redistricting plan. See Reno v. Bossier Parish Sch. Bd., 520 U.S. 471, 480, 117 S.Ct. 1491, 137 L.Ed.2d 730 (1997) (“Because the very concept of vote dilution implies—and, indeed, necessitates—the existence of an ‘undiluted’ practice against which the fact of dilution may be measured, a § 2 plaintiff must also postulate a reasonable alternative voting practice to serve as the benchmark ‘undiluted’ voting practice.”).

Section 5 of the VRA, on the other hand, forbids voting changes with “any discriminatory purpose” as well as voting changes that diminish the ability of citizens, on account of race, color, or language minority status, “to elect their preferred candidates of choice.” Shelby County, 133 S.Ct. at 2621. Sections 2 and 5 “differ in structure, purpose, and application. Section 5 applies only in certain jurisdictions specified by Congress and ‘only to proposed changes in voting procedures.’ ” Holder v. Hall, 512 U.S. 874, 883, 114 S.Ct. 2581, 129 L.Ed.2d 687 (1994) (quoting Beer, 425 U.S. at 138, 96 S.Ct. 1357) (emphasis added).

Section 5 was enacted as “a response to a common practice in some jurisdictions of staying one step ahead of the federal courts by passing new discriminatory voting laws as soon as the old ones had been struck down.” Beer, 425 U.S. at 140, 96 S.Ct. 1357. By requiring that proposed changes be approved in advance, Congress sought “ ‘to shift the advantage of time and inertia from the perpetrators of the evil to its victim,’ by ‘freezing election procedures in the covered areas unless the changes can be shown to be nondiscriminatory.’ ” Id. (quoting H.R.Rep. No. 94–196, pp. 57–58 (1970)). The purpose of this approach was to ensure that “no voting-procedure changes would be made that would lead to a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise.” Holder, 512 U.S. at 883, 114 S.Ct. 2581.

“Retrogression, by definition, requires a comparison of a jurisdiction's new voting plan with its existing plan. It also necessarily implies that the jurisdiction's existing plan is the benchmark against which the ‘effect’ of voting changes is measured.” Reno, 520 U.S. at 478, 117 S.Ct. 1491. Unlike the “natural floor” of § 2 ensuring equal ability to elect, the retrogression standard of § 5 creates a “relative floor” based upon the existing benchmark plan. Under § 5, the State must ensure that the new plan does not “lead to a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise” by diminishing the ability of minority voters to elect their preferred candidates of choice as compared to the State's existing plan.

Therein lies the rub. To comply with federal statutory command (the VRA), the State must consider and account for race in drawing legislative districts in order to craft a compliant plan. However, to avoid violating the federal constitution, the State must not subordinate traditional, neutral principles to racial considerations in drawing district boundaries.

Apologies to Shakespeare for the misquotation. See William Shakespeare, Hamlet, Act 3, Scene 1, 3:66 (“[A]y, there's the rub.”).

And, at the same time, the State must also comply with the “one person, one vote” constitutional requirement as specified in Reynolds v. Sims. That, of course, is not a traditional redistricting principle to be weighed as part of the predominance inquiry, as Alabama makes clear. But it is a federal constitutional requirement that, of necessity, is central to the redistricting process and that is highly instrumental in the drawing of district boundaries.

It is within the context of this legal framework that the Virginia General Assembly sought to design and enact a compliant redistricting plan. And these principles are central to the resolution of this case.

Before proceeding to the facts of the case, the Court feels it necessary to pause and recognize that Delegate Jones, members of the redistricting committee, and other legislators involved in the crafting and amendment of HB 5005 did not have the benefit of either the Supreme Court's guidance in the recent Alabama decision or the guidance provided in the opinion entered here today. Based on the evidence and testimony provided in the record, the Court believes that all of the legislators involved proceeded in a good faith attempt to comply with all relevant constitutional and statutory demands, as they understood them at the time.

III. Factual Background

A. The 2011 Redistricting Process

The first steps in the redistricting process began well before the United States Census Bureau released its population and demographic data. Trial Tr. 273:11 (Jones). On August 23, 2010, Delegate Mark Cole announced that the redistricting subcommittee of the House of Delegates Committee on Privileges and Elections had scheduled a series of six public hearings throughout the Commonwealth to solicit input into the House redistricting process. (Docket No. 85.) These public hearings were held between September 8, 2010 and December 17, 2010. Id.; Trial Tr. 273:14–19 (Jones). Following these hearings, Governor McDonnell signed Executive Order 31 on January 10, 2011, creating the “Independent Bipartisan Advisory Redistricting Commission” (“Governor's Commission”) to develop plan proposals, review public input, and analyze recommendations from other stakeholders in the voting public. (Docket No. 85.)

Redistricting began in earnest in February 2011 when the 2010 census data was released via Public Law 94–171. Trial Tr. 276:4–21 (Jones). On March 25, 2011, the House Committee on Privileges and Elections adopted a resolution setting out the criteria that the committee would follow in reviewing redistricting plans. Pls.' Ex. 48 at 6. The House Committee established six criteria, which were as follows:

The initial data released on February 3, 2011 contained an error. A corrected data set was provided a few weeks later. Trial Tr. 276:4–21 (Jones).

I. Population Equality: The population of legislative districts shall be determined solely according to the enumeration established by the 2010 federal census. The population of each district shall be as nearly equal to the population of every other district as practicable. Population deviations in House of Delegates districts should be within plus-or-minus one percent.
II. Voting Rights Act: Districts shall be drawn in accordance with the laws of the United States and the Commonwealth of Virginia including compliance with protections against the unwarranted retrogression or dilution of racial or ethnic minority voting strength. Nothing in these guidelines shall be construed to require or permit any districting policy or action that is contrary to the United States Constitution or the Voting Rights Act of 1965.
III. Contiguity and Compactness: Districts shall be comprised of contiguous territory including adjoining insular territory. Contiguity by water is sufficient. Districts shall be contiguous and compact in accordance with the Constitution of Virginia as interpreted by the Virginia Supreme Court in the cases of Jamerson v. Womack, 244 Va. 506 (1992) and Wilkins v. West, 264 Va. 447 (2002).
IV. Single–Member Districts: All districts shall be single-member districts.
V. Communities of Interest: Districts shall be based on legislative consideration of the varied factors that can create or contribute to communities of interest. These factors may include, among others, economic factors, social factors, cultural factors, geographic factors, governmental jurisdictions and service delivery areas, political beliefs, voting trends, and incumbency considerations.... Local government jurisdiction and precinct lines may reflect communities of interest to be balanced, but they are entitled to no greater weight as a matter of state policy than other identifiable communities of interest.
VI. Priority: All of the foregoing criteria shall be considered in the districting process, but population equality among districts and compliance with federal and state constitutional requirements and the Voting Rights Act of 1965 shall be given priority in the event of conflict among the criteria. Where the application of any of the foregoing criteria may cause a violation of applicable federal or state law, there may be such deviation from the criteria as is necessary, but no more than is necessary, to avoid such violation.

Pls.' Ex. 16. These criteria were substantially similar to the criteria adopted by the committee in the 2001 redistricting cycle, with two exceptions. Ints.' Ex. 27. First, the 2001 criteria had permitted a population deviation of “plus-or-minus two percent,” rather than one percent, which Delegate Jones stated was altered to better “approximate the one-person-one-vote [standard] in the Virginia constitution.” Trial Tr. 275:10–19 (Jones). Second, the 2001 criteria were updated to include a citation to the decision of the Supreme Court of Virginia in Wilkins v. West as part of the “Contiguity and Compactness” criterion. Id. at 275:13–15.

B. The 55% Black Voting Age Population Floor

At the time the redistricting process began, the twelve Challenged Districts had black voting-age populations (“BVAP”) ranging from 46.3% to 62.7%. Three of the districts had BVAPs below 55%. All others were above 55%. Several legislators believed that the twelve “ability-to-elect” districts found in the 2001 redistricting plan (or “Benchmark Plan”) needed to contain a BVAP of at least 55% in the 2011 redistricting plan to avoid “unwarranted retrogression” under Section 5 of the VRA and to comply with Criterion II of their own redistricting rules.

The existence of a fixed racial threshold can have profound consequences for the Court's predominance and narrow tailoring inquiries in a racial sorting claim, so a substantial amount of time at trial was devoted to questions related to this factual topic. However, the most important question—whether such a figure was used in drawing the Challenged Districts—was not disputed. Rather, the parties disputed whether the 55% BVAP was an aspiration or a target or a rule. In the end, it is not relevant whether the 55% BVAP was a rule or a target because all the parties agree—and the Court finds—that the 55% BVAP figure was used in structuring the districts and in assessing whether the redistricting plan satisfied constitutional standards and the VRA, and whether the plan would be precleared by the Department of Justice (“DOJ”).

Plaintiffs introduced a fair amount of evidence, such as e-mail communications and floor debate, pertaining to HB 5001 rather than HB 5005. For some purposes, such as whether the drafters employed a 55% rule during redistricting, the evidence pertaining to HB 5001 is equally relevant to HB 5005. See Ints.' Ex. 7 at 3–8 (“[MR. ARMSTRONG:] In order for me not to have to go through the extensive dialogue we did here the other day on HB 5001, I would ask the gentleman would ... his answers to my questions per HB 5001 essentially be applicable to HB 5005? [MR. JONES]: Mr. Speaker, I would say to the gentleman I would believe that will be correct.... [MR. ARMSTRONG]: I thank the gentleman for allowing me to streamline the questions.”). For other purposes, such as whether the 55% threshold impacted a particular boundary, the evidence pertaining to HB 5001 cannot necessarily be applied to HB 5005. Compare Pls.' Ex. 30 at 1 (e-mail from Delegate McClellan to Richmond Registrar Kirk Showalter regarding HB 5001, stating “[T]he changes we discussed ... would have pushed the [BVAP] in the 71st District down to 54.8%. The target criteria was 55%, so the change can't be made.”) with Ints.' Ex. 7 at 2–3 (floor testimony from Delegate Jones regarding HB 5005, stating, “There was a request made by the registrar of Richmond City working with the gentlewoman from Richmond to make some adjustments to those boundaries, and we did split a precinct in anticipation of moving a polling place this fall for the upcoming elections.”).

At trial, two additional questions regarding the 55% figure dominated the discussion. First, whether the BVAP figure included or excluded those who identified themselves in the census process as ethnically Hispanic and racially black. And second, what the source of the 55% BVAP figure was.

The parties hotly debated whether the appropriate measure of BVAP used in the redistricting process did or did not include individuals who identified as racially black and ethnically Hispanic in the census data. The supposed importance of this dispute was that, if black Hispanics were excluded from the black population count, three of the Enacted Plan's majority-minority districts would actually contain a BVAP percentage just shy of 55%. Trial Tr. 280:24– 281:10 (Jones); 862:4–7 (Intervenors). That, according to Intervenors, would support a finding that there was not a 55% BVAP floor in deciding on the twelve Challenged Districts.

The record shows that delegates attempting to comply with the 55% BVAP floor submitted their proposed changes using data that included black Hispanics in the BVAP count. See Pls.' Ex. 33 at 46; Trial Tr. 40:10–25 (McClellan); Trial Tr. 68:23–69:2 (Dance); Ints.' Pre–Trial Brief at 8. Although Delegate Jones claimed to personally believe that the DOJ would use a BVAP figure excluding black Hispanics, Trial Tr. 286:8–16 (Jones), this was not a distinction that he discussed with any other delegates, id. at 427:1–428:16 & 490:2–4, and he repeatedly asserted on the House floor that all majority-minority districts in the proposed legislation had a BVAP of 55% or higher, Pls.' Ex. 35 at 42, 66, 108. Moreover, Delegate Jones “assumed” that Virginia, in its preclearance submissions to the DOJ, would represent that all 12 majority-minority districts contained at least 55% BVAP. Trial Tr. 447:6–8 (Jones). This turned out to be the case. Pls.' Ex. 48 at 11 (“All 12 black majority districts were maintained ... with greater than 55% black VAP—a range of 55.2% to 60.7%.”).

At trial, Intervenors relied on a spreadsheet prepared by the Division of Legislative Services (“DLS”) in an attempt to show that including Hispanics in the BVAP count would be erroneous. The spreadsheet contains rows of data by district and, in each column, contains metrics such as total population, population by race, racial population by percentage, population by ethnicity, and ethnic population by percentage. Pls.' Ex. 60 at 13. After adding the racial and ethnic population totals column by column, the Intervenors dramatically revealed that the number exceeded that of the district's total population. Trial Tr. 282:10–286:7 (Jones). But this exercise reflects an error on the part of the Intervenors, not DLS. Because ethnicity measures a different variable than race, the racial and ethnic data are not meant to be added in the first place. If one removes the ethnicity column from the count (on the assumption that Hispanic individuals of any race are already counted in their respective racial columns), then the total population figure is corrected. That does not, however, imply that Hispanics who are racially black should be excluded from the black population count because to do so would undercount the number of black individuals in the BVAP percentage.

The record shows that the ethnic data provided by the census only has redistricting implications in states that may need to craft majority-Hispanic districts or majority-“black-plus-Hispanic” (or “coalition”) districts. In states such as Virginia, on the other hand, black Hispanics would count towards the total black population of a district for retrogression purposes. Id. at 747:14–749:12 & 752:17–754:17 (Ansolabehere). That appears to be consistent with the DOJ's (admittedly confusing) guidance on this question: “If there are significant numbers of responses which report Latino and one or more minority races (for example, Latinos who list their race as Black/African American), those responses will be allocated alternatively to the Latino category and the minority race category.” Pls.' Ex. 9 at 4–5 (76 Fed.Reg. Vol. 27 (Feb. 9, 2011) at 7472–7473). This “alternating” approach presumably applies to situations where the district would be majority-“black-plus-Hispanic,” in which case counting black Hispanic individuals as either black or Hispanic in alternating fashion would avoid counting those individuals twice in the same district. Trial Tr. 757:1–12 (Ansolabehere). Thus, the Court finds that the proper count includes black Hispanics within the BVAP percentage of each majority-minority district. This method of counting results in a BVAP above 55% for all twelve majority-minority districts, ranging from 55.2% to 60.7%.

The Court recognizes that “Hispanic” and “Latino” are not interchangeable designations but has been forced into this unfortunate conflation by the record.

Regardless, this debate—like the first—generated more heat than light. The actual differences in BVAP percentages were minute, and both parties eventually agreed that the distinction was not one of great legal significance. See id. at 816:5–9 (Plaintiffs) (“The distinction between how [these are] calculate [d] ... is simply irrelevant, and it doesn't matter what we call it. They used a racial target, and whether that was 53 or 54 or 55 or 56, whether you measure it this way or that way, it just doesn't matter.”) and id. at 862:8–11 (Intervenors) (“Do I believe the difference between these two numbers is in reality meaningful in actual reality? No, it isn't a significant difference one way or the other, let's be candid.”).

Unlike the first two questions, the answer to the third question—i.e., the source of the 55% rule—can carry great legal significance. Testimony on this question is a muddle. Delegate Dance testified that her understanding came from Delegate Jones and that the 55% figure was necessary in order to achieve DOJ approval, id. at 70:18–23 (Dance), but her speech from the House floor appears to represent it as her own understanding, see Pls.' Ex. 33 at 45 (“[W]e need 55 percent at least voting African–Americans[.]”). Delegate McClellan understood the committee's adopted criteria to require “each of the majority-minority districts ... to have a black voting-age population of at least 55 percent,” Trial Tr. 33:1–4 (McClellan), and testified that she came to this understanding “[t]hrough conversations with Delegate Jones and with Legislative Services,” id. at 33:6–8. Delegate Tyler testified that her understanding came from Delegate Spruill, (Docket No. 90–2, Ex. B at 57:5–8), and Delegate Armstrong testified that, “as far as [he] could tell, the number was almost pulled out of thin air,” Trial Tr. 98:1–2 (Armstrong).

Delegate Jones initially testified that the figure was drawn from the public hearings held with the community. See id. at 424:1–4 (Jones) (55% BVAP “is what the community had indicated to us that they felt would allow them to elect the candidate of their choice”); id. at 429:8–9 (“That was the testimony that we heard during the public hearings.”). Although this testimony is consistent with his prior statements from the House floor, see Pls.' Ex. 35 at 72, the trial record does not support it. At trial, Delegate Jones admitted that he had not read the transcripts from every hearing and could not recall a single instance of a member of the public requesting a 55% BVAP level. Trial Tr. 442:18–443:9 (Jones). Moreover, most of these hearings were transcribed and submitted as evidence. A review of the public hearing transcripts from the Fall of 2010 fails to reveal any mention of the 55% figure. See Pls.' Exs. 3–6, Ints.' Ex. I.

There is, admittedly, one comment made regarding the maintenance of 55 percent voting strength during a public hearing held on April 4, 2011, Pls.' Ex. 31 at 20, but this was the same day that the Joint Committee reported out a substitute for HB 5001, (Docket No. 85 at 3). In other words, the 55% floor was in effect well before this lone comment was offered.

Delegate Jones also claimed that the 55% figure came from “Delegate Dance, and Delegate Tyler, Delegate Spruill, and one or two othe[r] ... African–American members of the House.” Trial Tr. 431:4–7 (Jones). This was then narrowed to Delegates Dance, Tyler, and Spruill. Id. at 490:5–13. After further questioning, the 55% figure appears to have come from feedback that Delegate Spruill received from various groups in Virginia and from concerns that Delegate Tyler would be unable to hold her seat in HD 75 with a lower BVAP percentage. Id. at 494: 6–495:1. In discussing Delegate McClellan's seat, by contrast, Delegate Jones indicated that, while “no one” was comfortable leaving the BVAP percentage in HD 71 at 46%, “they felt that we needed to have a performing majority-minority district, and from the members that I spoke to, they felt that it needed to be north of 50 percent minimum.” Id. at 293:6–16 (emphasis added).

Based on the foregoing testimony, and the evidence set forth below, the Court finds—based on the record presented—that the 55% BVAP floor was based largely on concerns pertaining to the re-election of Delegate Tyler in HD 75 and on feedback received from Delegate Spruill and, to a lesser extent, Delegates Dance and Tyler. That figure was then applied across the board to all twelve of the Challenged Districts.

C. The Passage and Enactment of HB 5005

During the redistricting process, the General Assembly initially considered three plans: HB 5001, HB 5002, and HB 5003. HB 5001 was the plan designed and proposed by Delegate Jones. HB 5002 and HB 5003, on the other hand, were designed by university students and proposed by other members of the House of Delegates. Id. at 376:24–378:9. According to Delegate Jones, HB 5002 paired somewhere between 40 and 48 incumbents, contained six majority-minority districts, and had over a 9% population deviation. Id. at 378:10–379:4. HB 5003, on the other hand, paired somewhere between 32–34 incumbents, contained nine or ten majority-minority districts, and also did not meet the population deviation criteria. Id. at 379:8–17. The Governor's Commission also designed two plans that contained 13 and 14 majority-minority districts, respectively; however, those plans were never formally introduced or proposed. Id. at 379:18–380:11.

Once the House had coalesced around HB 5001 and the plan was married with the Senate's redistricting plan, the bill was ready for passage and enactment. On April 12, 2011, the Virginia General Assembly passed HB 5001. (Docket No. 83.) Based largely upon objections to the Senate plan, then-Virginia Governor Robert McDonnell vetoed HB 5001 three days later. Ints.' Ex. 10. After relatively minor revisions to the House plan and more substantial revisions to the Senate plan, Pls.' Ex. 48 at 10, the legislature passed HB 5005, which was signed by the Governor and enacted into law on April 29, 2011, (Docket No. 83).

To comply with its obligations under the VRA, the Commonwealth then submitted the Enacted Plan (or “the Plan”) to the DOJ for preclearance. Id. The DOJ precleared the Plan on June 17, 2011, (Docket No. 83), and the first election under the new districts was held on November 8, 2011, (Docket No. 85).

IV. ANALYSIS

The questions raised in a racial sorting claim are deceptive in their simplicity but profound in their implications. Resting at the crossroads of race, politics, and the constitutional limits of federal power, the claim raises vital questions about how we identify as citizens and how we project that identity in the halls of the legislature. The Supreme Court has crafted an interpretive standard for navigating this field: the legislature must not allow racial considerations to predominate over (i.e., to subordinate) traditional redistricting criteria. If this results from attempted compliance with the VRA, the State must show a “strong basis in evidence” that its use of race was necessary to comply with a constitutional reading of the statute.

What this standard provides in conceptual grace, however, it lacks in practical guidance. For legislators, it does little to signal when it may be constitutionally permissible to cut through a precinct or move a boundary line to alter the demographic composition of a district for purposes of complying with similarly mandatory federal law. For litigators, it provides an enticingly vague standard and invites litigation that can drive up the cost of conducting and defending the State's redistricting endeavor. See Abrams v. Johnson, 521 U.S. 74, 118, 117 S.Ct. 1925, 138 L.Ed.2d 285 (1997) (Stevens, J., dissenting) (“Any redistricting plan will generate potentially injured plaintiffs, ... [a]nd judges (unable to refer, say, to intent, dilution, shape, or some other limiting principle) will find it difficult to dismiss those claims[.]”). And for courts, it provides an uncomfortable amount of discretion in a field that the Supreme Court has repeatedly admonished “represents a serious intrusion on the most vital of local functions.” Miller, 515 U.S. at 915, 115 S.Ct. 2475. By asking courts attempting to identify predominance to engage in a searching factual inquiry and comprehensive balancing before applying strict scrutiny—and to justify strict scrutiny—the test gives the judicial branch the relatively broad power to strike down or uphold legislative districts without much guidance in how to do so, notwithstanding exhortations to exercise “extraordinary caution” to the contrary.

Therefore, to sharpen the judicial inquiry, to ensure that the requisite burden is satisfied, and to assess whether redistricting legislation has successfully navigated the narrow passage between constitutional and unconstitutional redistricting, it is appropriate to articulate how the Court understands the predominance and strict scrutiny inquiries are to proceed as a matter of law. The statewide and district-by-district evidence then will be assessed within that framework.

A. The Racial Sorting Framework

The essence of the racial sorting analysis is quite easy to articulate and comprehend. First, courts examine whether racial considerations predominated over—or “subordinated”—traditional redistricting criteria. If a court so finds, then the court applies strict scrutiny. Second, the court examines whether the legislature had a strong basis in evidence for believing federal law required its use of race, assuming this is the basis upon which the State seeks to justify its decision.

But, as this case demonstrates, the devil is in the details. The parties actually have proposed conflicting rules regarding the “subordination” test. And each believes that the Supreme Court's recent Alabama decision reinforces its position. But both cannot be right, and we think that neither is.

The Plaintiffs' case and our colleague's dissent revolve chiefly around the evidence that legislators employed a 55% BVAP floor when crafting the Challenged Districts. According to Plaintiffs' theory, “race predominates if it is the most important criterion.” Pls.' Post–Trial Brief at 4 (Docket No. 105). In other words, subordination “does not require open conflict with ‘traditional’ districting criteria.” Id. at 5.

Thus, the Plaintiffs, like the dissent, propose a per se rule: the drafters' use of the 55% BVAP floor in districting is verboten and automatically satisfies Miller's predominance standard. This, the Plaintiffs argue, is the central thrust of the Alabama case:

This case boils down to a very simple proposition: May Virginia's General Assembly utilize a fixed numerical racial threshold in establishing district lines.... The answer to this question has been addressed and definitively settled by the United States Supreme Court in its recent Alabama decision which unambiguously condemned the use of racial thresholds in redistricting[.]

Trial Tr. 811:1–10 (Plaintiffs).

Despite its tempting simplicity and visceral appeal, the Court must reject this proposal. Although the Alabama decision condemned the use of unwritten racial thresholds, it did not establish a per se predominance rule. In Alabama, the Court accepted the lower court's finding that legislators had employed BVAP percentage floors in the challenged districts. See Alabama, 135 S.Ct. at 1271 (“The legislators in charge of creating the redistricting plan believed, and told their technical adviser, that a primary redistricting goal was to maintain existing racial percentages in each majority-minority district, insofar as feasible.”). If the use of those thresholds constituted predominance per se, then there would have been little reason for the Supreme Court to have remanded the case to the district court to determine whether race predominated. Id. at 1272.

Rather, the Court pointed out that “[t]here [was] considerable evidence that this goal had a direct and significant impact on the drawing of at least some of [the district's] boundaries.” Id. at 1271 (emphasis added). “That [the State] expressly adopted and applied a policy of prioritizing mechanical racial targets above all other districting criteria (save one-person, one-vote) provides evidence that race motivated the drawing of particular lines in multiple districts in the State.” Id. at 1267 (emphasis added).

The Alabama case could not be clearer that use of racial BVAP floors constitutes evidence—albeit significant evidence—of predominance. But, we do not read Alabama to hold that use of a BVAP floor satisfies the Plaintiffs' predominance burden merely because the floor was prioritized “above all other districting criteria” in “importance.” Rather, the significance of the racial floor is its impact on the creation of the district. This demands “actual conflict between traditional redistricting criteria and race that leads to the subordination of the former, rather than a merely hypothetical conflict that per force results in the conclusion that the traditional criteria have been subordinated to race.” Page v. Virginia State Bd. of Elections, No. 3:13CV678, 2015 WL 3604029, at *27 (E.D.Va.2015) (Payne, J., dissenting).

To understand why this is so, one must remember the origin of—and the rationale for—the Shaw claim. The district boundaries in Shaw were so outlandish that—despite any express textual classification by race in the statute—“it rationally [could] be viewed only as an effort to segregate the races for purposes of voting, without regard for traditional districting principles.” Shaw I, 509 U.S. at 642, 113 S.Ct. 2816. In response, the Court treated the legislation as though it had employed a facial classification and subjected the legislation to strict scrutiny rather than requiring the plaintiffs to prove both discriminatory purpose and discriminatory effect.

In Shaw, the Court compared the districts to racial “balkanization” and “political apartheid” and cautioned that such districts threaten expressive harm—i.e., the stigmatization of individuals “by reason of their membership in a racial group” and the incitement of “racial hostility”—as well as representative harm—i.e., the threat that elected officials would begin to “believe that their primary obligation is to represent only the members of that group, rather than their constituency as whole.” Id. at 657, 643, 648, 113 S.Ct. 2816.

Unlike in its racial and political vote dilution cases, however, the Supreme Court did not charge plaintiffs with producing evidence that such discriminatory effects had, in fact, come to pass. See e.g., Rogers, 458 U.S. at 625–27, 102 S.Ct. 3272 (observing in racial vote dilution case that “[e]xtensive evidence was cited by the District Court to support its finding that elected officials of Burke County have been unresponsive and insensitive to the needs of the black community, which increases the likelihood that the political process was not equally open to blacks”); Davis v. Bandemer, 478 U.S. 109, 131–32, 106 S.Ct. 2797, 92 L.Ed.2d 85 (1986) (observing in political vote dilution case that “[a]n individual or a group of individuals who votes for a losing candidate is usually deemed to be adequately represented by the winning candidate and to have as much opportunity to influence that candidate as other voters in the district” and that the Court “cannot presume in such a situation, without actual proof to the contrary, that the candidate elected will entirely ignore the interests of those voters”) (emphasis added). Such evidence is not necessary in a racial sorting claim because “[e]xpress racial classifications are immediately suspect” and are subjected to strict scrutiny. Shaw I, 509 U.S. at 642, 113 S.Ct. 2816. This is similarly true for the functional equivalents of express racial classifications: statutes “unexplainable on grounds other than race” or statutes that are an “obvious pretext for racial discrimination.” See id. at 643–44, 113 S.Ct. 2816.

No sooner had the ink dried on the Supreme Court's opinion in Shaw, than it was faced with a slightly different question. What if the district's boundaries are not “bizarre” or “irrational,” but still reflect a clear manifestation of racial classification? In Miller, the Court recognized that Shaw represented an “analytically distinct” claim, 515 U.S. at 911, 115 S.Ct. 2475, but decided that the litigation before it “require[d] [the Court] further to consider the requirements of the proof necessary to sustain this equal protection challenge,” id. at 915, 115 S.Ct. 2475. Rather than abandoning the claim's animating principles, the Court altered the threshold showing and clarified that parties bringing a racial sorting claim are “neither confined in their proof to evidence regarding the district's geometry and makeup nor required to make a threshold showing of bizarreness.” Id.

The district challenged in Miller was not as bizarre as those found in Shaw, but, “when its shape [was] considered in conjunction with its racial and population densities,” it became “exceedingly obvious” that the district employed “narrow land bridges” in “a deliberate attempt to bring black populations into the district.” Id. at 917, 115 S.Ct. 2475. There, the district's various spindly appendages contained nearly 80% of the district's total black population. Id. These facially evident deviations from neutral districting conventions could only be explained on the basis of race. Id. at 918–19, 115 S.Ct. 2475. Thus, districts such as the one found in Miller still raise the specter of expressive or representative harms and still manifest, on the face of the law, the lawmakers' clear intent to “us[e] race as a basis for separating voters into districts.” Id. at 911, 115 S.Ct. 2475. Moreover, these districts necessarily reflect the kind of “very stereotypical assumptions the Equal Protection Clause forbids;” namely, the “demeaning notion that members of the defined racial groups ascribe to certain ‘minority views' that must be different from those of other citizens.” Id. at 914, 115 S.Ct. 2475.

Miller, the State conceded that “portions of Effingham and Chatham Counties” would not have been added “but for the need to include additional black population;” that “a substantial reason for [the district's precinct splits] was the objective of increasing the black population of that district;” and that the addition of the district itself was “the product of a desire by the General Assembly to create a majority black district”. Furthermore, “Georgia's Attorney General objected to the Justice Department's demand for three majority-black districts on the ground that to do so the State would have to ‘violate all reasonable standards of compactness and contiguity.’ ” 515 U.S. at 918–19, 115 S.Ct. 2475.

However, when racial considerations do not entail the compromise of neutral districting norms, the basis for a racial sorting claim evaporates. Traditional, neutral districting principles reflect certain judgments about voters, but these are the same judgments that animate all geographic—as opposed to proportional—representation systems: that those who live near each other in the same communities, counties, and cities have something in common, something that warrants their representation as a reasonably defined geographical—rather than racial or political—unit.

More importantly, holding that otherwise reasonably neutral districts are subject to strict scrutiny because of a merely theoretical or latent conflict between race and traditional districting criteria would unlash the Shaw claim from the mooring of facial classification jurisprudence. If this legal equivalence is forfeited, it is unclear why the “analytically distinct” nature of the claim should not unravel entirely, forcing plaintiffs to prove the expressive or representative harms postulated in Shaw.

Admittedly, the issue presented in this case is a difficult one. The Supreme Court reserved from the very outset the question of whether the intentional use of a 50% BVAP threshold was sufficient to sustain a racial sorting claim:

It is unnecessary for us to decide whether or how a reapportionment plan that, on its face, can be explained in nonracial terms successfully could be challenged. Thus, we express no view as to whether “the intentional creation of majority-minority districts, without more,” always gives rise to an equal protection claim.

Shaw I, 509 U.S. at 649, 113 S.Ct. 2816. Although the principal opinion in Bush v. Vera attempted to put this question to rest, 517 U.S. 952, 958, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996) (“Strict scrutiny does not apply ... to all intentional creation of majority-minority districts.”) (principal opinion), Justice Kennedy expressed some doubts in his concurring opinion:

I join the plurality opinion, but the statements in ... the opinion that strict scrutiny would not apply to all cases of intentional creation of majority-minority districts require comment. I do not consider these dicta to commit me to any position on the question whether race is predominant whenever a State, in redistricting, foreordains that one race be the majority in a certain number of districts or in a certain part of the State.

Id. at 996, 116 S.Ct. 1941 (Kennedy, J., concurring) (internal citation omitted).

Based on the Supreme Court's recent decision in Alabama, the Court now appears to be divided, or at least equivocal, on whether BVAP thresholds alone are sufficient to constitute predominance. Compare Alabama, 135 S.Ct. at 1267 (noting that the prioritization of “mechanical racial targets above all other districting criteria” only provides evidence that race predominated) with League of United Latin American Citizens v. Perry (LULAC), 548 U.S. 399, 517, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006) (Scalia, J., concurring in the judgment in part and dissenting in part, joined by Chief Justice Roberts, Justice Thomas, and Justice Alito) (arguing that the intentional use of a 50% BVAP threshold necessarily means race predominated).

Although the unwritten use of a racial floor by legislators may seem repugnant at first blush, the interpretation of predominance proposed by the Plaintiffs and the dissent has quite serious repercussions. If the use of a BVAP threshold—any BVAP threshold—is sufficient to trigger strict scrutiny in the absence of a facial manifestation in the lines themselves through the subordination of traditional redistricting principles, then the constitutionality of the Voting Rights Act—as applied to redistricting—would be drawn into question. More fundamentally, the compatibility of the Fourteenth Amendment's Equal Protection Clause and the Fifteenth Amendment's Enforcement Clause might be drawn into question. The Court does not believe that the Constitution—or that Supreme Court precedent—either requires or permits the Plaintiffs' view of predominance and, therefore, does not believe that the racial sorting claim extends any further than its original purpose: to strike down those districts that, on their face, reflect racial classifications.

The dissent contends that we need not grapple with the issues that follow because we are faced with a “more narrow question.” See post at 574–75. But incrementalism does not demand that the Court ignore the clear consequences of two different judicial constructions when weighing which to adopt. If one sets us on a path to constitutional conflict and one avoids that path, we think that the latter is to be preferred.

“[E]ven if § 1 of the [Fifteenth] Amendment prohibits only purposeful discrimination, the prior decisions of th[e] [Supreme] Court foreclose any argument that Congress may not, pursuant to § 2, outlaw voting practices that are discriminatory in effect.” City of Rome v. United States, 446 U.S. 156, 173, 100 S.Ct. 1548, 64 L.Ed.2d 119 (1980). The ability-to-elect standard, which inherently utilizes racial floors in its redistricting applications, would seem to provide just such a necessary and proper statutory prophylaxis. See id. at 175, 177, 100 S.Ct. 1548. No one doubts that redistricting legislation can threaten the right to vote on account of race in defiance of the Fifteenth Amendment's guarantee, see Gomillion v. Lightfoot, 364 U.S. 339, 346–48, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960), or that the VRA protects against this threat of deprivation, see Allen v. State Board of Elections, 393 U.S. 544, 569, 89 S.Ct. 817, 22 L.Ed.2d 1 (1969). And, of course, “no one doubts” that “voting discrimination still exists.” Shelby County, 133 S.Ct. at 2619. Therefore, unless the Enforcement Clause is to be read with a rigidity alien to all other positive grants of legislative power, then the use of racial targets by states acting under congressional mandate would not—by itself—seem an appropriate per se trigger for strict scrutiny.

Moreover, the Plaintiffs do not take umbrage at the use of racial targets, so long as those targets serve the ends of preserving minority voters' ability to elect. Quoting from the Alabama decision during their closing statement, the Plaintiffs observed that, in order to be narrowly tailored, the legislature must ask “to what extent must we preserve existing minority percentages in order to maintain the minorities' present ability to elect the candidate of its choice.” Trial Tr. 819:23–820:1 (Plaintiffs) (quoting Alabama, 135 S.Ct. at 1274). But, the inquiry into whether the targets are adequately justified only occurs after finding race predominant. If targets themselves constitute subordination, then it is hard to see how the Plaintiffs have not smuggled one inquiry into the next. This would again threaten the foundations of the VRA by making all its redistricting applications subject to strict scrutiny and set up a potential conflict between the Fourteenth Amendment's Equal Protection Clause and the Fifteenth Amendment's Enforcement Clause.

Plaintiffs have occasionally flirted with this notion: “The Shaw cases ... prohibit all unjustified race-based redistricting, whatever form it may take.” Pls.' Post–Trial Reply at 6. That said, counsel for Plaintiffs has claimed that there must be a floor of “50 percent plus one” under Section 2 of the VRA. Trial Tr. 842:17–19 (Plaintiffs).

After this journey, we thus arrive back where we started: Miller's predominance test. In Miller, the Court described the Plaintiffs' burden as follows:

The plaintiff's burden is 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 decision to place a significant number of voters within or without a particular district. To make this showing, a plaintiff must prove that the legislature subordinated traditional race-neutral districting principles, including but not limited to compactness, contiguity, and respect for political subdivisions or communities defined by actual shared interests, to racial considerations.

515 U.S. at 916, 115 S.Ct. 2475 (emphasis added). Plaintiffs would prefer we stop reading Miller at this exact punctuation mark. And, under that formulation, they could plausibly argue that they have proved racial predominance merely upon proof that legislators used a 55% BVAP floor. But the very next sentence in Miller leads where this Court must follow: “Where these or other race-neutral considerations are the basis for redistricting legislation, and are not subordinated to race, a State can ‘defeat a claim that a district has been gerrymandered on racial lines.’ ” Id. (quoting Shaw I, 509 U.S. at 647, 113 S.Ct. 2816) (emphasis added). The Court's quotation of Shaw in this instance rather clearly reflects its intention:

[T]raditional districting principles such as compactness, contiguity, and respect for political subdivisions ... are important ... because they are objective factors that may serve to defeat a claim that a district has been gerrymandered on racial lines .... Put differently, we believe that reapportionment is one area in which appearances do matter.

Shaw I, 509 U.S. at 647, 113 S.Ct. 2816 (emphasis added). Therefore, we rely on the principal opinion in Bush, which stated that the “neglect of traditional districting criteria” is “necessary, [but] not sufficient” for strict scrutiny to apply. Bush, 517 U.S. at 962, 116 S.Ct. 1941 (principal opinion) (emphasis added); accord Miller, 515 U.S. at 928, 115 S.Ct. 2475 (O'Connor, J., concurring) (“To invoke strict scrutiny, a plaintiff must show that the State has relied on race in substantial disregard of customary and traditional districting practices.”).

Our dissenting colleague advocates a different reading of predominance. The dissent views the 55% BVAP floor as a “filter through which all line-drawing decisions had to pass” and argues that this “racial filter necessarily ... rendered all traditional criteria that otherwise would have been ‘race-neutral,’ tainted by and subordinated to race.” Post at 575. According to the dissent, “a legislative district necessarily is crafted ‘because of race ’ ” when such a filter is employed. Post at 575–76 (emphasis added). The dissent takes the view that the “application of strict scrutiny in this suit was never a close question” because when the legislators “intentionally created [55% BVAP] districts,” this “was sufficient to show that race was a predominant factor in its redistricting.” Bush, 517 U.S. at 999–1000, 116 S.Ct. 1941 (Thomas, J., concurring in the judgment). We respectfully decline to adopt this reading of predominance.

First, the dissent's interpretation echoes the view that was rejected by the principal opinion in Bush v. Vera. See id. at 962, 116 S.Ct. 1941 (principal opinion). In his separate Bush concurrence, Justice Thomas wrote:

In my view, [the intentional creation of a 50% BVAP district] means that the legislature affirmatively undertakes to create a majority-minority district that would not have existed but for the express use of racial classifications—in other words, that a majority-minority district is created “because of,” and not merely “in spite of,” racial demographics. When that occurs, traditional race-neutral districting principles are necessarily subordinated (and race necessarily predominates), and the legislature has classified persons on the basis of race. The resulting redistricting must be viewed as a racial gerrymander.

Id. at 1001, 116 S.Ct. 1941 (Thomas, J., concurring in the judgment) (internal citations omitted) (emphasis added). Although Justice Thomas recognized that this question was “expressly reserved” in Shaw I, he believed that the Court had “effectively resolved it in subsequent cases.” Id. at 999, 116 S.Ct. 1941.

Justice Thomas first pointed to the Supreme Court's decision in Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995), as evidence that “all governmental racial classifications must be strictly scrutinized.” Id. at 999–1000, 116 S.Ct. 1941. But this presumes what must in fact be proven: that the Virginia legislature's facially neutral redistricting legislation was the legal equivalent of a facially racial classification. Predominance is itself the arbiter of this legal equivalency.

In Adarand, the question was whether a contracting clause providing “financial incentive[s] to hire subcontractors controlled by ‘socially and economically disadvantaged individuals' ... violates the equal protection component of the Fifth Amendment's Due Process Clause.” 515 U.S. at 204, 115 S.Ct. 2097. In that case, federal law required the use of the clause in most federal agency contracts, and expressly “require[d] the clause to state that ‘[t]he contractor shall presume that socially and economically disadvantaged individuals include Black Americans, Hispanic Americans, Native Americans, Asian Pacific Americans, and other minorities[.]’ ” Id. at 205, 115 S.Ct. 2097.

The dissent retreads this path by citing to City of Richmond v. J.A. Croson Co., 488 U.S. 469, 109 S.Ct. 706, 102 L.Ed.2d 854 (1989). As in Adarand, the Croson Court was faced with a city ordinance expressly requiring contractors to subcontract at least 30% of their work on city contracts to “Minority Business Enterprises” owned and controlled by “[c]itizens of the United States who are Blacks, Spanish-speaking, Orientals, Indians, Eskimos, or Aleuts.” Croson, 488 U.S. at 477–78, 109 S.Ct. 706.

We have no doubt that strict scrutiny is applied to all express racial classifications, but neither Adarand nor Croson help light our path to interpreting predominance. Adarand itself explicitly disclaimed any application to facially neutral legislation, stating that “this case concerns only classifications based explicitly on race, and presents none of the additional difficulties posed by laws that, although facially race neutral, result in racially disproportionate impact and are motivated by a racially discriminatory purpose.” Adarand, 515 U.S. at 213, 115 S.Ct. 2097 (emphasis added).

Justice Thomas next pointed to Miller and argued that the State's “concession that it intentionally created [50% BVAP] districts was sufficient to show that race was a predominant, motivating factor in its redistricting.” Bush, 517 U.S. at 1000, 116 S.Ct. 1941 (Thomas, J., concurring in the judgment). The dissent also relies upon Miller to argue that strict scrutiny is warranted when a legislature is “motivated by,” rather than merely “conscious of,” race in its districting. See post at 571. But this demands the impossible. We cannot ask legislators to accidentally wander into compliance with the VRA, and Miller cannot be read to invoke strict scrutiny whenever legislators intentionally create a district with a predetermined BVAP floor.

In Miller, there was considerable evidence showing “that the General Assembly was motivated by a predominant, overriding desire to assign black populations to the Eleventh District and thereby permit the creation of a third majority-black district.” 515 U.S. at 917, 115 S.Ct. 2475. It was the State's overriding assignment of voters on the basis of race, rather than other districting criteria, that made the third majority-minority district constitutionally offensive. If Miller stood for the proposition that the intentional creation of a 50% BVAP district alone constituted “predominance,” then all three majority-minority districts would have constituted racial gerrymanders. Instead, the opinion focused on the Eleventh District, which was a geographic “monstrosity” and required the State to add lengthy appendages, split precincts, and abandon “all reasonable standards of compactness and contiguity.” Id. at 909, 917–19, 115 S.Ct. 2475.

The Miller decision does, of course, recognize that “statutes are subject to strict scrutiny under the Equal Protection Clause not just when they contain express racial classifications, but also when, though race neutral on their face, they are motivated by a racial purpose or object.” 515 U.S. at 913, 115 S.Ct. 2475. But it is Miller's subordination test itself that mans the floodgates to ensure that the predominance exception to traditional facial classification jurisprudence does not swamp the standing rule that Equal Protection Clause claims against facially neutral statutes usually require plaintiffs to prove discriminatory purpose and discriminatory effect.

Subordination in the enacted plan (rather than subordination of hypothetical plans) is required because a map that reflects neutral conventions on its face eliminates the assumption of expressive and representative harm found in Shaw I without necessarily imposing any other constitutionally cognizable harms in its stead. The Supreme Court recognized as much in Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978).

In Bakke, the Supreme Court struck down a higher education admissions program that reserved a specific number of seats for minority applicants. See 438 U.S. at 275, 98 S.Ct. 2733. The problem with this scheme was that it “prefer[red] the designated minority groups at the expense of other individuals who [were] totally foreclosed from competition for the 16 special admissions seats[.]” Id. at 305, 98 S.Ct. 2733 (opinion of Powell, J.) (emphasis added). As Justice Powell wrote, “[w]hen a classification denies an individual opportunities or benefits enjoyed by others solely because of his race or ethnic background, it must be regarded as suspect.” Id.

Justice Powell contrasted this holding with the Supreme Court's holding the previous year in United Jewish Organizations v. Carey (UJO), 430 U.S. 144, 97 S.Ct. 996, 51 L.Ed.2d 229 (1977). In UJO, the State of New York had redrawn its voting districts “to enhance the electoral power of certain ‘nonwhite’ voters” and “meet [the] objections of the [DOJ] under § 5 of the Voting Rights Act [.]” Bakke, 438 U.S. at 304–05, 98 S.Ct. 2733 (opinion of Powell, J.). The Supreme Court affirmed the plan. According to Justice Powell, UJO was distinguishable “as a case in which the remedy for an administrative finding of discrimination encompassed measures to improve the previously disadvantaged group's ability to participate, without excluding individuals belonging to any other group from enjoyment of the relevant opportunity—meaningful participation in the electoral process.Id. at 305, 98 S.Ct. 2733 (emphasis added). When a legislature crafts a plan that reflects traditional, neutral, districting conventions and does not intentionally dilute any group's meaningful participation in the electoral process, there is no constitutionally cognizable offense to be found. The use of a quota does not change this. See UJO, 430 U.S. at 162, 97 S.Ct. 996 (principal opinion) (“[A] reapportionment cannot violate the Fourteenth or Fifteenth Amendment merely because a State uses specific numerical quotas in establishing a certain number of black majority districts. Our cases under [Section] 5 stand for at least this much.”).

Justice Powell also emphasized that Congress has “special competence ... to make findings with respect to the effects of identified past discrimination” and special “discretionary authority to take appropriate remedial measures.” Bakke, 438 U.S. at 302 n. 41, 98 S.Ct. 2733 (opinion of Powell, J.). This too distinguishes the case at hand from those cases wherein a school or municipality, acting on its own impulse, employs a racial quota.

From this vantage, the second problem with the dissent's reading comes into view: an interpretation of predominance that ignores “discriminatory effect” and deploys strict scrutiny when a neutral statute is adopted “because of” race-based motives would allow claims to proceed on “racial purpose” alone. Such an interpretation raises vexatious justiciability and balance of powers questions.

A redistricting plan struck down “solely because of the motivations of the men who voted for it” regardless “of its facial content or effect ... would presumably be valid as soon as the legislature or relevant governing body repassed it for different reasons.” See Palmer v. Thompson, 403 U.S. 217, 224–25, 91 S.Ct. 1940, 29 L.Ed.2d 438 (1971). That is because the offense is not in the legislative content of the enactment but only in the mental content of the legislators. Although divining the amalgamated motivations of an entire legislature may be tolerable when a showing of discriminatory effect further girds the inquiry, a “purpose only” equal protection claim would require courts to rest judgment upon the thoughts of a coequal branch alone.

We decline to take that path. As Chief Justice Burger once wrote,

The seductive plausibility of single steps in a chain of evolutionary development of a legal rule is often not perceived until a third, fourth, or fifth ‘logical’ extension occurs. Each step, when taken, appeared a reasonable step in relation to that which preceded it, although the aggregate or end result is one that would never have been seriously considered in the first instance. This kind of gestative propensity calls for the ‘line drawing’ familiar in the judicial, as in the legislative process: ‘thus far but not beyond.’

United States v. 12 200–Foot Reels of Super 8mm. Film, 413 U.S. 123, 127, 93 S.Ct. 2665, 37 L.Ed.2d 500 (1973). The dissent's interpretation might be a logical step in the evolution of the equal protection “predominance” test. But we think it would be one step too far. Predominance requires that racial considerations manifest in the enacted plan itself through the actual subordination of other districting criteria. That determination cannot be made without examining the respective roles of both race and the other redistricting factors in the actual plan before the Court. For the foregoing reasons, we reject the invitation to read the unwritten use of a 55% BVAP floor as a per se satisfaction of the predominance inquiry in a racial sorting claim. Of course, evidence of such thresholds is still significant when examining those districts that exhibit deviations from traditional, neutral districting principles. See Easley v. Cromartie (Cromartie II), 532 U.S. 234, 254, 121 S.Ct. 1452, 149 L.Ed.2d 430 (2001) (noting that the use of a 50% racial threshold was “significant” evidence in Bush and Miller ); Page, 2015 WL 3604029 at *35 (Payne, J., dissenting) (noting the significance in Shaw v. Hunt (Shaw II), 517 U.S. 899, 116 S.Ct. 1894, 135 L.Ed.2d 207 (1996), of a concession by the State to create two districts with 50% BVAP thresholds). Shaw II, for example, recognized that racial deviations from neutral principles cannot be saved by later resort to non-racial explanations. See 517 U.S. at 907, 116 S.Ct. 1894.

According to the dissent, Shaw II compels a finding of predominance whenever non-racial factors are only considered “consistent with the racial objective.” Post at 572. But the district at issue in Shaw II was “highly irregular and geographically non-compact by any objective standard that can be conceived.” Shaw II, 517 U.S. at 905–06, 116 S.Ct. 1894. Simply put, the Shaw II Court was faced with a situation wherein some “race-neutral” goals—such as partisan balance—could still be partially advanced despite the qualitative predominance of race, but it was not faced with a situation wherein racial districting goals posed no conflict with neutral districting criteria whatsoever.

Moreover, the author of Shaw II, Chief Justice Rehnquist, joined the principal opinion issued the same day in Bush v. Vera, suggesting that these two opinions can—and should—be read in harmony. The Bush opinion joined by Chief Justice Rehnquist explicitly rejected the interpretation that the dissent now attributes to his opinion in Shaw II.

We adopt a reading consistent with Shaw II, as evidenced by our finding of racial predominance in HD 75. A State cannot district predominantly on the basis of race and then insulate such racial line drawing by pointing to other non-racial goals advanced by the racial sort.

Alabama, like its predecessors in the Shaw–Miller line, holds that racial thresholds constitute evidence, not dispositive proof, of racial predominance. If the thresholds employed by the legislators crafting the bill do not manifest in the formation of the enacted district, then there is no facial classification equivalent upon which to rest Shaw's “analytically distinct” framework.

If one strict predominance rule were not enough, Intervenors advance a counter-theory that they claim is derived from Alabama. As the Intervenors stated during their closing argument:

“[T]he question you must answer to get to strict scrutiny ... is whether the use of race resulted in any district which violated Virginia law or traditional redistricting criteria of the state, or, as the state did here, their specifically adopted criteria.”

Trial Tr. 16:8–13 (Intervenors). Intervenors drew the Court's attention to a passage in the Alabama decision where the Court “talk[ed] about [the State] transgressing its own state guidelines, its own state criteria.” Id. at 853:15–854:9. And so it did:

There is considerable evidence that [the racial thresholds] had a direct and significant impact on the drawing of at least some of District 26's boundaries .... Transgressing their own redistrictinq guidelines, the drafters split seven precincts between the majority-black

District 26 and the majority-white District 25, with the population in those precincts clearly divided on racial lines.

Alabama, 135 S.Ct. at 1272 (emphasis added). But, as is clear from the cited passage, the drafters' transgression of their own redistricting guidelines—like their informal use of a racial threshold—is evidence of predominance, not dispositive proof. That is because “subordination” is not the same as a “violation” or “transgression.” Subordination requires a balancing of degree to determine whether non-racial criteria or racial criteria predominated.

For example, it is difficult to understand what a “transgression” of “compactness” would even entail. Compactness, like temperature, falls along a range, and there is no professional consensus about what degree of departure (from any of more than twenty measures) is enough to say a district is “not compact.” Trial Tr. 716:15–18 (Hofeller).

One of Intervenors' experts, for example, found “no issues” with every last one of the Challenged Districts, Trial Tr. 708:15–709:21 (Hofeller), despite testifying that there is no professional consensus on what is and is not compact. Id. at 716:10–18. Meanwhile, Plaintiffs' expert found some of the districts “not compact” based upon a .20 Reock “rule of thumb,” Pls.' Ex. 50 at 18, that other experts disputed as having any meaningful basis, Trial Tr. 716:5–25 (Hofeller).

More importantly, the “traditional” criteria discussed in the ShawMiller cases are informed by, but not defined by, state law. Rendering the predominance inquiry subject to state law would make the existence of a federal constitutional claim dependent upon an individual state's resolutions, statutes, or constitution.

The determinative question is not whether a State's individualized districting requirements are “violated,” but whether traditional, neutral districting criteria and other districting criteria have been generally “subordinated” to racial considerations on the whole. See Page, 2015 WL 3604029 at *11 (“To show that race predominated, Plaintiffs need not establish that the legislature disregarded every traditional districting principle.”). A State's violation of, or departure from, its own stated criteria can constitute evidence in the predominance analysis, but Alabama does not require that the State do so in order to make out a racial sorting claim. Intervenors' proposed interpretation is, accordingly, rejected.

1. Predominance Analysis

As common courtesy holds, one should not shoot down a suggestion without offering an approach to replace it. Although “predominance,” “subordination,” “dilution,” and “retrogression” are all standards not amenable to hard rules or safe harbors, the Court does have an obligation to the parties to explain its reasoning as clearly and definitively as possible. Therefore, the Court will walk through each of the steps of the analytical framework that it has applied to arrive at its conclusions with respect to the Challenged Districts.

A racial sorting claim is “one area in which appearances do matter.” Shaw I, 509 U.S. at 647, 113 S.Ct. 2816. Because a district must exhibit “substantial disregard of customary and traditional districting practices” in order to animate the racial sorting doctrine's central concern with facial classification, Miller, 515 U.S. at 928, 115 S.Ct. 2475 (O'Connor, J., concurring), the Court will evaluate each Challenged District for “subordination” in three steps.

First, the Court will review the district on the basis of its compliance with traditional, neutral districting criteria, including, but not limited to, compactness, contiguity, nesting, and adherence to boundaries provided by political subdivisions and natural geographic features.

Second, the Court will examine those aspects of the Challenged District that appear to constitute “deviations” from neutral criteria. These may be particular, isolated areas along the district's boundary, or—on occasion—the district itself may seem facially questionable. Based on the evidence submitted and testimony provided, the Court will examine the record to ascertain the underlying rationale for those deviations. In determining the reasons for deviations from the traditional neutral criteria, it will be necessary to determine whether a deviation was caused in part or entirely by the need to comply with the one-person, one-vote precepts or by political circumstances such as protection of incumbents.

Of course, evidence of compliance with equal population goals is not weighed against evidence of racial consideration, but it may be important in determining why a district appears to deviate from neutral criteria.

Third, the Court will weigh the totality of the evidence and determine whether racial considerations qualitatively subordinated all other non-racial districting criteria.

a. Neutrality

A racial sorting claim requires the Court find that the State subordinated traditional, neutral criteria, and other non-racial districting criteria to racial considerations. Traditional districting principles include, inter alia, compactness, contiguity, respect for political subdivisions, and communities “defined by actual shared interests.” See Miller, 515 U.S. at 916, 115 S.Ct. 2475; Shaw I, 509 U.S. at 647, 113 S.Ct. 2816. These conventions neutrally advance the values inherent in a geographic—rather than proportional—system of representation, such as responsiveness, accountability, familiarity, ease of access, ease of administration, and political engagement.

The specific traditional criteria outlined in Miller and Shaw are not constitutionally required. See Shaw I, 509 U.S. at 647, 113 S.Ct. 2816; Gaffney v. Cummings, 412 U.S. 735, 752 n. 18, 93 S.Ct. 2321, 37 L.Ed.2d 298 (1973) (“[C]ompactness or attractiveness has never been held to constitute an independent federal constitutional requirement for state legislative districts.”). Rather, these criteria are important because they reflect the neutrality that is central to a redistricting statute that complies with the Equal Protection Clause. See Reynolds, 377 U.S. at 558, 84 S.Ct. 1362. Traditional, neutral conventions are important to evaluate in a racial gerrymandering claim “because they are objective factors that may serve to defeat a claim that a district has been gerrymandered on racial lines.” Shaw I, 509 U.S. at 647, 113 S.Ct. 2816 (emphasis added).

“[T]he concept of equal protection has been traditionally viewed as requiring the uniform treatment of persons standing in the same relation to the governmental action questioned or challenged. With respect to the allocation of legislative representation, all voters, as citizens of a State, stand in the same relation.... Any suggested criteria for the differentiation of citizens are insufficient to justify any discrimination, as to the weight of their votes, unless relevant to the permissible purposes of legislative apportionment.”

Of course, states may continue to develop new neutral districting principles, and a State's consistent adherence thereto would also be considered an objective factor to help defeat a claim of gerrymandering. Existing traditional districting conventions “evolved over the years through the political process” itself. Bush, 517 U.S. at 1073, 116 S.Ct. 1941 (Souter, J., dissenting). What renders these guiding principles important for redistricting purposes is that they observe and advance neutral democratic values.

The fact that a district deviates from neutral criteria on its face does not, however, mean that those deviations were racially motivated. Other, non-racial districting criteria may also be used to defeat a claim of racial gerrymandering by demonstrating that the district's deviations from neutral criteria are attributable to race-neutral motives. Chief among these are political and incumbency considerations. See Alabama, 135 S.Ct. at 1270.

During the first stage of the predominance inquiry, the Court examines whether the redistricting legislation—on its face—raises questions about the use of discriminatory, individualized criteria (such as race, politics, or incumbency) or whether it appears to be predominantly explainable on the basis of traditional, neutral, geographic criteria (such as compactness, contiguity, or respect for political subdivisions).

In reviewing the Challenged Districts, the Court will consider neutral criteria in the following manner:

i. Compactness

As Justice Stevens stated in Karcher v. Daggett, “geographical compactness serves independent values; it facilitates political organization, electoral campaigning, and constituent representation.” 462 U.S. 725, 756, 103 S.Ct. 2653, 77 L.Ed.2d 133 (1983) (Stevens, J., concurring). Although “non-compact” districts may sometimes be necessary to serve these values—such as when a “major transport corridor might ... minimum[ize] travel time for a representative to travel around the district”—“drastic departures from compactness are a signal that something may be amiss.” Id. at 758, n. 20, 103 S.Ct. 2653.

Yet, compactness is surprisingly ethereal given its seemingly universal acceptance as a guiding principle for districting. All of the expert testimony provided reveals one deep conceptual dilemma: no one can agree what it is or, as a result, how to measure it. See, e.g., Trial Tr. 535:19–536:8 (Katz). There are “at least 20” measures, not one of which can claim any greater legitimacy than its peers. Id. at 555:16–17. The Reock test measures geographical dispersion and therefore is sensitive to—and its scoring punishes—elongated districts. Id. at 136:13–23 (Ansolabehere). The Polsby–Popper test measures perimeter dispersion and therefore is sensitive to—and its scoring punishes—oddly shaped district boundaries with large numbers of indentations. Id. Meanwhile, the Schwartzberg test looks at “a normalized standard deviation of the distance from every point to the center of the district,” id. at 558:4–7 (Katz), and the Boyce–Clark test measures the “center of inertia” or “how far is the farthest voter from the center of the district,” id. at 537:12–538:6. One notable political scientist has quipped that all of these measures are just variants of “the intraocular test”: “people look at distric[t] maps, they figure out which districts they think look ugly, and then they choose the compactness measure which comports with their eyeball view of the mapping.” Id. at 542:14–24 (Katz). See also id. at 697:20–698:9 (Hofeller) (noting that “the main measurement of compactness ... while you are drawing a map is to look at the shapes of districts, so-called eyeball test”).

But compactness is not important for its own sake. Rather, compactness is important because it serves certain values of geographic representation. Therefore, the “major transportation corridor” district discussed by Justice Stevens would fare poorly on the Reock metric, but would serve its purposes in a manner that might be reflected by another measure (such as driving time). Meanwhile, a district that adheres to highly irregular county lines, id. at 559:18–21 (Katz); 687:1–4 (Hofeller), or easily identifiable geographic features, id. at 538:14–19 (Katz); 687:1–4 (Hofeller), might score poorly on the Polsby–Popper test, but would enhance the values served by those neutral criteria, as discussed below. If the price of advancing these other neutral criteria is compactness, then the cost is not a judicial concern.

Nor does a district's “absolute” compactness score matter so much as its “relative” score. The Court's examination of a district's compactness measure may be informed by the average in the State (which is important to take account of a State's inalterable features), see Ints.' Ex. 14 at 12 (discussing Virginia's irregular shape, county lines, and geographic features), may be informed by the average in the nation (which is important to take account where a State's own averages may be far above or far below the national average), see Page, 2015 WL 3604029 at *33 (“A highly compact district in a state that adheres closely to compactness principles may be both the least compact in the state and among the most compact in the nation.”) (Payne, J., dissenting), and may be informed by historical averages (which is important to account for trends in compactness over several districting cycles), see Trial Tr. 560:2–10 (Katz) (noting it is “perfectly reasonable” to use compactness measures “in comparing two maps for the same state”). These are all factors that courts must consider when evaluating this criterion.

In short, the Court would be remiss to look at compactness scores in a vacuum, but that does not render them useless as evaluative tools in the predominance inquiry. The key is not “absolute” compactness, “relative” compactness, or even a State's adherence to its own constitutional or statutory compactness definitions (although these may be illuminating); rather, the key is whether compactness deviations are attributable to something meaningful, such as other neutral criteria or a legitimate use of non-neutral criteria. As Dr. Hofeller stated at trial, echoing Justice Stevens' sage advice, compactness is “more like a flag than a conclusion.” Trial Tr. 684:17–18 (Hofeller).

Virginia's constitutional compactness requirement only demands that districts not be “clearly erroneous, arbitrary, or wholly unwarranted.” Wilkins v. West, 264 Va. 447, 465–66, 571 S.E.2d 100 (Va.2002). That standard informs the Court's inquiry, but does not resolve it.

ii. Contiguity

Contiguity, like compactness, serves important democratic purposes, binding geographic communities together and helping to enable effective representation. In upholding a district under the Virginia constitution's contiguity provision despite its division by water, the Supreme Court of Virginia reflected upon this raison d'être:

Although the record shows that travel between [some] precincts and the remainder of the district requires travel through another district, there is nothing in this record showing that such access is unreasonable, unduly burdensome, or adversely impacts the ability of residents to secure meaningful representation of their interests or effective communication with their elected representative.

Wilkins v. West, 264 Va. 447, 465–66, 571 S.E.2d 100 (Va.2002). As the Page court reminded, “contiguity and other traditional districting principles are ‘important not because they are constitutionally required,’ but rather ‘because they are objective factors' courts may consider in assessing racial gerrymandering claims.” 2015 WL 3604029 at *11.

A district split by water has not “violated” contiguity for the purposes of a racial sorting claim any more than a district connected by a single point on land has “respected” contiguity. See Shaw I, 509 U.S. at 636, 113 S.Ct. 2816 (noting that one of the districts in that case “remain[ed] contiguous only because it intersect[ed] at a single point with two other districts before crossing over them”). As with compactness, contiguity admits of degrees. Districts that are not divided by water are more contiguous than those that are, and districts that are at least connected by a water crossing—such as a bridge—are more contiguous than districts that are not. Land contiguity is important not because it is determinative, but because it reflects the common understanding that bodies of water may mark the natural divide between communities of interest or constitute barriers to the effective function of democratic activities.

As one Norfolk resident put it during the legislature's public hearings: “Please deep six this specious concept of contiguity by water. To put [these communities] in the same district ... is patently ridiculous.” Pls.' Ex. 3 at 36:8–11.

Of course, deviations from land contiguity may also reflect adherence to other neutral districting criteria. Many cities lie across rivers or around harbors and, indeed, are built outward from the central focal point of the community: the waterfront. In such cases, a body of water that “divides” a community may actually be the primary factor that unites it. In other words, a “deviation” from “contiguity” standards may be an attempt to respect a distinct community of interest or political subdivision. The subordination of contiguity conventions is, like compactness, simply a factor that the Court must consider in conducting its predominance analysis.

iii. Political Subdivisions

A common and significant neutral districting criterion is respect for political subdivisions, such as counties or cities. “Subdivision boundaries tend to remain stable over time. Residents of political units such as townships, cities, and counties often develop a community of interest, particularly when the subdivision plays an important role in the provision of governmental services.” Karcher, 462 U.S. at 758, 103 S.Ct. 2653 (Stevens, J., concurring). Moreover, adherence to subdivision boundaries can facilitate civic engagement, enhance democratic accountability, and increase administrative convenience. See id. (“[L]egislative districts that do not cross subdivision boundaries are administratively convenient and less likely to confuse the voters.”); id. at 787 n. 3, 103 S.Ct. 2653 (Powell, J., dissenting). As Justice Powell once wrote:

Most voters know what city and county they live in, but fewer are likely to know what [legislative] district they live in if the districts split counties and cities. If a voter knows his [legislative] district, he is more likely to know who his representative is. This presumably would lead to more informed voting. It also is likely to lead to a representative who knows the needs of his district and is more responsive to them.

Id. at 787 n. 3, 103 S.Ct. 2653 (Powell, J., dissenting) (internal citations and quotation marks omitted).

When a legislative district is “nothing more than an artificial unit divorced from, and indeed often in conflict with, the various communities established in the State,” legislators cannot represent their constituents properly and voters cannot exercise the ballot intelligently. Id. at 787, 103 S.Ct. 2653 (Powell, J., dissenting). A report produced by the Governor's Commission distilled the overarching themes that were repeatedly voiced during its public forums from around the Commonwealth. As the Commission noted, “the splitting of municipal and county jurisdictions drew the ire of citizens, who ... pointed out the difficulties that citizens have in knowing who to contact, who to hold accountable, and who among several legislators should coordinate or lead the representation of local city and county interests in the General Assembly.” Pls.' Ex. 23 at 8.

In evaluating whether neutral criteria were subordinated, a legislature's adherence to city and county boundaries provides an important reference point for courts undertaking the predominance analysis. Of course, the legislature may, and often will, need to deviate from political subdivision borders to comply with federal- or state-mandated population constraints. In such situations, the Court will look to whether another neutral criterion—such as compactness, geographic boundaries, precinct boundaries, or communities of interest—helps to explain the method of departure. In this manner, neutral criteria can often form a “backstop” for one another when one criterion cannot be fully satisfied, thus ensuring that neutral criteria are still predominating in the balance.

vi. Natural Geography

Geographic features, such as mountains ranges or rivers, may also be used to provide a neutral boundary during the districting process. Oftentimes, these geographic indicators mark the boundaries of distinct communities of interest or can provide a point of reference for voters, candidates, and representatives. In many cases, these natural boundaries may already constitute the basis for governmental subdivision lines. See, e.g., Ints.' Ex. 14 at 12 (noting that, in Virginia, “[m]any county lines follow riverbeds, and the State's western boundary runs along 400 miles of mountain ridges and rivers”).

Over time, artificial geography may also come to play a similar role. Major transportation thoroughfares may slowly generate distinct communities of interest on either side of the divide, or the marker may be used as a useful reference point for voters, candidates, and representatives seeking to understand their own district's boundaries. These are important factors to consider, especially when adherence to traditional subdivision lines is not possible.

v. Nesting

Nesting refers to the practice of putting two or more districts of the lower chamber of the state legislature wholly within each district of the upper chamber. “By permitting voters readily to identify their voting districts and corresponding representatives, a nested plan can be expected to foster voter participation.” Bandemer, 478 U.S. at 179 n. 18, 106 S.Ct. 2797 (Powell, J., concurring in part, dissenting in part). Nesting may result in a House district boundary that appears inexplicable by neutral criteria until the corresponding Senate district is laid atop.

vi. Precincts

Precincts and Voting Tabulation Districts (“VTDs”) are often the smallest objectively identifiable geographic groupings that legislators use to organize legislative districts. They may occasionally correspond to towns, neighborhoods, or other identifiable communities of interest, but they are not “governmental jurisdictions” in their own right. Trial Tr. 234:11–16 (Ansolabehere); 605:4 (Hood). In Virginia, VTDs generally correspond to voting precincts. Id. at 253:14–17 (Ansolabehere).

Given their small size, compliance with precinct or VTD boundaries alone will rarely be sufficient to show adherence to neutral criteria. This is because VTDs can easily be strung together into grotesque formations having little regard for compactness, contiguity, political subdivisions, or other important neutral criteria advancing democratic values. In short, a district could avoid splitting any VTDs but remain highly suspicious on its face.

For these same reasons, however, VTD splits will often provide a flag for further inquiry. The unexplained splitting of several VTDs in a single district can call into question the criteria guiding that district's construction.

vii. Communities of Interest

Among traditional, neutral districting principles, the concept of respecting “communities of interest” is the most enigmatic. On the one hand, respect for such communities is often considered the guiding light of the other neutral principles. On the other hand, defining some “communities of interest” may involve straddling the fence between neutral and discriminatory criteria. For example, communities of interest may be defined by relatively objective factors, such as service delivery areas, media markets, or major transit lines. Similarly, communities may be somewhat objectively characterized as rural, suburban, or urban. These can be valid neutral criteria, assuming that legislators actually have access to this information and rely upon it. See Bush, 517 U.S. at 953, 116 S.Ct. 1941 (principal opinion) (discounting argument that legislature relied upon “urban character,” “shared media sources,” and “major transportation lines” because the “supporting data were largely unavailable to the legislature before the district was created” and the factors did not possess “the same degree of correlation to district lines that racial data exhibit”).

The “communities of interest” criterion becomes less neutral, however, when one considers “cultural,” “social,” or “religious” communities of interest. This tendency to morph into a more individualized metric explains the Miller Court's qualification that traditional districting principles include “respect for ... communities defined by actual shared interests.” 515 U.S. at 916, 115 S.Ct. 2475. To give effect to this elusive delineation, it is important to have demonstrable evidence of shared interest when the boundaries cannot be explained on an objective or neutral basis.

viii. State Criteria

For the reasons discussed above, a plaintiff does not need to prove that a State “violated” its own districting criteria in order to prove predominance. A State's deviation from its own constitutional, statutory, or adopted criteria does, however, constitute evidence that is probative of subordination.

b. Deviations

If the Challenged Districts, or significant parts of the Challenged Districts, appear inexplicable by reference to the consistent application of traditional, neutral principles, then the Court will examine the basis for those departures. Deviations from neutral criteria signal the presence of potential subordination and lay the foundation for the sorting claim; namely, that the districts reflect racial classifications of individual voters and do not constitute neutral, geographic representative units.

The Supreme Court has cited several sources of direct and circumstantial evidence that courts can rely upon in identifying racial deviations, including:

[S]tatements by legislators indicating that race was a predominant factor in redistricting; evidence that race or percentage of race within a district was the single redistricting criterion that could not be compromised; ... use of land bridges in a deliberate attempt to bring African–American population into a district; and creation of districts that exhibit disregard for city limits, local election precincts, and voting tabulation districts.

Page, 2015 WL 3604029, at *7 (internal citations omitted).

Because traditional, neutral principles advance fundamental democratic values and neutral state interests, districts that substantially disregard these principles can “caus[e] a severe disruption of traditional forms of political activity.” Bush, 517 U.S. at 974, 116 S.Ct. 1941 (principal opinion). In Bush v. Vera, Justice O'Connor described the impact that such districts can have:

Campaigners seeking to visit their constituents “had to carry a map to identify the district lines, because so often the borders would move from block to block”; voters “did not know the candidates running for office” because they did not know which district they lived in. In light of [the State's] requirement that voting be arranged by precinct, with each precinct representing a community that shares local, state, and federal representatives, it also created administrative headaches for local election officials[.]

Id. at 974, 116 S.Ct. 1941. Such complaints have been echoed by local election officials in Virginia who “end up taking the brunt of complaints from voters who can't understand why they can't vote in their old precinct, why they can't find any of their current office holders on the ballot, and why they are in the same district as a relative who lives nowhere near them[.]” Pls.' Ex. 26 at 17:6–18.

Of course, the presence of identifiable deviations alone does not satisfy the predominance inquiry because “subordination” requires “substantial disregard” for traditional, neutral districting criteria. The substantiality of any identified deviations—and whether it is sufficient to support a finding of predominance—is examined when the Court weighs the evidence as a whole in the final stage.

In reviewing the Challenged Districts, the Court will consider evidence bearing on legislators' bases for the deviations. Deviations may be attributed to any number of considerations, but legislators typically rely upon the following: population equality, race, political affiliation or preference, and incumbency. The Court will evaluate these bases for deviation in the following manner:

i. Population

“[A]n equal population goal ... is part of the redistricting background, taken as a given, when determining whether race, or other factors, predominate in a legislator's determination as to how equal population objectives will be met.” Alabama, 135 S.Ct. at 1270. Thus, achievement of the population goal is not a traditional redistricting factor that is considered in the balancing that determines predominance. However, the requirement to comply with federally imposed population goals is relevant to assessing why a district may appear to deviate from neutral criteria. This is particularly true where the census data shows significant losses or gains of population in certain geographic areas of a State.

The Court's analysis does not change just because the State has decided to adopt a lower percentage deviation threshold than constitutionally required. In Alabama, the legislature adopted “a more rigorous deviation standard than our precedents have found necessary under the Constitution.” Id. at 1263. There, as here, it seems that “[c]ompliance with these two goals”—BVAP targets and a ±1% population deviation rule—“posed particular difficulties with respect to ... the State's ... majority-minority districts[.]” Id. But “legislative efforts to create districts of approximately equal population” more stringent than the 5% deviation held generally permissible in Brown v. Thomson, 462 U.S. 835, 842, 103 S.Ct. 2690, 77 L.Ed.2d 214 (1983), cannot explain away deviations from neutral principles. Id. at 1270. The predominance inquiry examines the basis upon which voters were sorted into appropriately apportioned districts. Id. at 1271. Where apportionment by political subdivision must be sacrificed to equal population goals, for example, other neutral principles such as compactness and precinct boundaries can often pick up the slack. A substantial deviation from neutral principles, therefore, only admits of answer by other, non-neutral criteria, such as race or political affiliation.

Nor can the fact that a benchmark district possesses “almost exactly the right amount of population,” Trial Tr. 147:19–148:19, (Ansolabehere), taken alone, provide evidence that changes to the district were based on race. If adequately populated districts were presumptively required to stay the same, the remaining districts on the map would need to wrap around them in violation of neutral principles. Id. at 688:20–689:10 (Hofeller). Of course, if a district exhibits unexplained deviations from neutral principles and the population changes for that district reflect “remarkable feats” of racial math, then this would constitute strong evidence that race predominated in the drawing of the district boundaries. See Alabama, 135 S.Ct. at 1271 (noting that “[o]f the 15,785 individuals that the new redistricting laws added to the population of [the district], just 36 were white—a remarkable feat given the local demographics”).

ii. Racial Deviations

One explanation for a district's deviations from neutral districting criteria may be voters' race. The mere awareness or consideration of race by legislators in their districting decisions does not, on its own, provide sufficient evidence to support a claim of racial sorting under the Equal Protection Clause. Shaw I, 509 U.S. at 646, 113 S.Ct. 2816 (“[T]he legislature always is aware of race when it draws district lines, just as it is aware of age, economic status, religious and political persuasion, and a variety of other demographic factors. That sort of race consciousness does not lead inevitably to impermissible race discrimination.”). It takes more than consideration of race to prove that race predominated over traditional factors. Of course, if legislators' use of race entailed the subordination of other districting criteria, it must be adequately justified under the strict scrutiny regime.

iii. Political Deviations

Another explanation for a district's deviations from neutral districting criteria may be voters' political opinions, affiliations, and beliefs. As with race, the mere awareness or consideration of voters' political affiliation by legislators is both unavoidable and constitutionally permissible. Gaffney, 412 U.S. at 753–54, 93 S.Ct. 2321 (“It would be idle, we think, to contend that any political consideration taken into account in fashioning a reapportionment plan is sufficient to invalidate it.... The very essence of districting is to produce a different—a more ‘politically fair’—result.... Politics and political considerations are inseparable from districting and apportionment.”). Accordingly, districting on the basis of political affiliation may be a legitimate criterion for the legislature to consider. Alabama, 135 S.Ct. at 1270 (citing Bush for the proposition that legislators may rely on “political affiliation” in districting); Bush, 517 U.S. at 954–65, 116 S.Ct. 1941 (principal opinion) (citing Gaffney ). The Intervenors have raised the argument that some of the Challenged Districts have political, rather than racial, justifications.

However, deviations from neutral districting principles on the basis of political affiliation or preference may not always be constitutionally permissible. See Gaffney, 412 U.S. at 754, 93 S.Ct. 2321 (“What is done in so arranging for elections, or to achieve political ends or allocate political power, is not wholly exempt from judicial scrutiny under the Fourteenth Amendment.”); LULAC, 548 U.S. at 413–14, 126 S.Ct. 2594 (holding that political gerrymandering is unconstitutional); Arizona State Legislature v. Arizona Indep. Redistricting Comm'n, ––– U.S. ––––, 135 S.Ct. 2652, 2658, 192 L.Ed.2d 704 (2015) (reaffirming that “partisan gerrymanders ... are incompatible with democratic principles” and present justiciable claims) (internal brackets omitted). As in Page, the Plaintiffs have not raised the issue of political gerrymandering, and so this Court need not consider it further. See 2015 WL 3604029 at *20 n. 33 (Payne, J., dissenting).

See, e.g., Ints.' Pre–Trial Brief at 18 (“HD95 was crafted carefully to avoid taking HD94's Republican precincts and instead take Democratic-leaning population left behind by HD93 and reach into precincts surrounded by HD93 to dilute Democratic voting strength in that area.”); id. at 25 (“The changes on the eastern border to HD75 were drawn to load heavily Republican precincts into the district of Democrat William Barlow (who subsequently lost to a Republican in the 2011 election by 10 percentage points)[.]”).

iv. Incumbency Deviations

Yet another explanation for a district's deviations from neutral districting criteria may be incumbency considerations. In Gaffney v. Cummings, the Supreme Court observed that: “It would be idle, we think, to contend that any political consideration taken into account in fashioning a reapportionment plan is sufficient to invalidate it.... Redistricting may pit incumbents against one another or make very difficult the election of the most experienced legislator.” 412 U.S. at 753–54, 93 S.Ct. 2321. Accordingly, a district's impact on an incumbent may be a legitimate criterion for the legislature to consider. Alabama, 135 S.Ct. at 1270 (citing Bush for the proposition that legislators may consider “incumbency protection” in districting).

However, as with political deviations, deviations from neutral districting principles for incumbency purposes are not always permissible. In Bush, the Court recognized “incumbency protection, at least in the limited form of ‘avoiding contests between incumbent[s],’ as a legitimate state goal.” 517 U.S. at 964–65, 116 S.Ct. 1941 (principal opinion) (emphasis added). This state interest “aim[s] at maintaining existing relationships between incumbent congressmen and their constituents and preserv[es] the seniority the members of the State's delegation have achieved in the United States House of Representatives,” White v. Weiser, 412 U.S. 783, 792, 93 S.Ct. 2348, 37 L.Ed.2d 335 (1973), but does not necessarily invade the province of the voters. As the LULAC Court advised: “[I]ncumbency protection can be a legitimate factor in districting, but experience teaches that incumbency protection can take various forms, not all of them in the interests of the constituents.” 548 U.S. at 440–41, 126 S.Ct. 2594.

Here, the Intervenors allege that many of the Challenged Districts' deviations have “incumbency protection” justifications. See, e.g., Trial Tr. 825:5–7 (Intervenors) (“This was an incumbent-protection plan. That's the predominate motive of this plan[.]”). Some of these deviations reflect an interest in drawing district lines between incumbents' residences to avoiding pairing incumbents. See, e.g., id. at 304:6–21 (Jones). Other deviations, however, reveal an effort to fence in the incumbent's preferred voters or fence out the incumbent's detractors or challengers. See, e.g., id. at 325:19–326:23 (Jones). Whether this latter definition of “incumbency protection” states a legitimate government interest need not be decided here because no one has presented that issue. See Weiser, 412 U.S. at 792, 93 S.Ct. 2348.

That said, we share the dissent's concern over Intervenors' “implicit suggestion that approval by incumbent legislators” can somehow “rescue” a plan from a finding of racial predominance. Post at 576. We fully agree that “[t]he [VRA] and the Equal Protection Clause are intended to protect the rights of the individual voter, not to promote the self-interest of incumbents in majority-minority districts.” Post at 576. And, to be clear, the framework we adopt today condones no such thing. For example, if legislators attempt to “ ‘pac[k]’ minority voters into a particular majority-minority district for the purpose of protecting the incumbent,” post at 576 (emphasis added), this would still constitute racial sorting regardless of the “goal” of incumbency protection, see post at 543–44. This is precisely what we find occurred in HD 75, and we hold that race predominated accordingly. See post at 556–58.

On the other hand, if legislators attempt to pack supporters into their districts or attempt to remove detractors or challengers, then it could hardly be said that race drove the districting deviation. This does not imply that such actions are immune from constitutional challenge. Although the Supreme Court has only sanctioned a state interest in “incumbency pairing prevention,” the Plaintiffs simply did not raise any challenge to the Commonwealth's alleged interest in a wider definition of “incumbency protection.” Thus, we are in no position to decide that constitutional question.

Simply put, if incumbency interests constitute the predominate criterion driving the construction of the district, then a claim of racial gerrymandering must fail. That, however, does not imply that a claim of political gerrymandering would face a similar fate.

c. Weighing

The final step in the predominance inquiry of a racial sorting claim involves the weighing of the evidence in toto to determine whether the deviations attributable to race “predominate” over all other districting criteria employed by the legislature, including both neutral criteria and deviations attributable to non-racial motives. To demonstrate predominance, the Plaintiffs must show that the legislature “subordinated” or exhibited “substantial disregard” for these other criteria.

In making its predominance determination, the Court “must be sensitive to the complex interplay of forces that enter a legislature's redistricting calculus” and “exercise extraordinary caution.” Miller, 515 U.S. at 915–16, 115 S.Ct. 2475. “Federal-court review of districting legislation represents a serious intrusion on the most vital of local functions,” id. at 915, 115 S.Ct. 2475, and the Plaintiffs' burden is understandably “a demanding one,” id. at 928, 115 S.Ct. 2475 (O'Connor, J., concurring). Therefore, the redistricting enactments of a legislature are entitled to a presumption of correctness and good faith, and the burden is upon the plaintiff to dislodge that presumption. Id. at 915, 115 S.Ct. 2475 (majority opinion).

It should be noted, however, that the predominance balancing inquiry is qualitative rather than quantitative. In Miller, for example, the challenged district employed gangly arms at various points to capture black population centers, but the district's overall shape was not far from routine. See id. at 917, 115 S.Ct. 2475; id. at Appendix B. Looking at the complete picture, however, the district court found that “[r]ace was ... the predominant, overriding factor explaining the General Assembly's decision to attach to the [district] various appendages containing dense majority-black populations.” Id. at 920, 115 S.Ct. 2475.

In conducting the predominance balancing, two particular issues warrant the Court's careful attention.

i. Racial & Political Correlation

Occasionally, a deviation may appear equally explainable by racial or political motivations. Because the State is presumed to have acted lawfully and in good faith, the plaintiff must provide evidence that race—rather than politics—represented the primary basis for the classification. Evidence may include the sources of data relied upon in drawing the district, the use of fixed (or “aspirational”) political or racial targets or floors, and statements from legislators regarding the relative priority of their racial and political goals.

A political objective, however, does not immunize the use of race as a basis for classification because race cannot be used as a proxy for political characteristics, Bush, 517 U.S. at 968, 116 S.Ct. 1941 (principal opinion), even if there is a proven correlation between race and political preference in the state. This is because “to the extent that race is used as a proxy for political characteristics, a racial stereotype requiring strict scrutiny is in operation.” Id.

This is consistent with the Supreme Court's holding in Hunt v. Cromartie (Cromartie I). The lesson of Cromartie I was that a political classification would not be considered racial simply because the Democratic voters happened to be black. 526 U.S. 541, 542, 119 S.Ct. 1545, 143 L.Ed.2d 731 (1999) (“[A] jurisdiction may engage in constitutional political gerrymandering, even if it so happens that the most loyal Democrats happen to be black Democrats and even if those responsible for drawing the district are conscious of that fact.”). The lesson was not that a racial classification would be considered political simply because black voters happened to be Democrats.

In the latter scenario, the State still makes decisions about individuals based on the color of their skin. It is the act of using race as a proxy that constitutes an offensive stereotype. The fact that a stereotype might have some basis in fact—or is relied upon to achieve “non-racial” purposes—does not render it any less offensive.

Evidence of a racial floor will also lend support to the argument that race, rather than politics, can be attributed for particular deviations from neutral principles. Although such a floor will not result in per se predominance where a district is formed predominantly on the basis of neutral criteria, its use can buttress a plaintiff's argument that race was the primary reason for a deviation where race and politics would otherwise seem equally plausible.

Lastly, statements about the relative priority of districting goals may constitute evidence to support a finding of racial predominance. Taken alone, the parroting of federal requirements or the acknowledgment that certain compliance obligations are “mandatory” or “nonnegotiable” does not lend any weight in the predominance balance. If it did, the State would start the predominance balancing at an immediate disadvantage. However, if evidence is provided that demonstrates legislators held a false belief that certain artificial criteria—such as fixed BVAP floor—were necessary to comply with federal law, then statements by those particular legislators regarding compliance are relevant evidence in the predominance inquiry.

ii. Core Retention

Core retention—or “respecting existing district boundaries”—appears to be facially neutral and serves neutral political values, such as increased administrative ease, electoral accountability, and enhanced voter awareness and engagement. Unlike the other neutral criteria identified above, however, core retention holds a special place in the predominance balance. That is because “core preservation ... is not directly relevant to the origin of the new district inhabitants.” Alabama, 135 S.Ct. at 1271. Moreover, core retention may be used to insulate the original basis for the district boundaries.

Thus, where district lines track a path similar to their predecessor districts or where “core retention” seems to predominate, courts should also examine the underlying justification for the original lines or original district. Legislators' use of the core retention principle should certainly receive some degree of deference. But, the inquiry in a racial sorting claim examines the basis upon which voters were placed “within or without a particular district.” Miller, 515 U.S. at 916, 115 S.Ct. 2475. “That's the way we've always done it” may be a neutral response, but it is not a meaningful answer.

The Court applied the foregoing principles when weighing all of the evidence in the record and in ascertaining whether voters were sorted into a district predominantly on the basis of their race.

2. Strict Scrutiny Analysis

Having applied these precepts to the evidence, we found that the Plaintiffs met their burden to prove that race was predominant in the formation of HD 75, making it necessary to apply strict scrutiny as to that district. To survive strict scrutiny, the redistricting statute must be narrowly tailored to a compelling state interest. In the redistricting context, this familiar test takes on a somewhat different appearance, which the Court will now examine.

a. Compelling Interest

In prior cases, the Supreme Court has assumed, without deciding, that compliance with federal antidiscrimination laws can constitute a compelling state interest. See Shaw II, 517 U.S. at 915, 116 S.Ct. 1894 (“We assume, arguendo, for the purpose of resolving this suit, that compliance with § 2 [of the VRA] could be a compelling interest[.]”); Bush, 517 U.S. at 977, 116 S.Ct. 1941 (principal opinion) (“[W]e assume without deciding that compliance with the results test [of the VRA] ... can be a compelling state interest.”). Various members of the Court have also expressed their separate views on the matter. See Bush, 517 U.S. at 990, 116 S.Ct. 1941 (O'Connor, J., concurring) (“In my view ... the States have a compelling interest in complying with the results test [of the VRA] as this Court has interpreted it.”); LULAC, 548 U.S. at 517, 126 S.Ct. 2594 (Scalia, J., concurring in the judgment in part and dissenting in part, joined by Chief Justice Roberts, Justice Thomas, and Justice Alito) (“I would hold that compliance with § 5 of the Voting Rights Act can be [a compelling state] interest.”).

This already complex posture was rendered even less certain by the recent decision in Shelby County. There, the Supreme Court struck down the coverage formula under Section 4 of the VRA, but “issue[d] no holding on § 5 itself[.]” 133 S.Ct. at 2631. The Supreme Court did not help matters in Alabama when it stated, “[W]e do not here decide whether, given Shelby County v. Holder, continued compliance with § 5 remains a compelling interest [.]” 135 S.Ct. at 1274 (internal citation omitted).

Here, the Intervenors claim compelling interests founded on both Section 2 and Section 5 of the VRA. To resolve whether compliance with the VRA was a compelling interest at the time of enactment, the Court finds the rationale offered by Justice Scalia in his LULAC opinion convincing. As to Section 5, Justice Scalia wrote, in a passage joined by Chief Justice Roberts, Justice Thomas, and Justice Alito:

We long ago upheld the constitutionality of § 5 as a proper exercise of Congress's authority under § 2 of the Fifteenth Amendment to enforce that Amendment's prohibition on the denial or abridgment of the right to vote. If compliance with § 5 were not a compelling state interest, then a State could be placed in the impossible position of having to choose between compliance with § 5 and compliance with the Equal Protection Clause.

548 U.S. at 517, 126 S.Ct. 2594 (Scalia, J., concurring in the judgment in part and dissenting in part, joined by Chief Justice Roberts, Justice Thomas, and Justice Alito) (internal citations omitted). We find this reasoning persuasive, with the proviso that the State's interest must be in actual compliance with the standards articulated in federal antidiscrimination law as interpreted by the federal courts.

This reasoning is persuasive with respect to Section 2 as well. See Bush, 517 U.S. at 990–92, 116 S.Ct. 1941 (O'Connor, J., concurring) (noting that the Supreme Court has repeatedly enforced the obligations of § 2, lower courts have unanimously affirmed its constitutionality, and states would be “trapped between the competing hazards of liability” if § 2 were not a compelling interest). Because only a compelling interest in actual compliance with the non-retrogression standard of Section 5 is necessary to the resolution of this case, however, the Court need not address Section 2 at length.

This distinction is an important one. In Miller, the Supreme Court stipulated that “compliance with federal antidiscrimination laws cannot justify race-based districting where the challenged district was not reasonably necessary under a constitutional reading and application of those laws.” 515 U.S. at 921, 115 S.Ct. 2475. That fundamental limitation remains applicable. In drafting redistricting legislation, the State must pass a state law that complies with both federal law and the federal constitution. Thus, the goal of “actual compliance” is clearly compelling. If the State achieves actual compliance with the demands of a federal statute, and the federal statute is itself constitutional, then there can be little doubt that the state law is similarly constitutional.

The State also has an interest in avoiding preclearance denial under Section 5 (or liability under Section 2). This goal of “defensive compliance,” however, is not a compelling interest. See, e.g., id. at 921–27, 115 S.Ct. 2475. This is because defensive compliance could often entail a violation of constitutional law itself: subordinating traditional, neutral criteria and other districting criteria to racial considerations. See Harris v. Arizona Indep. Redistricting Comm'n, 993 F.Supp.2d 1042, 1054–55 (D.Ariz.2014) (noting that “[s]everal aspects of the preclearance process ... may work together to ... encourage a state that wants to obtain preclearance to overshoot the mark, particularly if it wants its first submission to be approved”).

But Section 5 does not require—and cannot be read to require—states to subordinate traditional, neutral districting principles to race in the redistricting process. The DOJ's own regulations state this explicitly. Pls.' Ex. 9 at 4 (76 Fed.Reg. Vol. 27 (Feb. 9, 2011) at 7472) (“[P]reventing retrogression under Section 5 does not require jurisdictions to violate Shaw v. Reno and related cases.”). Therefore, a state that finds itself engaging in predominant racial sorting to fulfill an interest in defensive compliance will begin to forfeit any credible interest in preventing retrogression and may be said to have adopted an interpretation of Section 5 that would itself render Section 5 unconstitutional as applied. In sum, we hold that Virginia's interest in actual compliance with the standards of federal antidiscrimination law—as the federal courts have interpreted them—was a compelling interest at the time the 2011 redistricting plan was designed and enacted.

Nor does Section 2 require states to engage in such behavior. That is because Section 2 requires a plaintiff to first prove that the minority group was “geographically compact” and could have constituted a numerical majority in a hypothetical single-member district. See Shaw II, 517 U.S. at 916, 116 S.Ct. 1894; LULAC, 548 U.S. at 433, 126 S.Ct. 2594; Bartlett, 556 U.S. at 26, 129 S.Ct. 1231.

The conceptual difficulty with the compelling interest arises when the State attempts actual compliance but does not achieve actual compliance. But this is not a dispute about whether the interest is compelling; it is a dispute about whether the State's attempt was “narrowly tailored.” If the State's goal was actual compliance with a proper reading of a constitutional federal standard, then the interest is compelling. Only the federal courts can ascertain whether the State “achieved” actual compliance with a constitutional reading of those statutes, so the State can only “attempt” actual compliance.

Apart from that question, the Court believes that an interest that is compelling at a redistricting plan's inception is capable of sustaining the plan until the next districting cycle. As the district court in Alabama stated, “We evaluate the plans in the light of the legal standard that governed the Legislature when it acted, not based on a later decision of the Supreme Court that exempted [the State] from future coverage under section 5 of the [VRA].” See Alabama Legislative Black Caucus v. Alabama, 989 F.Supp.2d 1227, 1307–08 (M.D.Ala.2013) (three-judge court), vacated and remanded, ––– U.S. ––––, 135 S.Ct. 1257, 191 L.Ed.2d 314 (2015). Because the legislature possessed a compelling interest in actual compliance with federal antidiscrimination laws as interpreted by the federal courts at the time the plan was enacted, and because redistricting plans are inherently subject to periodic revision on a reasonable, decennial basis, we conclude that the compelling interest underlying the statute at enactment remains a compelling interest during its effective duration.

b. Narrow Tailoring

The next question in the analytical calculus is whether the State's redistricting statute was “narrowly tailored” to this compelling interest. In particular, the question is whether a State's “attempt” at actual compliance could be viewed as “reasonably necessary under a constitutional reading and application of [federal antidiscrimination] laws.” Shaw II, 517 U.S. at 911, 116 S.Ct. 1894 (citing Miller, 515 U.S. at 921, 115 S.Ct. 2475). In Alabama, the Supreme Court explained that narrow tailoring is satisfied if there is a “strong basis in evidence” for the predominant use of race in drawing a challenged district. 135 S.Ct. at 1274.

The conceptual difficulty for the narrow-tailoring inquiry is this: if a finding of predominance means that race subordinated other considerations, and a constitutional reading of the antidiscrimination standards does not require race to subordinate other considerations, then how can an unconstitutional reading of a federal statute by the State be the interest that saves the State's unconstitutional racial gerrymander? The answer is this: if the disregard for non-racial criteria could have reasonably been viewed as not substantial, and the State shows a strong basis in evidence that its deviations appeared necessary to ensure actual compliance with the federal standard, then the district could still have been considered reasonably necessary under a constitutional reading of the statute.

Therefore, as the finder of fact, we employ a “preponderance” standard during the predominance inquiry, but apply a “sufficiency” standard during the narrow tailoring inquiry. Justice Breyer's dissent in Abrams v. Johnson makes this rationale clear:

This legal distinction—between whether a plan really violates § 2 or might well violate § 2—may seem technical. But it is not. A legal rule that permits legislatures to take account of race only when § 2 really requires them to do so is a rule that shifts the power to redistrict from legislatures to federal courts (for only the latter can say what § 2 really requires). A rule that rests upon a reasonable view of the evidence (i.e., that permits the legislature to use race if it has a “strong basis” for believing it necessary

to do so) is a rule that leaves at least a modicum of discretionary (race-related) redistricting authority in the hands of legislators.

521 U.S. at 114, 117 S.Ct. 1925 (Breyer, J., dissenting). In Abrams, a federal court was already required to undertake the districting endeavor, so Justice Breyer's dissent was unavailing. Because the lower court decided that it could not create a second majority-black district without subordinating neutral principles, it declined to do so. Id. at 84–85, 117 S.Ct. 1925 (majority opinion). This does not mean, however, that a court reviewing a State's plan cannot accept the State's alternate judgment, so long as the legislature had a strong basis for believing its plan was compliant.

Therefore, for predominance, the inquiry is whether, as a matter of fact, the State substantially disregarded non-racial criteria. For narrow tailoring, the inquiry is whether the State had good reason to believe that its actions were required for actual compliance with the non-dilution or non-retrogression standard. Because substantial disregard of non-racial criteria is not required under a constitutional reading of either standard, this inquiry necessarily entails also asking whether the State had good reason to believe that its own departure from non-racial criteria was not substantial.

Because the standards of the racial sorting claim and the standards of non-dilution and non-retrogression often stand in tension, the Court must recognize that the State is attempting to “toil with the[se] twin demands” and provide a fairway for the State's objectively reasonable efforts. Bush, 517 U.S. at 990, 116 S.Ct. 1941 (O'Connor, J., concurring). There may be a variety of plans that reasonably avoid dilution and retrogression and also reasonably respect traditional, neutral districting principles. If the legislature had a strong basis in evidence for its districting decision and reasonable individuals could have come to a different conclusion, then the court should accept that reasonable judgment during the narrow tailoring stage.

Thus, the question a court must ask at the narrow-tailoring stage is whether the legislature has shown that it had “good reasons” to believe—i.e., that it had a strong basis in evidence for believing—that its actions were reasonably necessary to achieve actual compliance with federal antidiscrimination standards based on a constitutional reading of those standards. Or, could a reasonable legislator have come to the conclusion that the challenged district violated neither federal law nor any constitutional limitations upon that federal law.

This formulation also explains why the Plaintiffs and Intervenors proposed seemingly different narrow tailoring inquiries. Plaintiffs argue that the State “must show that [it] had a ‘strong basis in evidence’ for believing that all of the Challenged Districts needed to meet or exceed a predetermined BVAP target to avoid retrogression.” Pls.' Post–Trial Brief at 28. Intervenors argued at trial that the narrow tailoring question is “how much that district violates the state's criteria.” Trial Tr. 855:20–21 (Intervenors) (emphasis added). Both of these inquiries are necessary, but neither is sufficient.

The narrow tailoring inquiry asks whether “the legislature ha[d] a ‘strong basis in evidence’ in support of the (race-based) choice that it has made.” Alabama, 135 S.Ct. at 1274.

This standard ... does not demand that a State's actions actually be necessary to achieve a compelling state interest in order to be constitutionally valid. And legislators may have a strong basis in evidence to use racial classifications in order to comply with a statute when

they have good reasons to believe such use is required, even if a court does not find that the actions were necessary for statutory compliance.

Id. (emphasis added). With respect to Section 5, for example, this inquiry into whether the “race-based choice” had a “strong basis in evidence” reaches both the standard of retrogression and—because a constitutional interpretation of retrogression does not require subordination—the standard of subordination.

With respect to subordination, the Supreme Court has noted that the extent of a State's disregard of neutral criteria “is not irrelevant to the narrow tailoring inquiry” when it “exhibit[s] a level of racial manipulation that exceeds what [the VRA] could justify.” Bush, 517 U.S. at 980–81, 116 S.Ct. 1941 (principal opinion) (emphasis added). Accord Miller, 515 U.S. at 921, 115 S.Ct. 2475 (“[C]ompliance with federal antidiscrimination laws cannot justify race-based districting where the challenged district was not reasonably necessary under a constitutional reading and application of those laws.”). In other words, part of showing that a district is narrowly tailored to an interest in actual compliance with a constitutional reading of the retrogression standard entails showing that the district is one that a reasonable legislator could believe entailed only reasonable and minor deviations from neutral districting conventions.

Nor is an inquiry into whether the State possessed a “strong basis in evidence” that its actions were necessary to “prevent retrogression” limited to the BVAP percentages in the Benchmark Plan's existing majority-minority districts. When Congress amended Section 5, it rejected the Supreme Court's decision in Georgia v. Ashcroft, 539 U.S. 461, 123 S.Ct. 2498, 156 L.Ed.2d 428 (2003), and “adopted the views of the dissent.” Alabama, 135 S.Ct. at 1273 (citing H.R.Rep. No. 109–478, pp. 68–69, and n. 183 (2006)). The dissent “made clear that courts should not mechanically rely upon numerical percentages but should take account of all significant circumstances.” Id. at 1273 (citing Ashcroft, 539 U.S. at 493, 498, 123 S.Ct. 2498 (Souter, J., dissenting)). Thus, there can be no argument that retrogression “locks in” the BVAP of each particular district. Ashcroft, 539 U.S. at 498, 123 S.Ct. 2498 (Souter, J., dissenting) (noting that the entire Court agrees that “the simple fact of a decrease in [BVAP] in some districts is not alone dispositive about whether a proposed plan is retrogressive”).

The retrogression standard also does not “lock in” a specific number of majority-minority districts. See id. at 492, 123 S.Ct. 2498 (“I agree with the Court that reducing the number of majority-minority districts within a State would not necessarily amount to retrogression barring preclearance under § 5 of the Voting Rights Act of 1965.”); Texas v. United States, 831 F.Supp.2d 244, 260 (D.D.C.2011) (“[T]he Supreme Court ... has never suggested that the inquiry required by Section 5 can be satisfied by examining only the number of majority-minority districts. In fact, the Court has acknowledged that the inquiry is a complex undertaking.”). This holds true not only as a legal principle, but as a matter of logic. Based on demographic changes within the State, it simply may not be feasible to create the same number of majority-minority districts because performing Section 5 districts must also avoid unreasonable deviations from neutral districting criteria. See Miller, 515 U.S. at 910, 115 S.Ct. 2475.

A retrogression analysis must “take account of all significant circumstances,” Alabama, 135 S.Ct. at 1273, while retaining Section 5's “anchoring reference to electing a candidate of choice,” Ashcroft, 539 U.S. at 493, 123 S.Ct. 2498 (Souter, J., dissenting). This mandate is now part of the statute itself. See 52 U.S.C. § 10304(b) (prohibiting covered jurisdictions from adopting changes that “ha[ve] the purpose of or will have the effect of diminishing the ability of any [minority] citizens ... to elect their preferred candidates of choice[.]”). “Clearly, ‘ability to elect’ is the statutory watchword.” Texas, 831 F.Supp.2d at 260.

Therefore, once a court finds that race predominated, the strong basis in evidence standard asks not only whether the legislature had good reasons for believing the BVAP percentage employed in the district—as well as the district itself—was necessary to avoid retrogression, but also whether the district is one that a reasonable legislator could believe generally respected neutral districting principles. As the Alabama Court reminded: “The standards of § 5 are complex; they often require evaluation of controverted claims about voting behavior; the evidence may be unclear; and, with respect to any particular district, judges may disagree about the proper outcome.” 135 S.Ct. at 1273. This applies to reasonable state judgments about subordination as well. In the context of redistricting, the “narrow tailoring” inquiry permits the State to overshoot the bull's-eye, so long as it hits the target.

The foregoing legal framework for analyzing a racial sorting claim provides the guidepost for the statewide and district-by-district findings that follow.

B. Evidence Of General Application To All Districts

“A racial gerrymandering claim ... applies to the boundaries of individual districts” and must be proven on a “district-by-district” basis. Alabama, 135 S.Ct. at 1265. However, the Plaintiffs provided some evidence that applied across all districts. Therefore, the Court will assess that evidence before proceeding to its district-by-district analysis. Id. (“Voters, of course, can present statewide evidence in order to prove racial gerrymandering in a particular district.”). In like fashion, the Commonwealth's evidence may apply across districts. Our findings on the evidence are based on our credibility determinations and how particular evidence squares with the record as a whole.

First, the Intervenors frequently discussed the substantial population changes experienced on both a statewide level and in the Challenged Districts. See, e.g., Ints.' Post–Trial Brief at 19–20 (Docket No. 104). That evidence has a role to play in the predominance analysis, but it is a limited one.

As the Supreme Court held in Alabama, “an equal population goal is not one factor among others to be weighed against the use of race to determine whether race ‘predominates.’ ” 135 S.Ct. at 1270. Instead, “it is part of the redistricting background, taken as a given, when determining whether race, or other factors, predominate in a legislator's determination as to how equal population objectives will be met.” Id. Although the equal population goal is not a traditional factor to be considered in the balance in deciding predominance, its “background” role is nonetheless important in assessing why certain redistricting actions were taken. For example, gains or losses in population affect where in a State new districts must be created or where old districts cannot stand. That, in turn, is pertinent to which neutral redistricting criteria can—or cannot—be fully satisfied.

The predominance question “concerns which voters the legislature decides to choose[.]” Alabama, 135 S.Ct. at 1271.

That is because, like compliance with the VRA, it is a “demand” that the State does not have the option of ignoring. See Page, 2015 WL 3604029 at *26 (Payne, J., dissenting). “Indeed, in light of the Constitution's demands, that role may often prove ‘predominant’ in the ordinary sense of that word. But, ... ‘predominance’ in the context of a racial gerrymandering claim is special. It is not about whether a legislature believes that [a goal] takes ultimate priority.” Alabama, 135 S.Ct. at 1270–71; accord Page, 2015 WL 3604029 at *26 (Payne, J., dissenting) (“[T]here is a difference between a State's ‘paramount concern’ with complying with federal law and a State's use of [a factor] as a ‘predominant criterion’ for allocating voters between districts.”).

Second, for the reasons provided in the factual discussion in Section III above, the Court finds that a 55% BVAP floor was employed by Delegate Jones and the other legislators who had a hand in crafting the Challenged Districts. Those delegates believed this necessary to avoid retrogression under federal law, and we do not doubt the sincerity of their belief.

The dissent believes that Virginia's “one-size-fits-all quota ... raises even more serious concerns” than the mechanical racial targets in Alabama because the Alabama legislature “sought to maintain preexisting racial percentages specific to each district with the aim of avoiding retrogression[.]” Post at 574. But, the legislators in Alabama mistakenly believed that any decrease in existing BVAP percentages would constitute retrogression. Any patina of district-specific treatment was no more than the residue of this misconception.

Third, the Plaintiffs' expert, Dr. Stephen Ansolabehere, testified about his analysis of VTDs in the Commonwealth. In particular, Dr. Ansolabehere used statistical models to examine the movement of VTDs into and out of the Challenged Districts and opined whether, in his view, those movements were predominantly “racial” or “political.” See id. at 149:19–152:6 (Ansolabehere).

With respect to Dr. Ansolabehere's analysis regarding race and politics as “predictors” of the likelihood of inclusion of VTDs in one of the Challenged Districts, the Court has both initial technical concerns and more fundamental substantive concerns about the method employed that cause us not to credit his views as to the reasons for VTD placement. First, even though Dr. Ansolabehere's analysis provides a “regional” control to avoid examining VTDs that could not have feasibly found their way into the Challenged Districts, id. at 163:19–25 (Ansolabehere), that does not account for whether a VTD in that region could be considered to “hop” over another VTD in the region en route to the target district in violation of contiguity conventions, see id. at 503:9–504:3 (Katz) and 514:23–515:13 (Katz) (noting that the analysis incorrectly assumes that a VTD “can be independently assigned to a given district” and that “doing [the same analysis] by subregions doesn't solve that problem”).

Admittedly, Dr. Katz's approach—which includes a variable for distance from the center of the target district—is, by his own description, “not a perfect fix” and a sort of “crude or poor approximation.” Trial Tr. 504:18–24 (Katz). Nonetheless, it offers a more reliable approach to the issue than Dr. Ansolabehere's analysis.

More fundamentally, however, Dr. Ansolabehere's “race versus politics” opinions miss the mark because they do not consider the extent to which the boundaries themselves are justifiable by neutral criteria or any other motivation besides race or political disposition. The models that he employed do not, for example, consider “economic factors, social factors, cultural factors, geographic factors, governmental jurisdictions and service delivery areas.” Id. at 230:14–21 (Ansolabehere). If a district is intentionally designed as a performing district for Section 5 purposes, there should be little surprise that the movement of VTDs into or out of the district is correlated—even to a statistically significant degree—with the racial composition of the population. This does not mean, however, that race “predominated” for the purposes of a racial sorting claim.

The predominance question requires an inquiry into whether the movement of VTDs into and out of a district subordinated other criteria in the process. See Backus v. South Carolina, 857 F.Supp.2d 553, 565 (D.S.C.2012), sum. aff'd, ––– U.S. ––––, 133 S.Ct. 156, 184 L.Ed.2d 1 (2012). Dr. Ansolabehere's analysis, for the most part, just does not provide any specific insights into this inquiry. Dr. Ansolabehere's partial correlation analysis, which holds other factors—including party—steady can be considered in determining whether a district's deviations from neutral criteria may be more attributable to race or politics, id. at 157:24–158:5 (Ansolabehere), but it can only be considered in assessing—not refuting—testimony that provides non-racial reasons for particular deviations from neutral principles. Moreover, using Dr. Katz's admittedly crude, but nonetheless reliable, approximation for the limitation that VTDs are not equally susceptible to being included in every district, the statistical significance of the racial justification disappears, at least with respect to the question of whether race or politics is a more significant predicator of VTD placement. See Ints.' Ex. 16 at 21, Table 1; Trial Tr. 505:22–510:25 (Katz) (“Statistically these are a tie.”). On balance, Dr. Ansolabehere's analysis on the VTD issue is not reliable proof on the predominance issue.

Lastly, the Court finds that some “statewide” compactness information is useful as a point of comparison for the district-by-district analysis set out in Section IV.C. below. In the Challenged Districts, the average Reock score was .320, the average Polsby–Popper Score was .192, and the average Schwartzberg score was 2.365. Pls.' Ex. 51 at 12, Table 2. In the Non–Challenged Districts, the average Reock score was .360, the average Polsby–Popper Score was .243, and the average Schwartzberg score was 2.128. Id. Under the Reock and Polsby–Popper measures, higher scores represent more compact districts. Id. Under the Schwartzberg measure, lower scores represent more compact districts. Id. Of the 100 House districts, seven of the Challenged Districts are in the “bottom 50”—with the lowest Reock scores—and five of the Challenged Districts are in the “top 50”—with the highest Reock scores. Trial Tr. 721:8–12 (Hofeller).

Dr. Katz utilized a modified Boyce–Clark measure in his analysis. Trial Tr. 537:2–4 (Katz). The Court declines to analyze the districts separately using this measure. Dr. Katz appeared to employ the Boyce–Clark measure simply to prove the more academic point that there is no agreed-upon standard and that different measures can lead to different outcomes. Id. at 540:19–542:9 (Katz). This point is not disputed.

None of the experts disputed the compactness calculations provided by the Plaintiffs. However, the Court reiterates that compactness is “more of a flag than a conclusion” and rejects the suggestion by Dr. Ansolabehere that districts under .20 on the Reock scale are presumptively “non-compact.” See ante at 533 n. 15.

With these generally applicable findings in mind, the Court now advances to the requisite district-by-district analysis. In so doing, the analysis is guided by the legal principles and the framework outlined in Section IV.A. above.

C. District–by–District Analysis

As with the generally applicable factual findings above, our district-by-district analysis itself is a factual one that we have based on our examination of the record as a whole and on our assessment of the credibility of the witnesses.

1. District 63

HD 63 is found in the Dinwiddie–Greensville area and was represented by then-Delegate Rosalyn Dance during the 2011 redistricting process. Under the Benchmark Plan, the district contained all of Dinwiddie and Petersburg City, and part of Chesterfield. Pls.' Ex, 50 at 69, Table 1. Under the Enacted Plan, the district now contains all of Petersburg City and parts of Chesterfield, Dinwiddie, Hopewell, and Prince George. Id. This increased the number of county and city splits from 1 to 4 and increased the number of split VTDs from 0 to 8. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 63 has a core retention percentage of 80.2, Ints.' Ex. 14 at 83, and is contiguous by land.

Dr. Ansolabehere and Dr. Hood come to different statewide conclusions regarding the number of VTD splits. See Ints.' Ex. 15 at 6 n. 5. This is because Dr. Hood counts the number of VTDs that are split, whereas Dr. Ansolabehere counts the number of splits in VTDs. The latter method accounts for VTDs that are split multiple times. We are not convinced that Dr. Ansolabehere's approach is entirely sound. See Pls.' Ex. 51 at 15 n. 3. But, because Dr. Hood only provides statewide splits data, the Court will rely upon Dr. Ansolabehere's district-by-district splits data, thereby giving Plaintiffs the benefit of the doubt. The Court expresses no opinion regarding the appropriate counting measure.

On its face, the district is unusually shaped. After chopping Dinwiddie County in half, the southern border of the district tends to follow precinct boundaries from west to east until it cuts through Dinwiddie precinct along Interstate 85. After that, the district line constricts, carving out a hook around New Hope. After a brief return to a rather normal configuration around Petersburg City, the district narrows to avoid the Jefferson Park area and the homes of Delegates Cox and Ingram. It then continues in a narrow form through Prince George, into various parts of Hopewell, and terminates at the James River. See Pls.' Ex. 66 at 1; Ints.' Ex. 94 at 1.

The district had Reock and Polsby–Popper scores of .61 and .48 under the Benchmark Plan and experienced a steep drop to scores of .25 and .16 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. This marks the largest Reock compactness reduction of any district in the Enacted Plan. Trial Tr. 140:7–9 (Ansolabehere). The district's Schwartzberg score is 2.506. Pls.' Ex. 51 at 11, Table 1.

The district's deviations from neutral redistricting criteria begin with the splitting of Dinwiddie County. This split appears to be avowedly racial. Delegate Dance testified that the southern half of Dinwiddie “went to Delegate Tyler to try to get her number ... [o]f African–American voters up to 55 percent.” Trial Tr. 80:11–17 (Dance). Within this deviation are two sub-deviations: (1) the splitting of Dinwiddie precinct; and (2) the hook that wraps around New Hope precinct.

The Dinwiddie precinct is split along I–85, but this is not listed among the redistricting criteria, which undermines its explanatory value as a districting criterion. See Alabama, 135 S.Ct. at 1271–72. Although established transit corridors may split areas into “communities of interest” over time, there was no evidence that this precinct is comprised of distinct communities on either side of the highway. On the other hand, the artificial border provided by I–85 may provide a clear boundary to voters and candidates alike that reside in Dinwiddie precinct and wish to know their House district. In the absence of any further explanation by the Intervenors or the Plaintiffs, however, the Court declines to identify any particular rationale for this “sub-deviation,” meaning that the Plaintiffs have not carried their burden of attributing it to race.

The other “sub-deviation”—the hook around New Hope—is decidedly not racial. After reviewing the evidence, the Court finds that the purpose for this deviation was “challenger prevention” and “incumbency protection.” This deviation was negotiated between Delegates Dance, Tyler, and Jones. Trial Tr. 325:24–25 (Jones). Delegate Jones testified that the cutout accounted for “the bulk of the splits in [the 75th] district,” id. at 326:18–19, that New Hope was retained in HD 63 because “a tremendous amount of [Delegate Dance's] employees or constituents had family” there, id. at 326:5–10, and that Delegate Dance had “a potential primary opponent she wanted to draw out of her district,” id. at 326:11–12; accord id. at 858:4–7 (Intervenors); Ints.' Pre–Trial Brief at 20. So, if it looks like the hook is reaching for something, that's because it is: a potential threat to the incumbent.

Thus, at this point the record is that one reason for the configuration of HD 63 was racial and one reason was purely political.

The other component of HD 63's unusual shape is its reach north and east from U.S. 460 to the James River in a way that runs through both Prince George County and the City of Hopewell. In so doing, this component of HD 63 increases the number of localities in the district from three to five, and it also splits a number of VTDs. Trial Tr. 140:16 (Ansolabehere); id. at 79:23–80:3 (Dance). According to Delegate Dance's testimony, “that's what it took to get [Delegate Tyler] to the 55 percent strength of African–American voters.” Id. at 81:15–18 (Dance). Not only did this help satisfy the 55% threshold in District 75, it also helped maintain a substantial African–American population in District 63. Delegate Dance “picked up parts of Prince George ... to get more African–Americans ... [a]nd then ... picked up the concentration of African–Americans in Hopewell[.]” Id. at 81:21–83:6 (Dance).

However, the record shows that the eastern border advanced other criteria, both neutral and political. In order to unwind the water crossing in the Benchmark HD 74, Delegate Jones decided to move precincts in Hopewell City out of HD 74 and into HD 63. Thus, HD 63's eastern configuration improved HD 74's adherence to contiguity conventions. See Wilkins, 264 Va. at 465, 571 S.E.2d 100 (examining whether HD 74's water continuity was permissible under the Constitution of Virginia). Moreover, by placing these precincts in HD 63 rather than HD 62 or HD 64, the District's eastern boundary avoids solving the water crossing problem to the detriment of Republican districts on either side. See Ints.' Ex. 92 at 2. Thus, it appears that this aspect of HD 63's unusual shape can be explained on a neutral, racial, and political basis.

It is the Plaintiffs' burden to show that the racial considerations subordinated all other criteria, including neutral criteria and other non-racial criteria. The evidence provided thus far is in equipoise, and the Plaintiffs have not yet satisfied their burden on the predominance issue.

Plaintiffs rely on the testimony of Dr. Ansolabehere to complete their task. To begin, Dr. Ansolabehere notes the drop in compactness scores but, as discussed above, that is more of a flag than a conclusion. If compactness has been sacrificed to enhance contiguity or serve political ends, then race alone has not subordinated this criterion. Dr. Ansolabehere also analyzed VTD movements but, as discussed above, that analysis fails to account for other criteria that may be shaping the district, such as incumbency considerations or solving contiguity issues in nearby districts. Finally, Dr. Ansolabehere notes the number of VTD splits. But the majority of splits are attributable to incumbency considerations rather than race. Moreover, some splits appear to be attributable to Delegate Jones' twin aims of solving the water crossing and limiting population deviations to ±1%. In sum, we find Dr. Ansolabehere's testimony on each point to be unconvincing. Thus, his evidence did not help the Plaintiffs in their obligation to prove predominance and to dislodge the presumption of lawful action to which the General Assembly's redistricting plan is entitled.

Based on the record, the Court finds that the Plaintiffs have not satisfied their burden to prove that racial considerations subordinated all other neutral and race-neutral districting criteria in the formation of HD 63. And, on the basis of the record, the Court holds, as a matter of fact, that race did not predominate in the drawing of HD 63.

2. District 75

HD 75 is found in the Dinwiddie–Greensville area and was represented by Delegate Roslyn Tyler during the 2011 redistricting process. Under the Benchmark Plan, the district contained all of Sussex County, Greensville, and Emporia City and parts of Brunswick, Franklin City, Isle of Wight, Lunenberg, and Southampton. Pls.' Ex. 50 at 69, Table 1. Under the Enacted Plan, the district now contains all of Emporia City and Greensville and parts of Brunswick, Dinwiddie, Franklin City, Isle of Wight, Lunenberg, Southampton, Surry, and Sussex. Id. This increased the number of county and city splits from 5 to 8 and increased the number of split VTDs from 4 to 13. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 75 has a core retention percentage of 78.64, Ints.' Ex. 14 at 83, and is contiguous by land.

On its face, the district appears relatively compact, despite its odd tendency to leak across county and city lines. Pls.' Ex. 66 at 6. The district had Reock and Polsby–Popper scores of .42 and .22 under the Benchmark Plan, which shifted to scores of .41 and .19 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2.282. Pls.' Ex. 51 at 11, Table 1. Although the district's technical compactness remained “about the same between the two plans,” Trial Tr. 141:4–5 (Ansolabehere), Delegate Tyler testified that her district has “[v]ery irregular borders” and is “not an easy district to follow,” (Docket No. 90–2, Ex. B, 23:2–7).

A review of HD 75's boundaries suggests that she is right. Although the district has a clear southern border, that provides no solace because her district borders North Carolina. Unlike population equality and VRA compliance, state borders are not just mandatory; they admit no variation. As such, state borders are a ity in the predominance balance. The only other county boundaries seemingly respected are those segments bordering Mecklenburg, Nottoway, Prince George, and Suffolk counties. Pls.' Ex. 66 at 6. Notable in this regard, is the addition of the district's lower left corner, which makes Brunswick County whole. Trial Tr. 323:8–10 (Jones); Ints.' Ex. 94 at 7.

Delegate Dance testified that the creation of HD 75 “gave us a little trouble to try to get to the 55 percent.” Trial Tr. 741:1–15 (Dance). To get to the 55% BVAP, the district “required some drastic maneuvering[.]” Id. Delegate Tyler herself testified that she “was concerned about the decrease in number of black people in my district.” (Docket No. 90–2, Ex. B, 88:15–16.)

Although the irregularity of the district boundaries can be seen to buttress Delegate Dance's testimony that HD 75 required “drastic maneuvering” in order to comply with the 55% BVAP floor, the Intervenors have offered their own explanations for the district's “very irregular borders.” Delegate Jones testified that Dinwiddie County was split because the district was in need of population. Trial Tr. 323:2–4 (Jones). That appears to be the case because HD 75 was underpopulated. The choice to go north, however, was “to try to get [Delegate Tyler's] number ... [o]f African–Americans voters up to 55 percent.” Id. at 80:11–17 (Dance). Therefore, while underpopulation may help explain the changes to the district, it cannot be weighed against race in the predominance analysis.

The district's irregular eastern and western borders can be also attributed to race because, according to Delegate Dance, moving coherently to the “east [or] west would have been Euro–Americans, and she needed some African Americans to get to that 55 percent.” Id. at 80:21–24 (Dance). Delegate Jones' testimony did not contradict that assessment.

Delegate Jones testified that many of the changes, such as swapping out the Wakefield and Dendron precincts, splitting Franklin City, and excluding the Berlin and Ivor precincts were done on the basis of a “member request” or because Delegate Tyler did not receive many votes in those removed precincts. See id. at 323:11–16; 324:12–16; 325:1–5 (Jones). Delegate Jones accepted these changes even though adherence to political subdivisions and compactness would be subordinated in the process. See id. at 323:11–16 (“[W]e had two other counties whole until ... she requested that we swap [Wakefield and Dendron] out.”); 325:14–16 (“I would have never done that had it not been requested because I wanted to split as few jurisdictional boundaries as I could [.]”). But attributing the changes to “member requests” or performance concerns begs, rather than answers, the relevant question: was the request racial or political?

Like in HD 63, the evidence admits of both a racial purpose and a political purpose. For instance, Delegate Jones himself testified that Delegate Tyler's request to swap Wakefield and Dendron was based on “real concerns” stemming from the fact that she “didn't break 51 percent” in a general election race “with a Caucasian” and that she “won by less than 300 votes” in a “five-way race in a primary with two Caucasians.” Id. at 323:19–324:3 (Jones). That bespeaks an effort to both protect the incumbent and prevent retrogression. Similarly, Delegate Jones testified: “[S]he was worried about too low of a black voting-age population for her to be able to be successful in an election.” Id. at 322:10–12. This too reflects an effort to protect the incumbent while also preserving minority voters' ability to elect their candidate of choice.

Unlike in HD 63, however, here there is no ambiguity about the basis upon which voters were sorted. Intervenors' Post–Trial Brief relies upon the overlapping racial and political purposes to argue that race did not “predominate.” According to the Intervenors, Delegate Tyler's deposition testimony “made crystal clear her view that ‘[w]hat I'm saying is most of the time blacks vote Democratic,’ and that ‘in [her] mind, the purpose of ensuring 55 percent BVAP was to help Democrats be elected.’ ” Ints.' Post–Trial Brief at 30–31 (citing Docket No. 90–2, Ex. B, 62:17–25 & 63:19–23). But, attributing a political purpose to—or justification for—the 55% BVAP floor does not somehow render it a non-racial classification. Whether the changes were made to comply with Section 5, enhance Democratic performance, or protect the incumbent, the changes were still made based on voters' skin color.

Weighing all the evidence and testimony provided on the record, the Court finds that racial considerations subordinated traditional districting principles and other non-racial districting criteria in the creation of HD 75. The testimony from the three delegates primarily responsible for shaping the district, Delegates Jones, Tyler, and Dance, shows that the overriding objective was to achieve a 55% BVAP in HD 75. Achieving a 55% BVAP floor required “drastic maneuvering” that is reflected on the face of the district and, according to Delegate Jones, would not otherwise have been undertaken due to the impact on traditional county boundaries. Delegate Tyler herself found the boundaries “very irregular,” worried about her ability to cover her district with ease, and was “concern[ed] about the decrease in number of black people in [her] district.”

Intervenors attempt to explain the boundary deviations by ascribing a political purpose to them. But that attempt is not successful. As in Bush, the record shows that, in building HD 75, race was used by Delegate Tyler herself as a proxy for Democratic voters in an effort to protect her own position as an incumbent at the expense of traditional districting principles. 517 U.S. at 972–73, 116 S.Ct. 1941 (principal opinion). When a legislator sorts voters by political affiliation or performance, then the deviation from neutral principles is a political one. But, when a legislator sorts voters by race, for whatever purpose, then the deviation is a racial one. As explained above, the lesson of Cromartie was that a political deviation would not be considered racial simply because the Democratic voters happened to be black. Cromartie I, 526 U.S. at 542, 119 S.Ct. 1545. The lesson was not that a racial deviation would be considered political simply because the black voters happened to be Democrats. That is using race as a proxy for political affiliation, an approach that is prohibited.

See Bush, 517 U.S. at 968–73, 116 S.Ct. 1941 (principal opinion) (“If district lines merely correlate with race because they are drawn on the basis of political affiliation, which correlates with race, there is no racial classification ... But, to the extent that race is used as a proxy for political characteristics, a racial stereotype requiring strict scrutiny is in operation .... the fact that racial data were used in complex ways, and for multiple objectives, does not mean that race did not predominate over other considerations. The record discloses intensive and pervasive use of race both as a proxy to protect the political fortunes of adjacent incumbents, and for its own sake in maximizing the minority population of [the District].”).

As to HD 75, the Plaintiffs have proved (without reference to Dr. Ansolabehere's testimony) that race was the predominate criterion leading to the disregard of neutral conventions in forming HD 75. Moreover, to the extent that political interests were considered and achieved, it appears that those criteria were secondary to, and only satisfied by, adherence to the 55% BVAP floor. Shaw II, 517 U.S. at 907, 116 S.Ct. 1894 (“That the legislature addressed these interests does not in any way refute the fact that race was the legislature's predominant consideration.”).

The dissent argues that our interpretation of predominance will allow legislators to “mask” racial sorting and only permit plaintiffs to challenge districts that “manifest extreme line-drawing unexplainable on race-neutral grounds, like the district at issue in Shaw I.” Post at 572, 575. Our holding with respect to HD 75 should put these fears to rest. The boundaries of HD 75 not only simultaneously advance racial and non-racial goals, but they are hardly egregious or “extreme.” That has not prevented us from carefully examining the actual basis upon which voters were sorted and finding predominance satisfied where non-racial criteria were subordinated in fact.

Based on the foregoing analysis, the Court finds that race was the predominate criterion driving the formation and configuration of HD 75; and, therefore, the legislature's decision is subject to strict scrutiny. To survive strict scrutiny, the Intervenors must show that the legislature had a “strong basis in evidence” for its racial districting decisions.

The Court finds that this burden has been satisfied and that, accordingly, HD 75 survives the Plaintiffs' challenge. First, Delegate Jones' determination that HD 75 (or its environs) reflected an “ability-to-elect” district requiring protection against retrogression was a reasonable determination. As Plaintiffs themselves point out, HD 75 appeared to be a performing ability-to-elect district before the State's redistricting efforts. Pls.' Post–Trial Brief at 33–34 (citing Pls.' Ex. 50 at 85, Table 14). Therefore, retaining this ability to elect reasonably can be viewed as necessary to ensure actual compliance with the federal non-retrogression standard.

Next, as to HD 75, the 55% BVAP floor is grounded in a “strong basis in evidence” because the primary source of the 55% BVAP threshold appears to have been an analysis of HD 75 itself. For example, Delegate Jones testified that he did not feel a 52% BVAP threshold across all districts would be acceptable “based on ... the functional analysis that I had done using the Tyler primary, for example, and the Tyler general election in 2005.” Trial Tr. 430:2–9 (Jones). These were close races, prompting “real concerns.” Id. at 323:19–324:3 (Jones). Delegate Jones met with Delegate Tyler “probably half a dozen times to configure her district as she felt it needed to be configured for ... [minority voters] to elect a candidate of their choice for her district.” Id. at 322:6–12 (Jones).

The Court does not suggest that those designing redistricting plans can always just add more BVAP every time a meaningful challenger appears. Like Section 2, Section 5 does not “guarantee minority voters an electoral advantage,” Bartlett, 556 U.S. at 20, 129 S.Ct. 1231, it only requires that the system not effect a retrogression in minority voters' effective electoral franchise. Interpreting the VRA to allow more than this would render it an instrument in service of the same discriminatory practices it was designed to eliminate. This would be contrary to the plain language of the Fifteenth Amendment itself, let alone the precepts of equal protection. Where an application of the VRA cannot reasonably be said to have gone beyond the “remedial,” however, it is this Court's duty to uphold it.

Delegate Jones examined turnout rates in HD 75, id. at 467:7–11 (Jones), an issue about which Delegate Tyler was particularly concerned, id. at 463:12–16 (Jones). In addition, Delegate Jones considered the district's prison population and relied upon his knowledge of the district's electoral history. Id. at 464:7–465:5; 458:18–459:18 (Jones). These are precisely the kinds of evidence that legislators are encouraged to use “[i]n determining whether the ability to elect exists in the benchmark plan and whether it continues in the proposed plan[.]” Pls.' Ex. 9 at 3 (76 Fed.Reg. Vol. 27 (Feb. 9, 2011) at 7471) (“[E]lection history and voting patterns within the jurisdiction, voter registration and turnout information, and other similar information are very important to an assessment of the actual effect of a redistricting plan.”).

Delegate Jones primarily testified about the 2005 election. See, e.g., Trial Tr. 458:15–459:18 (Jones). There were more recent elections in 2007 and 2009, but Delegate Tyler ran unopposed in those elections. See Pls.' Ex. 50 at 85, Table 14. The dissent suggests that these unopposed races “cas[t] significant doubt” on the contention that a 55% BVAP level remained necessary to prevent retrogression. Post at 578. But short of hiring a statistical analyst, it's hard to see how much useful information can be gleaned from the uncontested races. Should legislators have lowered the target by 1%, 2%, or 3%? Any preference for a 53% target instead of a 55% target would seem to rest upon speculation, not a stronger basis in evidence.

Plaintiffs dispute the need for raising the BVAP percentage in HD 75, arguing that the district was already a performing Section 5 district for minority-preferred candidates going into the 2011 redistricting. Pls.' Post–Trial Brief at 33–34 (citing Pls.' Ex. 50 at 85, Table 14). Here, that argument only strengthens the Intervenors' hand. Under the Benchmark Plan, BVAP in HD 75 was 55.3%. Under the Enacted Plan, BVAP in HD 75 was 55.4%. Id. at 34. Considering the intricacies of redistricting, the new HD 75 could effectively be considered to have the “same” BVAP level as the old HD 75. And, considering the evidence relied upon by Delegate Jones, it appears abundantly clear that he had “good reasons” for holding the BVAP in HD 75 just above 55% to ensure that the district remained a performing Section 5 district for minority-preferred candidates, as Plaintiffs' themselves suggest. Alabama, 135 S.Ct. at 1274.

Nor does the 55% floor appear unreasonable when subjected to expert review. Plaintiffs' own expert noted that HD 63 and 75 “exhibit high rates of [racial] polarization because large majorities of Whites vote in the opposite way as large majorities of African Americans.” Pls.' Ex. 50 at 51, 84, Table 14. Intervenors' expert agreed, observing that the 2011 and 2013 elections held in HD 75 were racially polarized. Ints.' Ex. 16 at 24, Table 4. Dr. Ansolabehere ultimately opined that a 55% BVAP threshold was not necessary in HD 75, Pls.' Ex. 50 at 55, but ex post statistical analyses cannot upset the State's ex ante judgment so long as that decision was “reasonably necessary” based on strong evidence. In this case, it was so based. Alabama, 135 S.Ct. at 1273. Simply put, there were “good reasons” to believe that a 55% BVAP threshold was necessary to ensure that minority voting influence did not retrogress in HD 75, and the Court will not upset that reasonable judgment. Id. at 1274.

The Court does not credit the racial polarization analysis conducted by Dr. Ansolabehere. His analysis drew from on-year statewide elections data (rather than off-year House of Delegates elections data). Trial Tr. 516:7–25 (Katz). We find that the use of the wrong elections led to unreliable results. Dr. Ansolabehere also relied on an ecological regression analysis (rather than an ecological inference analysis), which “doesn't make use of all ... available information” and results in “blatantly incorrect answers.” Id. at 521:10–14. As Dr. Katz testified, ecological regression “was great technology in 1950” when it was developed, but “[t]he world has come a long way in those intervening six decades.” Id. at 519:11–22. This too makes Dr. Ansolabehere's testimony unreliable.


The Plaintiffs offered Dr. Ansolabehere's testimony on racial polarization as pertinent to the predominance analysis even though it would (were the Court to accept it as reliable—which it does not) be more probative of the narrow tailoring analysis. But, either way, his testimony on racial polarization is flawed and cannot be credited.



The Court finds that legislators had good reason to believe that maintaining a 55% BVAP level in HD 75 was necessary to prevent actual retrogression (and not just to attain preclearance), and that this was achieved by reasonable deviations from traditional redistricting criteria (judged by a sufficiency standard). Because the State has provided a “strong basis in evidence” for its use of race-based districting in its configuration of HD 75, the Court holds that HD 75 passes constitutional muster under the Equal Protection Clause of the Fourteenth Amendment.

3. District 69

HD 69 is found in the Richmond area and was represented by Delegate Betsy Carr during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district contains parts of Chesterfield and Richmond City. Pls.' Ex. 50 at 69, Table 1. Although the number of county and city splits remained the same, redistricting increased the number of split VTDs from 2 to 4. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 69 has a core retention percentage of 74.7. Ints.' Ex. 14 at 83.

In Wilkins, the Supreme Court of Virginia found that race did not predominate over other districting criteria under Virginia's state constitution in Districts 69, 70, 71, 77, 80, 89, and 90. 264 Va. at 477–79, 571 S.E.2d 100. This Court finds the rationale and outcome stated in Wilkins, with respect to these districts, informative but not determinative. First, perhaps the simplest explanation is that the 2011 map is not the 2001 map, several similarities notwithstanding. Second, the Wilkins court observed that the “trial court did not reference any specific evidence or make any specific findings for any of these districts to support a conclusion that race was the predominant factor in creating each district.” Id. at 477, 571 S.E.2d 100. That is precisely the analysis this Court undertakes today. Third, the Wilkins court included population and core retention among the balancing criteria, which are either verboten or called into question by the Alabama decision. Compare id. at 478, 571 S.E.2d 100 with Alabama, 135 S.Ct. at 1270, 1271. Finally, there was no evidence before the Wilkins court suggesting the use of a racial floor in the subject districts.

On its face, the district appears to reflect a large, compact swath of Richmond below the Fan District and to the south of the James River. The district had Reock and Polsby–Popper scores of .37 and .20 under the Benchmark Plan, which increased to scores of .52 and .34 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 1.712. Pls.' Ex. 51 at 11, Table 1. As Delegate Jones testified, the changes from the Benchmark Plan made the district more “Richmond centric,” Trial Tr. 309:1 (Jones), which appears on its face to have enhanced the district's alignment with a distinct political subdivision and community of interest, Ints.' Ex. 94 at 2.

The Plaintiffs recognize that HD 69 has become more compact and retained its “core,” but argue that the district has become more compact “only by incorporating heavily African–American communities at the outskirts of the benchmark district.” Pls.' Post–Trial Reply at 15. Delegate McClellan also testified at trial that HD 69 had to satisfy the 55% BVAP floor, according to Delegate Jones. Trial Tr. 29:5–13 (Jones). But all of this is largely irrelevant. The question is whether the Commonwealth's consideration of race or a racial floor subordinated traditional, neutral criteria. Plaintiffs have offered no evidence to show subordination, relying instead on the erroneous view that proof of a 55% BVAP floor would be sufficient to carry their burden. As explained previously, it is not.

With respect to potential deviations from neutral criteria, it should be noted that HD 69 is not contiguous by land. Ints.' Ex. 94 at 2. However, the district contains multiple river crossings, id., and no evidence has been provided by the Plaintiffs to show that the district improperly combines two distinct communities of interest rather than uniting one community of interest. Moreover, the Plaintiffs have not provided any evidence that this split has diminished representation for communities on either side of the James. As such, there is no evidence that contiguity was “subordinated” to non-neutral criteria.

In short, the Plaintiffs have failed to carry their burden of proof with respect to HD 69, and the Court holds, as a matter of fact, that race did not predominate in the drawing of HD 69.

If anything, HD 69 seems to reflect the kind of district that might well be amenable to resolution on a motion for summary judgment based on a more structured understanding of the predominance inquiry, as provided above. See Abrams, 521 U.S. at 118, 117 S.Ct. 1925 (Stevens, J., dissenting) (“Any redistricting plan will generate potentially injured plaintiffs, ... [a]nd judges (unable to refer, say, to intent, dilution, shape, or some other limiting principle) will find it difficult to dismiss those claims[.]”).

4. District 70

HD 70 is found in the Richmond area and was represented by Delegate Delores McQuinn during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district contains parts of Chesterfield, Henrico, and Richmond City. Pls.' Ex. 50 at 69, Table 1. Although the number of county and city splits remained the same, redistricting increased the number of split VTDs from 2 to 3. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 70 has a core retention percentage of 67.31. Ints.' Ex. 14 at 83.

On its face, the district appears coherent and generally compact, perhaps with the exception of the “turret” on top of the district. HD 70 straddles the intersection of Richmond City, Chesterfield County, and Henrico County, Pls.' Ex. 66 at 3, with most of the boundaries therein drawn on the basis of precinct and VTD lines, Ints.' Ex. 94 at 3. The district had Reock and Polsby–Popper scores of .47 and .14 under the Benchmark Plan, which shifted to scores of .40 and .19 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. In other words, the district became slightly more elongated, but also removed some of its more convoluted and irregular boundaries in the process. The district's Schwartzberg score is 2.290. Pls.' Ex. 51 at 11, Table 1.

As the Plaintiffs contend, the redistricting “pull[ed] the district substantially out of the city of Richmond and pull[ed] it into the Chesterfield area and deeper into Henrico County.” Trial Tr. 142:7–10 (Ansolabehere). Plaintiffs believe that this shows a disregard for core retention, Pls.' Post–Trial Reply at 16, but this is precisely the reason the Court cautioned about “core retention” arguments above. Redistricting, by its very nature, involves the changing of districts. If a state completely abandoned its prior map and started from scratch, a hypothetical new “HD 70” might bear no resemblance whatsoever to the benchmark “HD 70,” but that would not—taken alone—be suspicious. Moreover, such a hypothetical would entail “removing” the entire population of HD 70 and then “adding” that entire number back. Again, nothing about that would be inherently suspicious.

The question is whether the boundaries—or the changes to the boundaries—are justifiable by reference to traditional, neutral criteria. Here, they are. Delegate Jones testified that HD 70's overall configuration was altered to better represent suburban interests—where population had expanded—and to cede more Richmond-centered population to HD 69 and HD 71. Trial Tr. 310:18–311:21 (Jones). The Plaintiffs' case supports that point. Id. at 142:11–20 (Ansolabehere) (“[HD 70 has] substantially shifted from being ... [a] plurality urban district to being a plurality suburban district.”). These represent objectively identifiable communities of interest.

Plaintiffs also argue that HD 70 was not under-populated before the redistricting process, but “the General Assembly added about 26,000 people and removed about 26,000 people in redrawing the district.” Pls.' Post–Trial Reply at 16. As discussed above, if properly populated districts were presumptively required to remain untouched, then all the other districts would need to wrap around them (in substantial disregard of neutral principles) in order to achieve population equality. See ante at 540–41 n. 20; accord Trial Tr. 310:7–311:2 (Jones). Nor is the substitution in population numbers particularly shocking. If a properly populated district must shift locations, then it will necessarily “remove” a large amount of people from its old location and “add” the same amount from its new location. That result seems rather obvious.

With respect to deviations, HD 70—like HD 69—is divided by the James, but contains a river crossing. Ints.' Ex. 94 at 3. And—like HD 69—Plaintiffs have offered no evidence to suggest that this has had any effect on representation or local communities of interest. As such, there is no evidence that contiguity was “subordinated” to non-neutral criteria.

The only facially odd deviation sits atop the northern edge of the district. This “turret” appears to deviate from districting norms, especially insofar as it pokes across Richmond City lines. However, Intervenors offered a simple, non-racial explanation for this deviation: Delegate McQuinn, the incumbent, lives there. As Delegate Jones testified: “[H]ad she not lived there, I could have actually had all of the 71st District in the city of Richmond because I could have taken these couple of precincts and there wouldn't have been any going into the Radcliffe precinct in Henrico County for 71.” Trial Tr. 311:3–17 (Jones).

In weighing the evidence, the Court recognizes that Delegate McClellan testified that HD 70 was drawn to comply with the 55% BVAP floor, id. at 29:5–13 (McClellan), but the legislature's pursuit of this goal is not the “predominate” criterion employed unless it subordinates all others. The Court finds that HD 70 is largely explained by reference to traditional, neutral districting criteria, and that the only deviation therefrom is explainable on the basis of “incumbent pairing prevention.” As a result, this Court holds, as a matter of fact, that race did not predominate in the drawing of HD 70.

5. District 71

HD 71 is found in the Richmond area and was represented by Delegate Jennifer McClellan during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district contains parts of Henrico and Richmond City. Pls.' Ex. 50 at 69, Table 1. Although the number of county and city splits remained the same, redistricting increased the number of split VTDs from 1 to 3. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 71 has a core retention percentage of 78.31, Ints.' Ex. 14 at 83, and is contiguous by land.

On its face, the district appears quite compact and generally follows normal districting conventions. The district had Reock and Polsby–Popper scores of .24 and .19 under the Benchmark Plan, which increased to scores of .33 and .24 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2.045. Pls.' Ex. 51 at 11, Table 1. The district remains bounded to the south by the James River—a natural geographic boundary—and became “more Richmond centric” with the 2011 redistricting thanks to the removal of Summit Court, Hilliard, and Stratford Hall precincts from its western edge. Trial Tr. 305:2–7 (Jones).

The district itself includes the Fan, moves east through Richmond's downtown, and continues up to Church Hill. The district contains the majority of the North Side, and contains one precinct in eastern Henrico County. Id. at 24:22–25:1 (McClellan).

The only facially evident deviations are along HD 71's eastern border. Here, the district's one Henrico precinct and the 701, 702, and 706 VTDs seem to form a set of “horns” on the eastern side of the district. See Pls.' Ex. 66 at 4; Ints.' Ex. 94 at 4.

In examining these deviations, it should first be noted that the northern-most horn adheres to the boundaries of Ratcliffe precinct, whereas the two other horns appear to adhere to the boundaries of VTDs 701, 702, and 706. Plaintiffs have argued that VTDs 701 and 702 were included because they were “heavily African American” and “very densely populated.” Id. 43:15–18 (McClellan). The Plaintiffs have not discussed whether Ratcliffe was added to capture black voters. Although Delegate McClellan testified that the 55% BVAP rule affected the districting decisions as to HD 71, id. 29:5–13 (McClellan), the Plaintiffs bear the burden of showing that the decision subordinated neutral criteria in the process.

Plaintiffs have not satisfied that burden. Delegate Jones offered a far more convincing reason for HD 71's eastern horns. As discussed above, Delegate McQuinn lives right on the border of VTDs 703 and 705. Ints.' Ex. 94 at 4. “[H]ad [Delegate McQuinn] not lived [in Richmond], I could have actually had all of the 71st District in the city of Richmond because I could have taken these couple of precincts and there wouldn't have been any going into the Radcliffe precinct in Henrico County for 71.” Trial Tr. 311:3–17 (Jones).

Plaintiffs also noted the split of VTD 505, which was previously wholly within HD 71. Id. at 42:20–43:4 (McClellan) (“That was split so that I got the VCU potion which is very densely populated, and [Delegate Carr] got the Oregon Hill neighborhood.”). Although a VTD split constitutes a deviation from neutral principles, the decision to split 505 advanced other neutral principles, such as compactness. Plaintiffs have not demonstrated that this split “subordinated” such neutral principles.

Delegate McClellan also spoke extensively about the removal of precinct 207 from her district, which split the Fan neighborhood. Id. at 39:14–20 (“207 and 208 are a majority of the Fan neighborhood where I live, and 207 was taken out [.]”). Precinct 207 had “highly democratic voter turnout,” and Delegate McClellan had “quite a base there[.]” Id. at 39:21–24.

But this split does not appear to substantially disregard neutral principles on its face. A local resident might wonder why the Fan straddled two House districts, but any observer of the map would see that precinct 207 was removed and replaced with precinct 204, making the district more compact.

Nor does that swap appear obviously racial. As Delegate McClellan testified, precinct 204 is “demographically similar to 207 racially.” Id. at 42:17–20. Delegate McClellan testified that she couldn't keep “any portion of 207” because it would “push the [BVAP] below 55 percent,” id. at 40:1–9, but if the 55% BVAP goal could be achieved without subordinating neutral principles on the whole, it does not matter what Delegate McClellan's personal preferences were.

And here, her personal preferences appeared in conflict with those of another legislator: Delegate Loupassi. According to Delegate Jones, Delegate Loupassi used to be on the Richmond City Council and his former ward abutted precinct 207 where he had strong support, so he “wanted that precinct in his district.” Id. at 305:15–307:12 (Jones). Delegate McClellan argued that adding precinct 207 to Delegate Loupassi's district “didn't help him” because he is a Republican, id. at 42:2–11 (McClellan), but Delegate Jones testified that Delegate Loupassi has “a broad base of support from the democratic side of the aisle” and had a personal “community of interest”—rather than partisan—connection to the area, id. at 485:7–14 (Jones).

There is a difference between pruning the edges of the political thicket and striding headlong into it. By verifying a district's overall compliance with neutral criteria that do not discriminate between citizens based on their race or other individualized characteristics, the Court fulfills its constitutional duty to ascertain whether state legislation violates the Equal Protection Clause. The Court should not, however, become embroiled in a credibility dispute between two legislators, especially when resolving that “factual” issue is unnecessary to find that neutral criteria predominated in the drawing of the district boundaries. HD 71 does not substantially disregard traditional, neutral districting principles, and that is sufficient for the Court to find that these principles were not subordinated to race. The existence of a 55% BVAP floor does not disturb that fact. Therefore, the Court holds, as a matter of fact, that race did not predominate in the drawing of HD 71.

The Plaintiffs also observe that a request from the Richmond Registrar was denied in HB 5001, and it is alleged that this change was rejected because the BVAP in HD 71 would have dropped to 54.8%. Pls.' Ex. 30. This provides strong evidence that a firm 55% BVAP rule was employed, as this Court has already held. See ante at 519 n. 7. But that finding does not imply that race “predominated” over neutral criteria in the drawing of HB 5005, especially because that particular “deviation” appears to have been addressed in HB 5005 itself. See Ints.' Ex. 7 at 2–3.

6. District 74

HD 74 is found in the Richmond area and was represented by Delegate Joseph Morrissey during the 2011 redistricting process. Under the Benchmark Plan, the district contained all of Charles City and parts of Henrico, Hopewell City, and Richmond City (as well as part of Prince George containing no population). Pls.' Ex. 50 at 69, Table 1. Under the Enacted Plan, the district now contains all of Charles City and parts of Henrico and Richmond City. Id. This decreased the number of county and city splits from 4 to 2, with the number of split VTDs remaining the same. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 74 has a core retention percentage of 80.08, Ints.' Ex. 14 at 83, and is contiguous by land.

On its face, the ax-shaped district arouses some suspicion. The “blade” of the ax encompasses all of Charles City, but the eastern “handle” is curious. The district had Reock and Polsby–Popper scores of .16 and .10 under the Benchmark Plan, which remained almost identical—with scores of .16 and .12–under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2 .839. Pls.' Ex. 51 at 11, Table 1. These low scores reflect the district's substantially elongated shape.

Despite its elongation, however, the district is not as unreasonable as it first appears. The north edge of the handle tracks the Henrico county line, while the lower edge is almost entirely retained within Henrico County. In fact, Delegate Jones' revision permitting the upper edge to track Henrico county lines “put some more good Republican precincts in there that the gentleman in the 97th did not want to lose[.]” Trial Tr. 317:13–17 (Jones). The district has also improved on neutral metrics over the last three districting cycles. See Ints.' Ex. 14 at 60. In particular, the 2011 plan removed the water crossing discussed in Wilkins v. West. See 264 Va. at 465–66, 571 S.E.2d 100; Trial Tr. 316:15–25 (Jones).

The Intervenors also noted that the BVAP percentage in the district had been lowered substantially from the Benchmark Plan. See Trial Tr. 313:3–315:6; Pls.' Ex. 50 at 72. But the fact that the BVAP percentage dropped does not, taken alone, indicate that race was not the predominate criterion influencing the district's construction. As the Plaintiffs observe, much of the black population ceded from HD 74 went to other Challenged Districts, such as HD 63 and HD 71. See Pls.' Post–Trial Reply at 17. Unlike in a racial vote dilution claim, a racial predominance inquiry does not necessarily concern itself with whether the BVAP went up or down. A district formed primarily to eject black voters would employ the same racial classification as a district formed primarily to include black voters.

In the end, however, the primary objection to this district amounts to a criticism that the district is too long. But predominance is not merely a beauty contest centered on Reock-style compactness. Although this district certainly does not earn high marks in a qualitative predominance analysis, the Plaintiffs have failed to demonstrate that neutral criteria were substantially disregarded in the formation of HD 74. The district contains all of Charles City and, for most of its length, has readily identifiable boundaries. Moreover, the shifting of black population into HD 63 and HD 71 largely improved HD 74's compliance with neutral criteria, such as contiguity and compactness.

Moreover, the district has retained roughly the same long shape since 1991. Trial Tr. 315:19–318:25 (Jones). Core retention alone cannot be used to save an otherwise offensive district, but it is worth holding in the balance if the familiarity of the boundaries has “allow[ed for the] development of relationships and communities of interest relative to election of delegates.” Wilkins, 264 Va. at 466, 476, 571 S.E.2d 100.

On the whole, the Court finds that the Plaintiffs have failed to meet the predominance inquiry's “demanding burden” to show that racial considerations subordinated both neutral criteria and other race-neutral explanations in the formation of HD 74. Therefore, the Court holds, as a matter of fact, that race did not predominate in the drawing of HD 74.

7. District 77

HD 77 is found in the Portsmouth area and was represented by Delegate Lionel Spruill during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district contains parts of Chesapeake and Suffolk. Pls.' Ex. 50 at 69, Table 1. The number of county and city splits remained the same, and the number of split VTDs decreased from 4 to 3. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 77 has a core retention percentage of 74.4. Ints.' Ex. 14 at 83.

At first glance, this jagged and elongated district is suspect. However, upon closer inspection, the top-right corner of the district hews to strange county lines, while many curious features on the lower side of the district track natural water boundaries and precincts that are themselves rather jagged and elongated. The district had Reock and Polsby–Popper scores of .18 and .17 under the Benchmark Plan, which shifted to scores of .19 and .15 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2.542. Pls.' Ex. 51 at 11, Table 1. With respect to neutral criteria, it appears that compliance therewith could still result in an inherently oddly-shaped district, but the record lacks guidance in this regard.

The record is similarly unclear and incomplete respecting deviations from traditional criteria. The district's large western chunk is admittedly attributable to a single precinct, but that does not answer why that whole half of the district is thrust so far into HD 76 as to nearly sever it in half. Ints.' Ex. 91 at 152. As Delegate Jones observed, the 76th and 77th districts share the most geographical boundary area on the map. Trial Tr. 334:2–4 (Jones).

Based on the alternative districting plans referenced by the Plaintiffs, see, e.g., Pls.' Ex. 23 at 40, it appears that it was possible to create the same number of performing districts in this region without resorting to this westward leap. So was this deviation necessary to reach the 55% BVAP floor (in which case, race might predominate), see Ints.' Ex. 92 at 15, or was this deviation motivated by a desire to remove Democrat performing precincts from Delegate Jones' district (in which case politics might predominate), see Ints.' Ex. 92 at 14? Or, is this overall structure attributable to the “knock-on” effects of avoiding pairing incumbents in this region? If so, incumbency considerations might predominate, political performance might predominate, or racial considerations might predominate. These are all questions that Plaintiffs bore the burden of answering. The Court is not in a position to guess based on the skimpy evidence submitted.

But, the record does show that the district's already-strange 2001 design was somewhat ameliorated in HB 5005 by moving the “Airport District” precinct from HD 77 to HD 76, id. at 336:7–12 (Jones), and “reuniting” the “old city of South Norfolk” at Delegate Spruill's request, id. at 334:8–10 (Jones), which allowed segments of the new district to more closely track county boundaries and water boundaries, Pls.' Ex. at 7. These changes also served political ends. The Airport District is primarily Republican, so this transfer helped Delegate Jones, Trial Tr. 336:7–12 (Jones), whereas the “old city of South Norfolk” surrounds Delegate Spruill's residence, which was seen as politically advantageous for him as well, id. at 336:1–4. Although the neighborhoods added around Delegate Spruill also contained meaningful black populations, Tanglewood, Oaklette, Norfolk Highlands, Indian River, and Johnson Park were all majority-white precincts. Ints.' Ex. 92 at 15.

The Court also observes that the district is not contiguous by land and does not appear to possess a water crossing within its bounds, see Pls.' Ex. 66 at 7; Ints.' Ex. 94 at 9, but Plaintiffs have offered no substantive evidence on whether this deviation relates in any way to the attainment of the district's BVAP level, which is 58.8% in the Executed Plan, see Pls.' Ex. 50 at 72.

Based on the testimony, evidence, and arguments, the Court cannot ascertain from the record whether race, politics, or other criteria predominated in the formation of HD 77. Frankly, if the presumption of correctness and good faith has any meaning, it is applicable in this instance. The Plaintiffs simply point to the threshold's attainment of the 55% BVAP floor, evidence of racial correlation, and a low compactness score to prove that race predominated. There is no evidence-based explanation to show how, if at all, the racial floor impacted the boundaries of HD 77 or why voters were placed there in the redistricting process. The Plaintiffs cannot hand the Court a stone and expect back a sculpture.

It is at least as likely that politics and traditional districting factors account for the configuration and composition of HD 77 as it is that race was responsible. Because the Plaintiffs have failed to provide evidence as to the ways in which racial considerations might have had a “direct and significant impact” on the District's formation, the Court finds that the Plaintiffs have failed to meet the burden of proof required to show that race predominated in the construction of HD 77.

8. District 80

HD 80 is found in the Portsmouth area and was represented by Delegate Matthew James during the 2011 redistricting process. Under the Benchmark Plan, the district contained parts of Chesapeake, Norfolk, and Portsmouth. Pls.' Ex. 50 at 69, Table 1. Under the Enacted Plan, the district now contains parts of Chesapeake, Norfolk, Portsmouth, and Suffolk. Id. This increased the number of county and city splits from 3 to 4 but decreased the number of split VTDs from 2 to 1. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 80 has a core retention percentage of 59.94. Ints.' Ex. 14 at 83.

At trial, Intervenors stated, “I think it's fair to say honestly that this district looks a little irregular.” Trial Tr. 349:3–5 (Intervenors). But “a little irregular” is “a little bit of an understatement.” The district is quite unusually configured. The district had Reock and Polsby–Popper scores of .39 and .26 under the Benchmark Plan, which experienced a substantial drop to scores of .26 and .11 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 3.054—the highest of all the Challenged Districts. Pls.' Ex. 51 at 11, Table 1.

Because the district makes little rational sense as a geographical unit, the Court will move directly to ascertaining the predominant purpose of the deviations. To begin, it is hard to identify what is now a “deviation” because it is hard to identify what is now the core of the district. The district is split by water twice without any apparent crossing enabling residents to stay within the district on either occasion. See Pls.' Ex. 66 at 8; Ints.' Ex. 94 at 10.

The Plaintiffs correctly note that HD 80's western border “winds its way around low BVAP precincts like Silverwood (14.9%), Churchland (8.3%), and Fellowship (14.2%) to capture high BVAP precincts such as Yeates (56.3%) and Taylor Road (48.8%).” Pls.' Post–Trial Brief at 19. Considering the district's attainment of the BVAP floor, this is the kind of detailed explanation that might lead the Court to find that racial considerations subordinated all others. In this case, however, the Plaintiffs' racial explanation must contend with other “dominant and controlling” considerations: incumbency protection as well as geographic features and a naval base.

In addition to the constraints imposed by the James River, the Atlantic Ocean, and the Norfolk naval base, the district needed to retain the residence of Delegate James while avoiding the residences of Delegate Johnny Joannou (HD 79) and then-Delegate Kenneth Alexander (HD 89). Ints.' Post–Trial Brief at 34. The general—and relatively simple—problem was “a loss of population” in the area and the need to move district boundaries “from the oceanfront back ... western to Suffolk” to capture population. Trial Tr. 349:6–11 (Jones). This problem became far more complex, however, because Delegates Alexander, Joannou, and Jones all live in relatively close proximity. Ints.' Ex. 94 at 10. To avoid pairing incumbents, Trial Tr. 350:23–24 (Jones), the westward shift of the districts had to wrap around the residences of the incumbents, resulting in the distortion found here. Thus, the map needed to “roll the population around ... to make sure Delegate Joannou had a sufficient number of residents in his district” and narrow the neck of the district before leaping further out westward to avoid Delegate Joannou while capturing Delegate James. Id. at 350:10–20.

That explanation addresses why neutral criteria were subordinated, but it does not provide the basis upon which voters were sorted into the corresponding districts. “Incumbent pairing prevention” may have resulted in “population rolls,” but an equal population goal itself is not part of the predominance balance. Alabama, 135 S.Ct. at 1271 (“[Predominance asks] whether the legislature placed race above traditional districting considerations in determining which persons were placed in

appropriately apportioned districts.”) (internal quotation marks omitted).

“Incumbency protection,” on the other hand, does provide an explanation for the amalgamation of precincts selected for HD 80. As the Intervenors explained:

Although HD80 could have been drawn to take territory from HD76—represented by Delegate Jones—the precincts there were Republican strongholds, and neither Jones nor HD80's representative, Democrat Matthew James, wanted that trade. Drawing HD80 into the former territory of HD79 gave those Democratic-leaning precincts to James, and not Jones. This arrangement made HD80 less compact than it would have been had it taken territory from Jones, but it was politically preferable. HD80 was also drawn to protect other incumbents, Johnny Joannou (HD79) and Kenneth Alexander (HD89), who resided near the borders they shared with HD80, making it impossible for HD80 to take territory to the north and northeast without pairing incumbents.

Ints.' Pre–Trial Brief at 16–17. Based on this record, it appears just as likely that precincts were selected for being highly Democratic and avoided for being highly Republican, see Ints.' Ex. 92 at 16, as it is that precincts were selected for being highly African–American and avoided for being highly Caucasian, see id. at 17. And, just because “the most loyal Democrats happen to be black Democrats” does not mean that a political gerrymander is thereby transformed into a racial gerrymander. Cromartie I, 526 U.S. at 551, 119 S.Ct. 1545.

On the whole, the Court finds that the Plaintiffs have not carried the burden of demonstrating that racial considerations subordinated neutral districting criteria and other non-racial districting criteria, including incumbent pairing prevention and incumbency protection. Although the existence of the BVAP floor itself weighs in favor of a racial predominance finding, the Court finds, as a matter of fact, that—qualitatively—the “dominant and controlling” factor dictating the construction of HD 80 was incumbency protection, and that race did not predominate in the drawing of HD 80.

9. District 89

HD 89 is found in the Norfolk area and was represented by then-Delegate Kenneth Alexander during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district is contained wholly within Norfolk. Pls.' Ex. 50 at 69, Table 1. There were no county or city splits and the number of split VTDs remained the same under both plans. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 89 has a core retention percentage of 76.86. Ints.' Ex. 14 at 84.

On its face, the district appears reasonably compact and generally follows precinct lines within Norfolk. The district had Reock and Polsby–Popper scores of .58 and .31 under the Benchmark Plan, which dropped to scores of .40 and .20 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2 .263. Pls.' Ex. 51 at 11, Table 1.

Although the district is not contiguous by land, it does contain water crossings within the district. See Pls.' Ex. 66 at 9; Ints.' Ex. 94 at 11. One of these crossings is largely to blame for the district's relative drop in compactness. Trial Tr. 144:9–145:1 (Ansolabehere). The added precinct—Berkley—contains a high BVAP percentage, see Ints.' Ex. 92 at 19, but is also relatively close to Delegate Alexander's residence, see Ints.' Ex. 94 at 11.

In addition, the district added a small “pipe” to its northernmost border, which includes a funeral home owned by Delegate Alexander. Trial Tr. 345:1–5. As Delegate Jones explained, Virginia state legislators are “part-time citizen legislators,” many of whom regularly interact with their constituents in their professional capacities. Id. at 346:2–18. As such, having a business within the district enables incumbents to more readily engage with their constituents.

Weighing all evidence, it appears that a couple of small deviations possibly could be attributable either to racial or to incumbency considerations, but the district's composition is predominantly attributable to traditional, neutral principles. Therefore, the Court holds that the Plaintiffs did not carry the burden of proving that race predominated in the drawing of HD 89.

10. District 90

HD 90 is found in the Norfolk area and was represented by Delegate Algie Howell, Jr. during the 2011 redistricting process. Under the Benchmark Plan, the district contained parts of Chesapeake, Norfolk, and Virginia Beach. Pls.' Ex. 50 at 69, Table 1. Under the Enacted Plan, the district now contains parts of Norfolk and Virginia Beach. Id. This decreased the number of county and city splits from 3 to 2 and the number of split VTDs remained the same. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 90 has a core retention percentage of 63.21. Ints.' Ex. 14 at 84.

On its face, the district appears to represent a reasonably compact geographic unit. The district had Reock and Polsby–Popper scores of .35 and .24 under the Benchmark Plan, which shifted to scores of .46 and .20 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 2.221. Pls.' Ex. 51 at 11, Table 1.

Apart from the district's two extensions into Virginia Beach and lack of land contiguity, HD 90 seems to largely comply with traditional, neutral districting conventions. See Pls.' Ex. 66 at 10; Ints.' Ex. 94 at 12. Even these “deviations,” however, must be viewed in context. Specifically, the 2011 redistricting plan improved the district's compliance with the “political subdivisions” criterion by removing a segment from Chesapeake. And, the southern appendage that reaches into Virginia Beach tracks the county line on its western border. Id. Moreover, one of the district's jumps across water connects parts of Norfolk. Id. As such, this land-contiguity failure simultaneously serves to unite a political subdivision and community of interest.

On the record submitted, neutral criteria appear to predominate. Even if the southern appendage reaching into Virginia Beach were enough for the district as a whole to exhibit a “substantial disregard” for neutral principles, it hardly appears that this offending piece of land could be viewed as racially driven. In fact, that segment of Virginia Beach contains some of the lowest BVAP percentages in the entire district. See Ints.' Ex. 92 at 21. Therefore, the Court holds that the Plaintiffs did not carry the burden to prove that race predominated in the drawing of HD 90, notwithstanding that it satisfies the 55% BVAP floor.

11. District 92

HD 92 is found in the Hampton area and was represented by Delegate Jeion Ward during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district is contained wholly within Hampton. Pls.' Ex. 50 at 69, Table 1. The district contains no county or city splits, and redistricting lowered the number of split VTDs in the district from 3 to 0. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 92 has a core retention percentage of 77.27. Ints.' Ex. 14 at 84.

On the whole, the Court finds it hard to imagine a better example of a district that complies with traditional, neutral districting principles. The district had Reock and Polsby–Popper scores of .28 and .15 under the Benchmark Plan, which increased to scores of .34 and .26 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. The district's Schwartzberg score is 1.970. Pls.' Ex. 51 at 11, Table 1.

As a result of the 2011 redistricting process, the district became more compact, reunified downtown Hampton, Trial Tr. 356:13–20 (Jones), and eliminated all precinct splits. Moreover, most of the district's southern border is marked by the waterfront and much of the district's western border now follows the Hampton boundary, making it easily identifiable to voters. See Pls.' Ex. 66 at 11; Ints.' Ex. 94 at 13. Although the district is not contiguous by land, it contains water crossings to allow voters to travel between parts of the district without traversing other districts. Id. The Court holds, as a matter of fact, that traditional, neutral criteria—not race—predominated in the construction of HD 92.

12. District 95

HD 95 is found in the Hampton area and was represented by Delegate Mamye BaCote during the 2011 redistricting process. Under both the Benchmark Plan and the Enacted Plan, the district contains parts of Hampton and Newport News. Pls.' Ex. 50 at 69, Table 1. Although the number of county and city splits remained the same, redistricting increased the number of split VTDs from 1 to 6. Pls.' Ex. 50 at 69–70, Tables 1, 2. HD 95 has a core retention percentage of 62.15, Ints.' Ex. 14 at 84, and is contiguous by land.

Their proximity notwithstanding, HD 92 and HD 95 share little in common. From bottom to top, the district begins by encompassing the full width of Newport News but soon departs from any observable neutral criteria. As the district moves northwest, a sliver attributable to the River precinct extends into HD 94 before the district works its way entirely over into Hampton City. There it remains for a period before extending briefly back into Newport News via the South Morrison precinct. After retreating back into Hampton City the district then hits water and York County, which it weaves around before running up through the middle of Newport News in a narrow spike. See Pls.' Ex. 66 at 12; Ints.' Ex. 94 at 14. If there is any reasonably neutral explanation for the route followed, this Court was not informed. The district had Reock and Polsby–Popper scores of .43 and .28 under the Benchmark Plan, which dropped to scores of .14 and .14 under the Enacted Plan. Ints.' Ex. 15 at 15, Table 9. This rendered HD 95 the least compact district on the map under the Reock metric. See Ints.' Ex. 14 at 76–78, Table 9. The district's Schwartzberg score is 2.657. Pls.' Ex. 51 at 11, Table 1.

Rather than attempting to explain the district through neutral criteria, the Intervenors themselves acknowledge that the construction of the district was “significantly political.” Trial Tr. 359:6–8 (Jones). According to Delegate Jones, the district's movement north follows heavily Democratic precincts and then narrowly jumps through two Republican precincts in order to capture another strongly Democratic voting area at its northernmost tip. Id. at 369:1–4; Ints.' Ex. 92 at 24. Moreover, the district's eastward “zig” followed by its westward “zag” managed to avoid including the residence of Delegate Robin Abbott in HD 95. See Ints.' Ex. 94 at 14. This avoided pairing female Democratic incumbents and, in conjunction with the partisan maneuvering above, placed Delegate Abbott in a more heavily Republican swing seat. Trial Tr. 369:6–372:12 (Jones). As Intervenors explained: “HD95 was crafted carefully to avoid taking HD94's Republican precincts and instead take Democratic-leaning population left behind by HD93 and reach into precincts surrounded by HD93 to dilute Democratic voting strength in that area.” Ints.' Pre–Trial Brief at 18. The Court finds that explanation persuasive. Where there is a correlation between race and party, the burden is upon the Plaintiffs to dislodge the evidence showing that voters were sorted predominantly on the basis political preference rather than race. Delegate Jones had access to political performance data as well as racial data. As the Intervenors asked during closing argument: “[I]f race was the principal factor, why [did the legislature] pass by all these areas which have more black voters [in the southern part of the peninsula and] go up there [to the northern tip of the district]? ... We don't hear any analysis from the other side on that point. There's no contradictory testimony.” Trial Tr. 827:6–19 (Intervenors). On the evidence submitted, political advantage (based on partisan performance data) has been shown to have been the dominant and controlling consideration guiding the district's unorthodox boundaries. As a result, the Court holds, as a matter of fact, that race did not predominate in the construction of HD 95.

V. CONCLUSION

For the foregoing reasons, the Court holds that each of the twelve Challenged Districts withstands constitutional scrutiny under the Equal Protection Clause, and judgment will be entered for the Defendants and the Intervenor–Defendants.

It is so ORDERED.

BARBARA MILANO KEENAN, Circuit Judge, dissenting:

Today, despite the Supreme Court's clear warning against the mechanical use of racial targets in redistricting, this court upholds the Virginia General Assembly's application of a one-size-fits-all racial quota to twelve highly dissimilar legislative districts. This quota was used to assign voters to districts based on the color of their skin without the constitutional protection afforded by strict scrutiny.

I recognize that the legislature in this case did not have the benefit of the Supreme Court's decision in Alabama, and I do not doubt that individual legislators acted in good faith in the redistricting process. Nevertheless, the resulting legislative enactment has affected Virginia citizens' fundamental right to vote, in violation of the Equal Protection Clause. Accordingly, I would invalidate Virginia's 2011 redistricting plan.

I.

Redistricting decisions are almost always made with a “consciousness of race,” Bush v. Vera, 517 U.S. 952, 958, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996) (principal opinion of O'Connor, J.), and such awareness does not necessarily result in a violation of the Equal Protection Clause, see Miller v. Johnson, 515 U.S. 900, 916, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995). However, when a legislature is “motivated” by racial considerations, this inherently suspect system of racial classification must satisfy the rigorous requirements of strict scrutiny. Miller, 515 U.S. at 916, 115 S.Ct. 2475.

A plaintiff asserting a race-based equal protection claim in a redistricting case has the burden of proving “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. (emphasis added). Under this predominance test, a plaintiff must show that “the legislature subordinated traditional race-neutral districting principles ... to racial considerations.” Id.; see also Ala. Legislative Black Caucus v. Alabama, ––– U.S. ––––, 135 S.Ct. 1257, 1271, 191 L.Ed.2d 314 (2015) (“[T]he ‘predominance’ question concerns which voters the legislature decides to choose, and specifically whether the legislature predominantly uses race as opposed to other, ‘traditional’ factors when doing so.” (emphasis in original)). When a legislature has “relied on race in substantial disregard of customary and traditional districting principles,” such traditional principles have been subordinated to race. Miller, 515 U.S. at 928, 115 S.Ct. 2475 (O'Connor, J., concurring).

Strict scrutiny is required when race was the predominant factor that categorically was accorded priority over race-neutral districting factors. As the Supreme Court has explained, traditional factors have been subordinated to race when “[r]ace was the criterion that, in the State's view, could not be compromised,” and when traditional, race-neutral criteria were considered “only after the race-based decision had been made.” Shaw v. Hunt, 517 U.S. 899, 907, 116 S.Ct. 1894, 135 L.Ed.2d 207 (1996) (Shaw II ); see also Page v. Va. Bd. of Elections, No. 3:13cv678, 2015 WL 3604029, at *7 (E.D.Va. June 5, 2015). Thus, while a redistricting plan may reflect certain traditional districting criteria, that plan nevertheless remains subject to strict scrutiny when those criteria have been subordinated to a process that has sorted voters primarily by race.

Contrary to the majority's view, this predominance inquiry does not require that the use of race in drawing district boundaries be in “conflict” with traditional districting criteria. Maj. Op. at 524. In fact, the race of a voter often correlates with other districting considerations, including partisan preference, incumbency protection, and communities of interest. See Bush, 517 U.S. at 964, 116 S.Ct. 1941 (principal opinion). The conclusion logically follows, therefore, that racial sorting frequently will not be in “conflict” with these and other districting criteria.

Because such districting criteria can be used to mask racial sorting, courts must carefully examine the evidence under the test for predominance articulated in Miller and Shaw II. Under that test, race necessarily predominates when the legislature has subordinated traditional districting criteria to racial goals, such as when race is the single immutable criterion and other factors are considered only when consistent with the racial objective. Shaw II, 517 U.S. at 907, 116 S.Ct. 1894.

II.

This case presents a textbook example of racial predominance, in which a uniform racial quota was the only criterion employed in the redistricting process that could not be compromised. This one-size-fits-all quota automatically made racial sorting a priority over any other districting factor. Although a legislature is entitled to a presumption of good faith, this presumption must yield when the evidence shows that citizens have been assigned to legislative districts primarily based on their race. See Miller, 515 U.S. at 915–16, 115 S.Ct. 2475; Page, 2015 WL 3604029, at *8 (“[T]he good faith of the legislature does not excuse or cure the constitutional violation of separating voters according to race.” (citation omitted)). For this reason, I disagree with the majority's conclusion that a uniform racial quota merely is “evidence” of predominance, and instead would hold that the existence of such a widely applied quota establishes predominance as a matter of law.

A.

I first observe that while the parties have engaged in a semantical debate whether the 55% BVAP threshold was an “aspirational target” or a “rule,” the evidence presented at trial clearly established that the legislature employed the 55% BVAP figure as a fixed, non-negotiable quota. Three individual delegates testified regarding their understanding of the mandatory nature of the quota. Pl.Ex. 33 at 45 (Sen.Dance); Trial Tr. at 70 (Sen. Dance); Trial Tr. at 29–30 (Del.McClellan); Trial Tr. at 92 (Del. Armstrong). And, despite Delegate Jones' trial testimony that the 55% BVAP figure was merely an “aspirational ... rule of thumb,” he promoted the plan during the House of Delegates floor debates as having achieved a 55% minimum BVAP for all majority-minority districts. Trial Tr. at 491; Pl.Ex. 35 at 42, 66, 70, 72, 108, 113. The legislators' subjective understanding that the 55% figure operated as a mandatory floor further was corroborated by the fact that, in the 2011 plan, the BVAP in most of the twelve challenged districts converged toward 55% while each district satisfied the 55% BVAP floor. Pl.Ex. 50 at 72 Table 4; DI Ex. 15 at 14.

Delegates Dance and Armstrong no longer serve in the House of Delegates, though Dance currently serves as a senator in the Virginia Senate. Trial Tr. at 65, 90.

B.

The “disregard of individual rights” is the “fatal flaw” in such race-based classifications. Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 320, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978) (opinion of Powell, J.); see also City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493, 109 S.Ct. 706, 102 L.Ed.2d 854 (1989) (opinion of O'Connor, J.) (explaining that the “rights created by the first section of the Fourteenth Amendment are, by its terms, guaranteed to the individual. The rights established are personal rights.” (quoting Shelley v. Kraemer, 334 U.S. 1, 22, 68 S.Ct. 836, 92 L.Ed. 1161 (1948))). By assigning voters to certain districts based on the color of their skin, states risk “engag[ing] in the offensive and demeaning assumption that voters of a particular race, because of their race, think alike, share the same political interests, and will prefer the same candidates at the polls.” Miller, 515 U.S. at 911–12, 115 S.Ct. 2475 (quoting Shaw v. Reno, 509 U.S. 630, 647, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993) (Shaw I )) (internal quotation marks omitted). Quotas are especially pernicious embodiments of racial stereotypes, because they threaten citizens' “ ‘personal rights' to be treated with equal dignity and respect.” Croson, 488 U.S. at 493, 109 S.Ct. 706 (opinion of O'Connor, J.).

Because individual voters suffer the harm alleged in a racial sorting claim, I disagree with the majority's contention that “intentional [ ] dilut[ion] [of a] group's meaningful participation in the electoral process” is required to sustain an equal protection challenge like the one the plaintiffs have raised in this case. Maj. Op. at 531 (emphasis omitted). See Miller, 515 U.S. at 911–13, 115 S.Ct. 2475.

Here, the plan contravened the rights of individual voters by applying a one-size-fits-all racial quota for black voters in twelve highly dissimilar districts, without regard to the characteristics of the voters or of their communities. The 55% quota thus is a classic example of race-based stereotyping and unequal treatment prohibited by the Equal Protection Clause.

The Supreme Court's skepticism of racial quotas is long-standing. See generally Croson, 488 U.S. 469, 109 S.Ct. 706 (minority set-aside program for construction contracts); Bakke, 438 U.S. 265, 98 S.Ct. 2733 (higher education admissions). However, the Court has yet to decide whether use of a one-size-fits-all racial quota in a legislative redistricting plan or, in particular, use of such a quota well exceeding 50%, establishes predominance as a matter of law under Miller.

The Court recently has cautioned against “prioritizing mechanical racial targets above all other districting criteria” in redistricting. Alabama, 135 S.Ct. at 1267, 1272–73. Although the Court in Alabama did not decide whether the use of a racial quota well exceeding 50%, of itself, can establish predominance, the Court made clear that such “mechanical racial targets” are highly suspicious. Id. at 1267; see id. at 1272–73 (discussing racial targets as part of narrow tailoring analysis). After issuing this admonishment and identifying several errors in the district court's analysis, the Court ultimately remanded the case to the district court to reconsider the question of predominance. Id. at 1270–74.

I disagree with the majority's conclusion that the Supreme Court in Alabama would not have remanded the case if the use of racial thresholds in that case constituted predominance as a matter of law. See Maj. Op. at 524. Appellate courts frequently remand issues to trial courts for reconsideration when a trial court initially has employed an incorrect legal analysis.

The uniform racial quota employed in the present case is more suspicious on its face than the racial thresholds at issue in Alabama. The legislature in Alabama sought to maintain preexisting racial percentages specific to each district with the aim of avoiding retrogression under Section 5. Id. at 1263. In contrast, the racial quota used in the present case was applied indiscriminately to all twelve districts irrespective of the particular characteristics of those districts. The Virginia plan's one-size-fits-all quota thus raises even more serious concerns that the legislature's districting decisions were driven primarily by race.

In view of the Virginia legislature's application of a single racial quota to numerous districts in the case before us, this court is not presented with the question whether a particular fixed BVAP percentage would trigger strict scrutiny if applied to a single district. Nor is this court asked to decide whether strict scrutiny is required every time a legislature intentionally creates a majority-minority district. See Bush, 517 U.S. at 998, 116 S.Ct. 1941 (Kennedy, J., concurring) (reserving the question); Alabama, 135 S.Ct. at 1272 (declining to decide whether “the intentional use of race in redistricting, even in the absence of proof that traditional districting principles were subordinated to race, triggers strict scrutiny”); League of United Latin Am. Citizens v. Perry, 548 U.S. 399, 517, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006) (Scalia, J., concurring in the judgment in part and dissenting in part) (“[W]hen a legislature intentionally creates a majority-minority district, race is necessarily its predominant motivation and strict scrutiny is therefore triggered.”).

Instead, the more narrow question before this court is whether strict scrutiny is required when a uniform racial quota of 55% has been applied by a legislature in drawing twelve legislative districts that are highly dissimilar in character. Here, because traditional districting criteria were considered solely insofar as they did not interfere with this 55% minimum floor, see Shaw II, 517 U.S. at 907, 116 S.Ct. 1894, the quota operated as a filter through which all line-drawing decisions had to pass. Such a racial filter necessarily had a discriminatory effect on the configuration of the districts, because it rendered all traditional criteria that otherwise would have been “race-neutral” tainted by and subordinated to race. See Miller, 515 U.S. at 916, 115 S.Ct. 2475 (holding that when “race-neutral considerations are the basis for redistricting legislation, and are not subordinated to race, a State can defeat a claim that a district has been gerrymandered on racial lines” (citation and internal quotation marks omitted)). Under these circumstances, although a legislature may take into account traditional districting criteria, race-neutral application of those criteria becomes impossible and all decisions necessarily are affected by race. Therefore, I would hold that the plaintiffs have established as a matter of law under Miller that race predominated in the legislative drawing of each of the challenged districts, and I would apply strict scrutiny in examining the constitutionality of those districts.

I therefore disagree with the majority's contention that this question was answered by the principal opinion in Bush and by the majority in Shaw II. Maj. Op. at 528, 532. Neither Bush nor Shaw II presented the unique factual circumstances at issue in this case, namely, the application of an across-the-board 55% racial quota to twelve variable districts.

Although the majority is correct that the district at issue in Shaw II exhibited more facial irregularities than the districts here, such distinctions do not preclude application of relevant principles from the case. Shaw II, 517 U.S. at 905–06, 116 S.Ct. 1894. Maj. Op. at 532. As the Court noted in Shaw II, the fact that a legislature is able to achieve certain traditional districting goals in a race-based plan “does not in any way refute the fact that race was the legislature's predominant consideration.” Shaw II, 517 U.S. at 907, 116 S.Ct. 1894.

III.

In stark contrast, the majority's predominance analysis accepts the use of this facially suspicious racial quota. In doing so, the majority places an unwarranted burden on the plaintiffs to show that the quota had identifiable effects on the drawing of particular district lines. The majority thus effectively would require the plaintiffs to present an alternative legislative map showing how lines could have been drawn differently without imposing the 55% quota. Such an onerous burden, however, far exceeds the required showing for establishing predominance.

I further observe that the plaintiffs presented testimony from Delegate McClellan that she did not propose certain desired changes to the plan because the resulting lines would not comply with the 55% quota. Trial Tr. at 41.

Additionally, under the majority's test, visual inspection of a district would be fatal to an equal protection claim if the district's boundaries appear to be consistent with traditional criteria, irrespective of direct evidence that the line-drawing was racially motivated at the outset. Thus, as a result of the majority's analysis, and its requirement that the use of race be in actual “conflict” with traditional districting criteria, future plaintiffs asserting a racial sorting claim will be restricted to challenging districts that manifest extreme line-drawing unexplainable on race-neutral grounds, like the district at issue in Shaw I.

As the Supreme Court has emphasized, however, a district that is bizarre in shape is not the constitutional harm prohibited by the Equal Protection Clause. Rather, as stated above, the constitutional harm results from individual voters being sorted into districts based on the color of their skin. Miller, 515 U.S. at 911–15, 115 S.Ct. 2475 (explaining that it is “the presumed racial purpose of state action, not its stark manifestation, that [is] the constitutional violation”). By requiring that use of race actually “conflict” with traditional redistricting criteria, the majority's predominance test often will fail to identify constitutionally suspect racial sorting.

IV.

In reviewing a redistricting plan, courts typically examine whether a plan complies with traditional districting factors, such as compactness and contiguity, when evaluating whether there is evidence of racially motivated decision making. See Shaw I, 509 U.S. at 647, 113 S.Ct. 2816 (traditional districting factors are not constitutionally required, but “they are objective factors that may serve to defeat a claim that a district has been gerrymandered on racial lines”). When a legislative district is bizarre in shape, that fact “may be persuasive circumstantial evidence that race for its own sake, and not other districting principles, was the legislature's dominant and controlling rationale in drawing its district lines.” Miller, 515 U.S. at 913, 115 S.Ct. 2475. Here, however, the majority relies on shape and other traditional districting factors to uphold the 2011 plan, even in the face of the overwhelming, direct evidence of racial motivation evidenced by the use of a one-size-fits-all racial quota.

The majority's analysis is not aided by Cromartie II and Bush. In Cromartie II, the Court described the predominance inquiry as requiring plaintiffs to show that a district's boundaries were drawn “because of race rather than because of” other districting criteria. Easley v. Cromartie, 532 U.S. 234, 257, 121 S.Ct. 1452, 149 L.Ed.2d 430 (2001) (emphasis omitted). However, a legislative district necessarily is crafted “because of race” when a racial quota is the single filter through which all line-drawing decisions are made.

Similarly, the principal opinion in Bush explained that “[s]ignificant deviations from traditional districting principles ... cause constitutional harm insofar as they convey the message that political identity is, or should be, predominantly racial.” Bush, 517 U.S. at 980, 116 S.Ct. 1941 (principal opinion). The import of this language is obvious. The harm caused by racial stereotyping is apparent when racial sorting manifests itself in odd district boundaries that are visible to any observer. But the incidence of constitutional harm is not limited to the presence of a district that is odd in shape. In the present case, the legislature's use of a racial quota resulted in constitutional harm, because that methodology “convey[ed] the message that political identity is, or should be, predominantly racial.” Id.

I also disagree with the intervenors' implicit suggestion that approval by incumbent legislators in the challenged districts somehow rescues the plan from a finding of racial predominance. The Voting Rights Act (VRA) and the Equal Protection Clause are intended to protect the rights of the individual voter, not to promote the self-interest of incumbents in majority-minority districts. See League of United Latin Am. Citizens, 548 U.S. at 440–41, 126 S.Ct. 2594 (“If ... incumbency protection means excluding some voters from the district simply because they are likely to vote against the officeholder, the change is to benefit the officeholder, not the voters.”). To the contrary, immunizing incumbents from challenge could entrench them in overwhelmingly safe districts and undermine the representatives' accountability to their constituents. One can easily imagine how such entrenchment could harm minority voters by discouraging challengers from running and by preventing voters from electing a new candidate who better represents their interests. “Packing” minority voters into a particular majority-minority district for the purpose of protecting the incumbent also can reduce minority voters' ability to influence elections in nearby districts.

I recognize that the plaintiffs in this case do not raise a vote dilution claim under Section 2 of the VRA, but instead bring an “analytically distinct” racial sorting claim under the Equal Protection Clause. See Miller, 515 U.S. at 911, 115 S.Ct. 2475 (citing Shaw I, 509 U.S. at 652, 113 S.Ct. 2816). I note the potential detrimental effects of the plan only to highlight that a so-called “benign” racial quota, ostensibly intended to benefit minority voters, may in fact have the opposite effect.

A true predominance analysis also is not affected by the fact that, at the time of the 2010 census, nine of the twelve challenged districts already had a BVAP of 55% or higher. DI Ex. 15 at 13–14 & Table 8; Pl.Ex. 50 at 9 ¶ 17, 72 Table 4. Even assuming that such figures could protect the configuration of those nine districts in the 2011 plan, the three remaining districts still would be subject to strict scrutiny. Moreover, given the significant population deficits in most of the challenged districts, our inquiry must focus on “which voters the legislature decide[d] to choose” when moving voters between districts in order to achieve population equality. Alabama, 135 S.Ct. at 1271 (emphasis in original). Here, the legislature's decision to move certain voters in order to maintain a preexisting 55% BVAP floor in the new plan is still a “mechanically numerical” method of redistricting that is subject to strict scrutiny. See id. at 1273.

I therefore conclude that the majority's approach effectively and improperly places on plaintiffs asserting racial predominance in redistricting a burden never assigned by the Supreme Court. Under the majority's analysis, plaintiffs now will be required to show circumstantial evidence of racial motivation through “actual conflict” with traditional districting criteria, when such plaintiffs already have presented dispositive direct evidence that the legislature assigned race a priority over all other districting factors.

V.

Even upon applying its heightened predominance standard, the majority concludes that race was the predominant factor in the drawing of District 75. I would hold that, under the majority's test, the same conclusion of predominance holds true for neighboring District 63 as well.

As a result of the “drastic maneuvering” required to reach a 55% BVAP in District 75, portions of a county previously in District 63 were shifted into District 75, a move that the majority agrees was “avowedly racial.” Trial Tr. at 74, 80; Maj. Op. at 553. The plan compensated for this loss of BVAP in District 63 by adding to the district new areas with high BVAP concentrations. Trial Tr. at 81–83. Due to the changes in the 2011 plan, District 63 experienced a startling reduction in compactness and an increase in the number of split cities, counties, and VTDs. DI Ex. 15 at 15 Table 9; Pl.Ex. 50 at 7, 70 Table 2, 71 Table 3. This and other evidence showed that implementation of the 55% racial quota had a marked impact on the configuration of both Districts 63 and 75.

VI.

I further conclude that none of the challenged districts can survive the test of strict scrutiny, because the legislature's use of the 55% quota was not narrowly tailored to achieve a compelling state interest in any of the challenged districts. See Miller, 515 U.S. at 920, 115 S.Ct. 2475. Evidence of narrow tailoring in this case is practically non-existent.

Assuming that compliance with the VRA is a compelling state interest, attempts at such compliance “cannot justify race-based districting where the challenged district was not reasonably necessary under a constitutional reading and application” of federal law. Id. at 921, 115 S.Ct. 2475; see also Bush, 517 U.S. at 977, 116 S.Ct. 1941 (principal opinion). Thus, narrow tailoring requires that the legislature have a “strong basis in evidence” for its race-based decision, that is, “good reasons to believe” that the chosen racial classification was required to comply with the VRA. Alabama, 135 S.Ct. at 1274 (emphasis omitted).

In the present case, the intervenors presented virtually no evidence supporting the need for application of a 55% BVAP in any of the challenged districts. In fact, Delegate Jones even had difficulty articulating the original source of the 55% figure. Trial Tr. at 429, 431, 443, 490–95. The only evidence suggestive of any tailoring involved District 75. Delegate Jones testified that he conducted a “functional analysis” of Delegate Tyler's primary and general election results in 2005, and considered the significant prison population in that district, which together supported the imposition of a 55% racial floor. Trial Tr. at 323–24, 430, 458–59, 462–67, 494; Pl.Ex. 40 at 39 (Del. Tyler). However, Jones' statements were merely general and conclusory in nature and, therefore, fell far short of demonstrating a “strong basis in evidence” for the application of a racial quota. Not only did the 2005 elections occur six years prior to the 2011 redistricting, but Tyler ran unopposed in the two elections since, casting significant doubt on Jones' contention that District 75 was so competitive that a minority-preferred candidate required at least a 55% BVAP to be re-elected from 2011 onward. See Pl.Ex. 50 at 85 Table 14. And, critically, Jones failed to provide any explanation of how his “functional” review led him to conclude that a 55% BVAP was required in District 75 to ensure compliance with the VRA.

The evidence supporting the use of the 55% racial quota in the remaining challenged districts was even weaker. The House of Delegates did not conduct an analysis regarding the extent of racially polarized voting in any of these districts. Trial Tr. at 469. Although Delegate Jones stated that he was aware of low registration rates among black voters, he also admitted that he did not review voter registration figures when drawing the plan. Trial Tr. at 462–64. Nor did he examine minority turnout rates in most of the challenged districts, or consider state Senate districts, congressional maps, or other maps that had been pre-cleared or rejected by the Department of Justice. Trial Tr. at 462–69. And, in attempting to justify imposition of the 55% BVAP quota in District 63, Jones stated that he “t[hought] there was a primary” in which Delegate Dance ran as an independent, which results he reviewed, but he did not specify how those results led him to select a 55% BVAP threshold in District 63. Trial Tr. at 466–68. Such unsubstantiated and general comments plainly do not constitute the strong basis in evidence required to satisfy strict scrutiny.

Finally, I do not think that the outcome of this case, in favor of either party, is dependent on any of the expert testimony. However, I pause to note that I find the testimony offered by Dr. Katz to be singularly unpersuasive on the issue of narrow tailoring. Dr. Katz admitted that he provided only a “crude” analysis of the likelihood that a candidate preferred by minority voters would be elected. Trial Tr. at 531. According to Dr. Katz, this “crude” method demonstrated that a 55% BVAP correlates with an 80% chance of electing a black candidate. DI Ex. 16 at 18–19; Trial Tr. at 532.

I agree with the majority's criticism that Dr. Ansolabehere did not consider any factors other than race and politics as predictors of VTD inclusion in the challenged districts. Maj. Op. at 551. Nevertheless, my conclusion, that the legislature's use of the 55% racial quota per se establishes predominance as a matter of law, renders Dr. Ansolabehere's opinions regarding VTD movement superfluous to a proper predominance analysis.

Dr. Katz' crude analysis exhibits two glaring flaws. First, it underrepresents the likelihood that the preferred candidate of minority voters would be elected by evaluating only the likely success of black candidates, when minority voters had elected non-minority delegates in certain of the challenged districts. Trial Tr. at 532–34, 549–51, 769–71. Second, and more fundamentally, Dr. Katz' analysis is flawed because the VRA does not guarantee the success of a candidate of a particular race in a given election. Rather, the VRA ensures that minority voters do not “have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice,” and that minority voters retain their existing ability to elect their preferred candidates. 52 U.S.C. § 10301(b); League of United Latin Am. Citizens, 548 U.S. at 428, 126 S.Ct. 2594 (VRA Section 2); 52 U.S.C. § 10304(b); Alabama, 135 S.Ct. at 1272 (VRA Section 5).

Although my conclusions do not depend on the testimony of Dr. Ansolabehere, I am not persuaded by the majority's dismissal of Dr. Ansolabehere's racial polarization analysis. See Maj. Op. at 559 n. 37. In particular, I credit Dr. Ansolabehere's conclusion that none of the challenged districts required a 55% BVAP in order to ensure minority voters' opportunity to elect their preferred candidate. Trial Tr. at 203.
--------

For these reasons, I would find that the record utterly fails to show that the legislature had a “strong basis in evidence” for using the 55% racial quota in any of the challenged districts. Accordingly, I would hold that all the districts fail the test of strict scrutiny.

VII.

The promise of the Equal Protection Clause is the guarantee of true equality under the law, enforced by our courts for the protection of our citizens irrespective of the power of any governmental entity. The Virginia legislature's use of the racial quota in this case violated this core constitutional principle in the absence of a strong basis in evidence supporting its race-based decision. Thus, I would invalidate Virginia's 2011 redistricting plan. I respectfully dissent.