dissenting.
In these two cases, the Alabama Legislative Black Caucus, various elected black officials, and others challenge the redistricting plans for the Alabama House and Senate. Specifically, they challenge each majority-black House and Senate District in addition to Senate Districts 7,11 and 22. Despite the multiplicity of claims and responses in this litigation, in my view the two cases are actually quite simple. As explained below, the drafters of these plans labored under the false belief that § 5 of the Voting Rights Act (“VRA”), 42 *1313U.S.C. § 1973c, required them to adopt for each majority-black district a particular percentage of black population, ranging as high as 78.1% black. Therefore, the drafters sifted residents by race across the State of Alabama in order to achieve for each such district, where possible, what I believe can only be characterized as naked “racial quotas.” Bush v. Vera, 517 U.S. 952, 976, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996) (plurality opinion) (approvingly quoting Vera v. Richards, 861 F.Supp. 1304, 1341 (S.D.Tex.1994)) (internal quotation marks omitted).
I must reject Alabama’s redistricting plans for essentially five reasons. First, Alabama’s use of such a quota for any district warrants strict scrutiny. Second, the State’s argument that its solution was required by § 5 is not supported by the correct interpretation of that statute. Third, in any event, because Alabama is no longer subject to preclearance under § 5, that statute cannot serve as the basis for the quotas. Fourth, the quota for each district in which it was used is not grounded in current political, social, and racial conditions in that district that would warrant its use.. Fifth, the State’s redistricting plans “threaten[ ] to carry us further from the goal of a political system in which race no longer matters — a goal that the Fourteenth and Fifteenth Amendments embody, and to which the Nation continues to aspire.” Shaw v. Reno, 509 U.S. 630, 657, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993). “As a Nation we share both the obligation and the aspiration of working toward this end.” Miller v. Johnson, 515 U.S. 900, 927, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995). I respectfully dissent.
I. BACKGROUND
I agree with the majority that the complaints in this matter are best construed as bringing three sets of claims: claims of vote dilution in violation of § 2 of the Voting Rights Act, 42 U.S.C. § 1973; claims that the plans were drafted with invidious racial discrimination in violation of the Fourteenth and Fifteenth Amendments; and claims that the plans constitute a racial gerrymander in violation of the equal protection clause of the Fourteenth Amendment. As to the last, I would read both complaints as alleging that the plans in their entirety constitute racial gerrymanders and, as stated, also specifically challenging each majority-black House and Senate District in addition to Senate Districts 7, 11 and 22. Because I believe the plaintiffs are entitled to relief on their racial-gerrymandering claims and because that relief would require the drafting of new plans, I do not reach the other claims.
The majority opinion thoroughly recites the testimony and evidence presented in these consolidated cases. I will therefore summarize only the facts relevant to the racial-gerrymandering claims on which I would strike down these plans.
These cases arise out of Alabama’s process for redistricting its state legislative maps following the 2010 census. Three men steered that process: Senator Gerald Dial, Representative Jim McClendon, and Randy Hinaman. Dial and McClendon were the co-chairs of the Permanent Legislative Committee on Reapportionment. Hinaman is a political consultant who drew the actual maps using specialized computer software called Maptitude under the supervision of Dial and McClendon.
The Reapportionment Committee adopted guidelines to govern the reapportionment process, setting forth a number of factors to consider in drafting the new maps. One key factor was compliance with § 5 of the VRA. See Tr. Vol. I at 54 (Dial testifying that “[tjrying to meet the voting rights requirements was the basis *1314on which I drew those plants]”); id. at 118 (Dial’s “first qualification” was “not regressing minority districts”); id. at 42 (Dial believed “our job was to get a plan ... that would meet Justice”); Tr. Vol. Ill at 220-1 (McClendon’s goal was Justice Department approval). Other factors included a newly adopted rule limiting the total population deviation among districts to 2%, preserving of the core of existing districts, avoiding conflicts between incumbents, ensuring compactness, and accommodating incumbent preferences.
Under § 5, a covered jurisdiction must seek preclearance of new redistricting plans from either the Attorney General of the United States or the United States District Court for the District of Columbia. Alabama was a covered jurisdiction until the Supreme Court’s decision in Shelby County v. Holder, 133 S.Ct. 2612, decided after this case was filed but before trial.
Each of the drafters shared the same very specific (but incorrect) understanding of what compliance with § 5 involved: they believed they would need (1) to maintain the same number of majority-black districts as had existed under the 2001 redistricting scheme; and (2), more importantly for this case, to maintain, to the extent possible in each such district, the same percentage of black residents as that district was determined to have had when the 2010 census data were applied to the 2001 district lines. See Tr. Vol. Ill p. 142 (Hinaman explaining that “[w]hen I was adding population to majority black districts, my goal was not to retrogress the number that they had in 2001, meaning 2010 census, as applied to the 2001 lines”); Tr. Vol. I at 54 (Dial agreeing that his understanding of retrogression “required ... that you maintain the black majority percentage” as measured by the 2001 districts with 2010 census data); Dial Dep., APX 66, at 81 (Dial stating that lowering the black population by even one percentage point would have been retrogression); Tr. Vol. Ill at 221 (McClendon stating that “we tried to look at the 2010 census, overlay it on the districts, and try not to change the percentages of the citizens, the black citizens”). The drafters acknowledged that this might not always be possible; but they believed § 5 required them to match the previous percentage of black population insofar as it was possible to do so.
This understanding meant that for each majority-black district, the drafters adopted a district-specific racial quota. For example, if the 2010 census data indicated that a particular district as drawn in 2001 was 75% black in 2010, then the drafters believed that § 5 required them to draw that district’s new boundaries such that it remained 75% black.
These quotas, supposedly required by § 5, posed a challenge for the drafters. Many of the majority-black districts as drawn in 2001 were ‘under-populated’ once the 2010 census data were applied. ‘Under-population’ refers to a district which has fewer residents than is required by the constitutional principle of one-person-one-vote. See Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964) (holding that equal protection requires that state legislative districts be roughly equal in population). This meant that if no changes were made to their boundaries those districts would have less population than the Constitution required.
In order to address the under-population of the majority-black districts, the drafters needed to add people, often many thousands of people. The drafters’ quotas for those districts meant, in turn, that the large majority of those newcomers would need to be black. To illustrate, if 10,000 people needed to be added to the 75% black district discussed above in order to *1315address its under-population, then per the drafters’ understanding of retrogression under § 5 they would need to add at least 7,500 black people to maintain the same percentage of black residents overall. See Affidavit of Gerald Dial, APX 63, at 4 (“... we had to add population that was both contiguous to the old district line and had about the same percentage of black population in it”); Affidavit of Jim McClendon, APX 64, at 3 (same).
This problem was exacerbated by the rule the committee adopted mandating that the population of the least and most populated districts differ by no more than 2% of the ideal population.1 The “ideal population” refers to the population each district would have if the State’s total population were evenly divided among them. Prior to the current round of redistricting, the Alabama legislature had consistently used a 10% total deviation rule in drafting its state legislative redistricting plans. Because of the new 2% rule, under-populated districts needed to add even more population than they would have needed with a more traditional 10% deviation rule; often, the 2% rule required thousands of additional residents.2 And adding those thousands of additional residents meant, in turn, that the drafters would need to find many more black people to satisfy their quotas.
Indeed, the challenge the drafters created for themselves was enormous, as the sheer numbers show. The drafters’ (incorrect) understanding of the requirements of § 5, in combination with their adoption of the 2% rule, meant that they needed to find over 120,000 additional black people to add to the majority-black House Districts.3’
*13164 This amounted to 19.7% of the black people in the State of Alabama who did not already live in a majority-black House District.5 The story is similar in the Senate.6 There, the drafters would need over 106,-000 additional black people to satisfy their twin goals for the majority-black districts.7 This is 15.8% of the black population of Alabama not already living in a majority-black Senate District.8
But even those percentages understate the challenge the drafters faced. Many of the black people not already living in majority-black districts were likely dispersed around the State; but the drafters sought to find those additional 120,000 black people in areas contiguous to the existing majority-black House Districts. Affidavit of Gerald Dial, APX 63, at 4; Affidavit of Jim McClendon, APX 64, at 3. If a given majority-black district were surrounded by overwhelmingly black areas that were not already part of one of the majority-black districts, then this task might prove relatively easy. For example, if a majority-black district needed 10,000 additional residents, and 75% of those .residents needed to be black to comply with the drafters’ quota, then adding a nearby neighborhood containing 10,000 people of whom 75% were black would fit the bill. But if the available areas near a majority-black district were racially diverse, or even predominantly white, then a more artful approach would be required to add the requisite population without lowering the percentage of black residents. Thus, for example, if the 75% black district were surrounded by areas in which only 50% of the population was black, then the drafters would need to find some method of sorting the black people from the white in order to add population that was 75% black. They would not be able to add *1317population en masse, but would need to finely craft lines in order to include enough black residents and exclude enough white ones.
With this view of the challenges he faced, Hinaman set to work drafting these plans. Underscoring the focus on compliance with the drafters’ understanding of § 5, he began his work by drawing the majority-black districts. The maps Hinaman drew contain 27 majority-black House Districts (“HD”) and eight majority-black Senate Districts (“SD”); this is the same number as existed under the 2001 plan.9
However, the districts are not all drawn in the same place. Faced with underpopulation in the majority-black districts, Hinaman concluded that he could not draw the same number of majority-black districts in Jefferson County without lowering the percentages of black population in those districts. See Hinaman Dep., APX 75, at 60 (the alternative to moving HD 53 was to “retrogress every one of those districts by adding in adjoining white areas”). That outcome was unacceptable; Hinaman, like the other drafters, believed that § 5 required him to meet the quota of the previous percentage of black population. Hinaman never actually tried to draw nine majority-black districts in Jefferson County, and so could not say how much lower the black percentages would have been; in fact, he believed it would have been possible to draw nine such majority-black districts. Id. at 60-61, 86. Instead of doing so, he concluded that the prospect of lower black percentages in the majority-black House Districts left him no choice: he had to eliminate one of the districts, HD 53, from Jefferson County, relocate it elsewhere, and use its black population to maintain the black percentages in the remaining Jefferson County districts.
Hinaman took similar action in Montgomery County. There, he was again confronted with under-population in the majority-black districts. This time, his approach was to eliminate HD 73, a plurality-black district, and use its substantial black voter population to maintain the level" of black population in the majority-black districts. As McClendon put it, “The minority districts in Montgomery were underpopulated” and so “[w]e needed to pick up minorities from somewhere.” McClendon Dep., APX 67, at 90. In other words, the previous HD 73, like the previous HD 53, was eliminated in order to satisfy the drafters’ racial quotas for the surrounding majority-black districts.10
Eliminating two districts and redrawing them in another part of the State created conflicts between incumbents. Under the new plan, the incumbents of HD 53, Demetrius Newton, and HD 73, Joe Hubbard, were left living in another legislator’s district.11 One of the drafters’ goals was to avoid such conflicts among incumbents. But as the elimination of these two districts demonstrates, the drafters’ priority *1318of meeting the racial quotas for majority-black districts trumped the goal of incumbent protection. As Hinaman testified when he was asked about separating incumbents: “Well, it was a goal. It was a nice goal. Didn’t always work out.” Tr. Vol. Ill at 161.
Hinaman took such dramatic steps to achieve the racial quotas, which he believed § 5 required, throughout the State. One glaring example is SD 26. That district, represented by Senator Quinton Ross, was under-populated by nearly 16,-000 people from the ideal population. With the 2010 census data, his district was already 72.75% black. At trial, Ross noted that if only white people had been added to repopulate his district, it still would have been about 64% black; Ross testified he would have been comfortable with an even lower percentage of black residents.12 Instead, the Senate plan added 15,785 people to his district, of whom only 36 were white; 14,806 were black. That is, just .2% of the net population addition to SD 26 was white; as the Alabama Democratic Conference plaintiffs note, “This compares unfavorably to the 1.00 percent of the black voters who were left in the City of Tuskegee after the racial gerrymander in Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (I960).” ADC Pfs. Proposed Findings of Fact and Conclusions of Law (Doc. No. 195-1), at 64. Ross testified that, given the demographics of the area, to locate so many black people and so few white people near his district, “You have to make sure you look hard to find them.” Tr. Vol. II at 128.
Hinaman indeed went out of his way to locate so many black people in the vicinity of SD 26 and to exclude white people from the district. Ross testified that the population in the current SD 26 is highly segregated and that the boundaries in the new plan track those racial lines. Ross stated that, despite the under-population of his district, the new plan actually split precincts that were already part of SD 26, moving white portions of those precincts out of his district while retaining only the black portions; in other words, despite needing a huge number of new residents, Hinaman removed white residents already living in SD 26. This followed the pattern of the precinct splits between Ross’ district and white-majority SD 25, which gave the black-majority portions of precincts to SD 26 and the white-majority portions to SD 25. The new SD 26 wraps around and excludes a portion of Montgomery which Ross testified is predominantly white, and the resulting district is not compact. See Map, SDX 476, State Demonstrative Exhibit Tab 6. By taking these various steps to remove white residents and add black ones, the drafters achieved and even exceeded their quota of 72.75% black for this district; in the new plan, SD 26 is over 75% black.
Hinaman followed a similar pattern of ‘looking hard’ for black people throughout the State in order to achieve the quotas. Precinct splits like those Ross described were a major characteristic of these plans. One of the plaintiffs’ expert witnesses, William Cooper, testified that there was “massive precinct splitting” statewide. Tr. Vol. II at 105.13 Indeed, about 25% of all precincts were split, and dozens of precincts were split among two, three, or four different districts. Id.
*1319Furthermore, Hinaman split those precincts largely along racial lines. See Arrington Report, Ex. 323, at 37 (noting the precinct splits “mainly divided heavily minority blocks from heavily white ones”); Hinaman Dep., APX 75, at 117-8 (stating that avoiding retrogression and creating majority-black districts were two of his three normal reasons to split precincts). Indeed, Hinaman acknowledged that he used precinct splits in hunting for black residents. He agreed that, to avoid retrogression, he would first “reach out to find black precincts.” Tr. Vol. Ill p. 141-2. But, he testified, when adding whole precincts lowered the percentage of black residents in the new district, he would split precincts to achieve the racial quotas. Id. at 143.
In fact, the evidence establishes that Hinaman principally relied on the race of individuals living in split precincts in deciding how to distribute them among districts. As I will explain, this is clear because in deciding how to split precincts, Hinaman had access to the racial makeup of mapping units smaller than precincts; but he had no accurate data about the political makeup of those sub-precinct units. Thus, the fact that Hinaman’s precinct splits track race cannot be explained by race correlating with party affiliation, for example. For in choosing which residents of split precincts would be in majority-black districts and which would not, Hinaman knew those residents’ race but not their political affiliation or voting history.
Precincts are the basic unit of elections; in each precinct, voters cast their votes at a specified location. Precincts in turn are made up of census blocks, which are the smallest geographic unit the United States Census Bureau uses for statistics from its decennial population survey.
At the precinct level, there are “political” data: for example, what candidates won and lost in that precinct in past elections, and by how many votes. This can show the partisan breakdown of the population of a precinct. But because of the secret ballot, no political data are available at the block level. Cooper, who has 25 years of experience drawing redistricting maps, explained that there were no accurate political data at that level because “you don’t really know where ... individuals who turned out to vote for X or Y candidate actually live” within a precinct. Tr. Vol. II at 104.14
By contrast, because demographic data are collected by the Census Bureau on a house-by-house basis and aggregated at the census block level, accurate racial data are available for particular census blocks. Another of plaintiffs’ experts, Theodore Arrington, noted that the census file from which Hinaman was working was “rich in racial data.” Arrington Report, NPX 323, at 37. Hinaman acknowledged that “when I was working on the [mjajority black districts I had the racial data.” Hinaman Dep., APX 75, at 112.
When Hinaman split precincts, as he did in SD 26, he relied on those racial data. He could not have done so based on political data, because none were available at the census block level. The only reasonable conclusion is that he split precincts based on the information that was avail*1320able: namely, demographic data reflecting the race of the individuals who lived in each census block. And the evidence establishes that the reliance on racial data at the census block level was common statewide: as Cooper observed, because so many precincts were split, “[c]learly there was a focus on census blocks.” Tr. Vol. II at 106. In other words, clearly there was a focus on race.
The drafters’ belief that § 5 required a particular quota for each majority-black district also meant that they would reject suggestions from legislators when suggested changes failed to achieve those quotas. For example, Senator Marc Keahey (SD 22) testified at trial that he submitted to Dial close to ten proposed maps for his district, to each of which the incumbents of the neighboring black-majority SD 23 and SD 24 had agreed. Dial had told Keahey that he would consider such proposals with the other senators’ support as long as they did not cause retrogression in majority-black districts; Keahey understood his proposals to meet that requirement. However Dial rejected all of Keahey’s proposals as retrogressive. Eventually, Keahey came to understand the source of the disagreement. Keahey had sought to match the previous percentage of black residents in those districts using the 2000 census data, because that is what he thought Dial required. But Dial’s understanding of § 5 meant that the new districts needed to match the percentage of black population in the 2001 districts with the 2010 census data. That is, in Dial’s view Keahey had used the wrong quota; because Keahey’s proposals did not achieve the correct quota, the drafters would not even consider them, despite the preferences of all the affected incumbents.
Keahey’s testimony demonstrates that Dial’s adherence to particular quotas was strikingly rigid. For example, one of the majority-black districts that borders Keahey’s district is SD 23, represented by Senator Sanders. Using the 2000 census data, as Keahey originally did, SD 23 was 62.31% black. 2001 Plan Statistics, APX 4, at 5. But using Dial’s actual standard, namely the 2001 districts with 2010 census data, SD 23 was 64.79% black. APX 7. Thus if Keahey offered a suggested change, to which Senator Sanders had agreed, that maintained 62.31% black population in SD 23 but did not achieve 64.79% black population, Dial would automatically reject such a change as “retrogressive.” 15 Indeed, Dial agreed that he rejected Keahey’s proposals on just this basis. He testified that according to his understanding of § 5, a drop of even one percentage point would be retrogressive. Dial Dep., APX 66, at 81. The use of a rigid quota could not be clearer.
In sum, then, the drafters believed that § 5 required them to sift through surrounding districts for black people in order to achieve particular racial quotas for each district. In seeking to meet those quotas, they eliminated existing districts, created conflicts between incumbents, ignored legislators’ preferences, and split a huge volume of precincts.
The drafters were quite successful in achieving their quotas. See Comparisons of 2001 and 2012 plans with 2010 census data, APX 6 & APX 7. Of the majority-black districts, the black percentage of the population in 13 House Districts16 and *1321three Senate Districts17 is within one percentage point of the goal of maintaining the same percentage of black residents even after repopulating the districts, often with thousands of new individuals. Seven House Districts18 and three Senate Districts 19 have an even higher percentage of black population than before.
In some districts, the rigidity of these quotas is on full display.20 HD 52 needed an additional 1,145 black people to meet the quota; the drafters added an additional 1,143. In other words, the drafters came within two individuals of achieving the exact quota they set for the black population, out of a total population of 45,083; those two people represent .004% of the district. In HD 55, the drafters added 6,994 additional black residents, just 13 individuals more than the quota required, and in HD 56 they added 2,503 residents, just 12 individuals more than the quota required, both out of a total population of 45,071. In the Senate, SD 23 contains 116 more black individuals than were needed to achieve the drafters’ quota of adding an additional 15,069 black people, out of a total population of 135,338; in other words, the difference between the quota and the additional black population in the ultimate plan represents .086% of the district.
The plans were enacted over the opposition of every black legislator in the State, and precleared by the Justice Department. Two sets of plaintiffs, including the Alabama Legislative Black Caucus, the Alabama Democratic Conference, and a number of individuals, brought these lawsuits challenging their legality.
II. DISCUSSION
The majority rejects the plaintiffs’ racial-gerrymandering claims on two bases: first, that race was not the predominant factor in drawing these plans; and, second, that even if strict scrutiny applies, the maps were narrowly tailored to achieve the compelling purpose of compliance with § 5 of the VRA. I disagree on both points.211 will first review the standard for *1322a racial-gerrymandering claim, and will then address the majority’s conclusions in turn.
A. Legal Standard
The Fourteenth Amendment’s equal protection clause provides that, “No State shall ... deny to any person within its jurisdiction the equal protection of the laws.” U.S. Const., Arndt. 14, § 1. The central purpose of the clause “is to prevent the States from purposefully discriminating between individuals on the basis of race.” Shaw v. Reno, 509 U.S. 630, 642, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993). “ ‘[A]t the heart of the Constitution’s guarantee of equal protection lies the simple command that the Government must treat citizens as individuals, not as simply components of a racial, religious, sexual or national class.’ ” Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 730, 127 S.Ct. 2738, 168 L.Ed.2d 508 (2007) (plurality opinion of Roberts, C.J.) (quoting Miller v. Johnson, 515 U.S. 900, 911, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995)) (internal citation omitted); see also Fisher v. Univ. of Texas at Austin, — U.S. -, 133 S.Ct. 2411, 2422, 186 L.Ed.2d 474 (2013) (Thomas, J., concurring) (“The Constitution abhors classifications based on race because every time the government places citizens on racial registers ... it demeans us all”) (internal quotation marks omitted).
In Shaw v. Reno, the Supreme Court recognized a claim under the equal protection clause that was “analytically distinct” from somewhat similar vote-dilution claims. 509 U.S. at 652, 113 S.Ct. 2816. Where a purposeful-dilution claim alleges that a redistricting plan was enacted with the purpose of “minimiz[ing] or canceling] out the voting potential of racial or ethnic minorities,” Mobile v. Bolden, 446 U.S. 55, 66, 100 S.Ct. 1490, “the essence of the equal protection claim recognized in Shaw is that the State has used race as a basis for separating voters into districts.” Miller, 515 U.S. at 911, 115 S.Ct. 2475. If race is so used, then the redistricting plan is subject to strict scrutiny.
Redistricting legislation generally does not explicitly refer to race; rather, it “classifies tracts of land, precincts, or census blocks, and is race neutral on its face.” Hunt v. Cromartie, 526 U.S. 541, 547, 119 S.Ct. 1545, 143 L.Ed.2d 731 (1999). In addition, the Supreme Court has consistently recognized that redistricting is a complex process, and that legislatures will nearly always be “aware” of racial demographics. Miller, 515 U.S. at 916, 115 S.Ct. 2475. Such awareness of race is never enough to trigger strict scrutiny.
Instead, the Court has required that a Shaw plaintiff show “that race was the predominant factor motivating the legislature’s decision to place a significant number of voters within or without a particular district.” Miller, 515 U.S. at 916, 115 S.Ct. 2475. More specifically, a plaintiff must establish 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.” Id.
The plaintiff in such a case may carry this burden in a number of ways. In some instances, circumstantial evidence, including the shape of the district and the demographic splits created by its borders, is sufficient to establish that the boundaries are “unexplainable on grounds other than race.” Hunt, 526 U.S. at 546, 119 S.Ct. 1545 (citing Shaw, 509 U.S. 630, 113 S.Ct. 2816). In other cases, there is direct evidence that race was the predominant factor in the legislature’s decision-making. See, e.g., Miller, 515 U.S. at 917-8, 115 S.Ct. 2475. But, in any event, the rule is *1323clear: if race was the predominant factor, strict scrutiny applies.
To survive strict scrutiny, a racial classification must be narrowly tailored to achieve a compelling state interest. Id. at 904, 115 S.Ct. 2475. While such scrutiny is not “strict in theory, but fatal in fact,” Johnson v. California, 543 U.S. 499, 514, 125 S.Ct. 1141, 160 L.Ed.2d 949 (2005) (internal quotation marks omitted), the State is required to establish the “most exact connection between justification and classification.” . Parents Involved, 551 U.S. at 720, 127 S.Ct. 2738 (majority opinion of Roberts, C.J.) (quotation marks omitted).
B. Race Predominated
Race was the predominant factor in the drafters’ decisions to draw the majority-black districts as they did.22 This is clear from an examination of the racial quotas they adopted, even standing alone. Such quotas, under the circumstances of this case and without any justification other than race, require the court to strike down this plan unless the State can satisfy strict scrutiny. Furthermore, although no additional evidence is necessary in this case, there is ample circumstantial evidence that various other districting factors were subordinated to race in the drafting of those majority-black districts. The majority’s arguments to the contrary are unpersuasive; strict scrutiny must apply.
i.
From start to finish, Hinaman, Dial and McClendon were focused on drafting majority-black districts that would be precleared under § 5 of the VRA. See Tr. Vol. I at 113 (Dial’s “first qualification” was “not regressing minority districts”); Tr. Vol. III at 220-1 (McClendon’s goal was Justice Department approval); Tr. Vol. III at 145 (Hinaman “was concerned about ... retrogression that would be looked upon unfavorably by the Justice Department under Section 5”).
They believed that § 5’s non-retrogression principle required them to maintain (as nearly as possible) the same percentage of black residents in any given majority-black district as that district had when the 2010 census data was applied to the 2001 district boundaries. See Tr. Vol. I at 54 (Dial agreeing that his understanding of retrogression “required ... that you maintain the black majority percentage” as measured by the 2001 districts with 2010 census data); Tr. Vol. Ill at 221 (McClendon stating that “we tried to look at the 2010 census, overlay it on the districts, and try not to change the percentages of the citizens, the black citizens”); Tr. Vol. Ill p. 145 (Hinaman, when asked to define retrogression, stating that he would look to “2010 census as applied to 2001 lines” and then “tried to be as close to that as possible”).
The direct evidence of the drafters’ goals and intentions comes straight from their lips. Dial, for example, had the following exchange at his deposition:
*1324“Q. So you did not want the total population of African-Americans to drop in [SD 23]?
“A. That’s correct.
“Q. Okay. And if that population dropped a percentage, in your opinion that would have been retrogression?
“A. Yes, sir.
“Q. So if — And I’m not saying these are the numbers, but I’m just saying if Senator Sanders’ district had been 65 percent African-American, if it dropped to 62 percent African-American in total population, then that would have been retrogression to you?
“A. In my opinion, yes.
“Q. And so that’s what you were trying to prevent?
“A. Yes.”
Dial Dep., APX 66, at 81. By their own candid admissions, the drafters acknowledged that they understood § 5 to mean that for each majority-black district they needed to achieve a set percentage of black population, defined by the percentage in that district as drawn in 2001 with the 2010 census data.
This kind of requirement has a name: racial quota. “Quotas impose a fixed number or percentage which must be attained.” Grutter v. Bollinger, 539 U.S. 306, 335, 123 S.Ct. 2325, 156 L.Ed.2d 304 (2003) (internal quotation marks omitted). The Supreme Court’s equal protection cases have time and again treated this type of “rigid racial quota,” City of Richmond v. J.A. Croson Co., 488 U.S. 469, 109 S.Ct. 706, 102 L.Ed.2d 854 (1989), with the highest skepticism. See id.; Grutter, 539 U.S. at 335, 123 S.Ct. 2325; Regents of Univ. of California v. Bakke, 438 U.S. 265, 315, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978) (Opinion of Powell, J.).
The drafters did not deny adopting such percentages or quotas. To the contrary, when confronted with the suggestion that partisan politics, rather than race, actually motivated the how the majority-black districts were drawn, the drafters vehemently denied it. When asked about the use of partisan data at his deposition, Dial explained:
“... what I did was begin with the minority districts to ensure they were not regressed, and each one of them had to grow. And as we did those, then I filled in the blanks around those with what was left of the districts. So I didn’t look at partisan to say how many Republicans are here or how many Democrats are here. I began my process by filling in the minority districts, not to do away with any of those and not to regress any of those. And as they grew, we made sure that they grew in the same proportion [of black residents] that they had or as close to it as possible. And what was left, we just — it was basically fill in the blanks with what was left.”
Dial Dep., APX 66, at 19-20. When asked at trial, “Weren’t you aware when you were drawing the [S]enate [Districts that the Republicans’ goal in this state was to maintain your super majorities in the Legislature?,” Dial denied that was his goal:
“... I established my goal as maintaining the minority districts and passing a plan that would meet Justice Department. That was my ultimate goal, and that’s what I worked for ... The numbers themselves were actually to insure that we did not regress the minority districts, and we filled in what was left.”
Tr. Vol. 1 at 61. In this case, time and again the drafters have emphasized that in drawing the majority-black districts they were motivated by a desire to obtain preclearance. And time and again they have articulated their understanding that § 5 *1325meant they needed to achieve racial quotas.
ii.
These percentages or quotas in the State’s legislative plans must fall of their own weight unless they can survive strict scrutiny. Bush v. Vera, the Supreme Court’s first effort to apply the Miller predominant-factor standard to a legislative plan in which many of the districts were being challenged, is particularly instructive. At first blush, it might appear that Vera is of little precedential value because the decision is so fractured, with a plurality opinion, three concurrences, and two dissents. However, the array of opinions is helpful for two reasons: First, they offer a nuanced view of how the Justices think the Miller predominant-factor standard should be applied. Second, and perhaps most importantly here, under all of the opinions, the quotas in Alabama's legislative redistricting plans would fail unless they can survive strict scrutiny. Or, to put it another way, no matter how one defines the Miller predominant-factor standard, the quotas warrant strict scrutiny.
First, there is the Vera plurality opinion. The opinion first acknowledged that it was confronted with an array of factors that went into a legislative redistricting plan. The opinion therefore explained that, “Because it is clear that race was not the only factor that motivated the legislature to draw irregular district lines, we must scrutinize each challenged district to determine whether the District Court’s conclusion that race predominated over legitimate districting considerations, including incumbency, can be sustained.” Vera, 517 U.S. at 965, 116 S.Ct. 1941. Similarly here, because it is contended that race was not the only factor that motivated the Alabama Legislature to draw the challenged district lines the way it did, we must scrutinize each challenged district individually to determine whether race predominated over legitimate districting considerations.
For each district, the critical question is whether race was “the predominant factor motivating the legislature’s [redistricting] decision” for that district. Id. at 959, 116 S.Ct. 1941 (emphasis and internal quotation marks omitted). In this case, we are confronted with districts in which (1) the drafters announced a racial percentage or quota; (2) the drafters achieved that quota; and (3) there is no explanation for those actions other than race. For example, it is clear that one factor and one factor alone explains the fact that SD 26 is over 75% black: race. Nothing else explains that percentage. And the same is true for SD 24. One factor and one factor alone explains the fact that SD 24, with a quota of 62.8% black, is 63.3% black: race. And the same is true for SD 23. One factor and one factor alone explains the fact that SD 23, with a quota of 64.79% black, is 64.81% black: race.
Also, the same is true for majority-black House Districts. One factor and one factor alone explains the fact that HD 55, with a quota of 73.54% black, is 73.6% black: race. One factor and one factor alone explains the fact that HD 67, with a quota of 69.14% black, is 69.2% black: race. One factor and one factor alone explains the fact that HD 57, with a quota of 68.49% black, is 68.5% black: race.
The State has not offered, and on this record cannot offer, any alternative explanation that would explain away the State’s apparent use of race. In Vera, the State had argued that incumbency protection, rather than race, had motivated what appeared to be racial gerrymandering. Because the State had pointed to a race-neutral factor that might correlate to race, the plurality found it necessary to examine each district closely to determine whether *1326that race-neutral factor explained the apparently racial lines the State had drawn better than race did. But here the State has offered no race-neutral explanation for the black percentages in the majority-black districts; no race-neutral explanation for why SD 26, for example, is 75% black. In fact, Dial explicitly rejected the idea that partisan politics, rather than the racial quotas, motivated the drawing of the majority-black districts. In the absence of such an explanation, the plurality in Vera would have no difficulty striking down districts like those presented in this case, namely districts drawn to achieve racial quotas.
Second, there is Justice O’Connor’s concurring opinion. While she wrote separately to explain why “compliance with the results test of § 2 of the Voting Rights Act (VRA) is a compelling state interest” and why “that test can coexist in principle and in practice with Shaw,” Vera, 517 U.S. at 990, 116 S.Ct. 1941 (O’Conner, J., concurring), she accepted the plurality opinion’s understanding of the Miller predominant-factor standard; this, of course, is unremarkable, since she wrote the plurality opinion as well. Therefore, for the same reason that Alabama’s quotas warrant strict scrutiny under the plurality opinion, they warrant the same under Justice O’Connor’s concurrence.
Third is Justice Kennedy’s concurrence. While he joined the plurality opinion, he expressly and unequivocally stated in his discussion of the Miller predominant-factor standard that, “In my view, we would no doubt apply strict scrutiny if a State decreed that certain districts had to be at least 50 percent white, and our analysis should be no different if the State so favors minority races.” Id. at 996, 116 S.Ct. 1941 (Kennedy, J., concurring). Similarly, because Alabama has decreed that SD 26 must be 72% black, no matter what the other demographics are, and because it drew SD 26 so as to make it 75% black, it would be difficult, if not impossible to explain to Justice Kennedy why SD 26 would not be subject to strict scrutiny. And the same would apply to SD 23’s 64% quota, SD 24’s 62% quota, and so forth. And the same would apply to HD 55’s 73% quota, HD 57’s 68% quota, and so forth.
Fourth, there is Justice Thomas’s concurrence, in which Justice Scalia joined. Justice Thomas stated that “Georgia’s concession that it intentionally created majority-minority districts was sufficient to show that race was a predominant, motivating factor in its redistricting.” Id. at 1000, 116 S.Ct. 1941 (Thomas, J., concurring in the judgment) (emphasis added). He further stated that, “Strict scrutiny applies to all governmental classifications based on race, and we have expressly held that there is no exception for race-based redistricting.” Id. One does not need to think long to know what Justice Thomas’s views on Alabama’s quotas would be.
Fifth and finally, four Justices in Vera dissented and concluded that the challenged legislative plan did not warrant strict scrutiny. Though the majority in this case reaches a similar result about the Alabama plan, I do not think the majority can take solace from the reasoning of the Vera dissenters. Justice Stevens wrote a dissent in which Justices Ginsburg and Breyer joined; Justice Souter wrote a separate dissent, but stated that he agreed with Justice Stevens’s application of the Miller predominant-factor standard. For this reason, I will discuss only Justice Stevens’ opinion. Justice Stevens stated that “the typically fatal skepticism that we have used to strike down the most pernicious forms of state behavior” need not apply only if three conditions are met: “the state action (i) has neither the intent nor effect of harming any particular group, (ii) is not *1327designed to give effect to irrational prejudices held by its citizens but to break them down, and (iii) uses race as a classification because race is relevant to the benign goal of the classification.” Id. at 1010, 116 S.Ct. 1941 (Stevens, J., dissenting) (quotation marks omitted). There is absolutely nothing in the record to support the conclusion that these conditions are present as to Alabama’s redistricting plans. Indeed, it appears that the only racial dynamic at play in Alabama’s plans is that white members of the Alabama legislature, and the white ones alone, have expressly and specifically targeted black legislators and the members of their districts for difference in treatment solely because of the race of those legislators and over those black legislators’ deep and vocal objections.
This aspect of this case, in particular, bears a disturbing similarity to Gomillion, 364 U.S. 339, 81 S.Ct. 125, where the Supreme Court condemned the redrawing of Tuskegee, Alabama’s municipal boundaries by white members of the Alabama Legislature so as to exclude almost all the black citizens of that community. Admittedly, the there are some fundamental differences between this case and Gomillion: This case is based on the Fourteenth Amendment, and Gomillion was based principally on the Fifteenth Amendment, although it has also been read as a Fourteenth Amendment case, see Shaw, 509 U.S. at 645, 113 S.Ct. 2816; this case involves a Shaw claim, whereas Gomillion involved an invidious discrimination claim, although again Shaw itself drew on Gomillion; and, in this case, blacks are being brought into or kept in a district solely because of their race, and, in Gomillion, blacks were being excluded from a district solely because of their race. Nevertheless, in both cases, white members of the Alabama legislature, and the white ones alone, expressly and specifically targeted black people and treated them differently in the drawing of district lines solely because of the race. And despite the fact these black people object to, and are even offended by, this racial targeting and treatment, they are powerless to do anything about it politically. Or, to put it another way, a white majority has unwelcomely imposed its will on how a black minority is to be treated politically.
The injustice of this was poignantly brought home in the testimony of Senator Vivian Figures, an African-American, at the trial of this case. Senator Figures acknowledged at trial that the Republican Party had won a supermajority in the 2010 elections “fair and square.” Tr. Vol. II at 51. She therefore “expected to be outvoted” as a Democrat. Id. But what she did not expect was for her “voice to be squashed.” Id. This voicelessness, this complete powerlessness to do anything about the fact that white members of the Alabama legislature expressly and specifically targeted her and treated her differently in the drawing of her district lines solely because she is black, belies the idea that these plans could be considered “benign” under Justice Stevens’s analysis. In this sense, Senator Figures’s plight today is no different from that of Dr. Gomillion. Like Dr. Gomillion, she has no means to be heard and no avenue for relief — except through this court. In light of these considerations, it is clear that, under Justice Stevens’ opinion in Vera, Alabama’s plans are not saved from the court’s “typically fatal skepticism.” Vera, 517 U.S. at 1010, 116 S.Ct. 1941.
Thus, under any of the analyses articulated in Vera, the racial quotas here, supposedly required by § 5, were the predominant factor motivating how the majority-black districts were drawn. Under any of those analyses, this plan is subject to strict scrutiny. For the plurality, strict scrutiny is required because the drafters adopted *1328racial quotas, achieved those quotas, and there was no other factor to explain why they added so many black people to the majority-black districts. For Justices Kennedy, Thomas, and Scalia, the adoption of a racial quota is enough standing alone. And for Justice Stevens and the other dissenters, the factors which would allow for an exception to the rule of strict scrutiny for racial classifications are simply not present in this case. Under the analyses announced in each of the opinions in Vera, the use of quotas in this case cannot stand unless they survive strict scrutiny,
iii.
This conclusion that the strictest scrutiny should apply in this ease because of the use of racial quotas is reinforced by an examination of United Jewish Organizations of Williamsburgh, Inc. v. Carey, 430 U.S. 144, 97 S.Ct. 996, 51 L.Ed.2d 229 (1977) (“UJO”). UJO was a predecessor to the Shaw line of cases. In UJO, a “highly fractured” majority, Shaw, 509 U.S. at 651, 113 S.Ct. 2816, upheld New York’s reapportionment plan against a constitutional challenge. The Court, first in Shaw and later in Miller, read the UJO majority to have decided the case on a vote-dilution theory, rather than a racial-gerrymandering theory. The Court was clear in Miller: “To the extent any of the opinions in [UJO ] can be interpreted as suggesting that a State’s assignment of voters on the basis of race would be subject to anything but our strictest scrutiny, those views ought not be deemed controlling.” Miller, 515 U.S. at 915, 115 S.Ct. 2475. Thus the fractured UJO majority’s views are not relevant to this case, as the Court has since read them as either inapposite or overruled.
Nonetheless, the UJO dissent is instructive. Of all the opinions in the case, only Chief Justice Burger’s dissent applied the kind of equal protection analysis eventually adopted by the Court in Shaw, namely one focused on the sorting of voters by their race. Therefore, while dissents ordinarily carry little authority, the dissent in UJO is different: it persuasively applied the analysis which the Court subsequently adopted as the law of the land in Shaw and Miller to a set of facts similar to those in the instant case. Thus, the UJO dissent carries substantial persuasive authority.
Chief Justice Burger perceived grave problems with New York’s plan, which mandated a particular minority population percentage, 65%, for majority-minority districts. The State believed that percentage was required by the VRA based on a comment from a Justice Department official. UJO, 430 U.S. at 181-2, 97 S.Ct. 996 (Burger, C.J., dissenting). Chief Justice Burger concluded that the State had “mechanically adhered” to the 65% figure, even rejecting an alternative that would have reduced the minority percentage by a mere 1.6%. Id. at 183, 97 S.Ct. 996. The Chief Justice rejected this approach as unconstitutional: “Although reference to racial composition of a political unit may, under certain circumstances, serve as a starting point ... rigid adherence to quotas” like the one at issue in UJO violates the Constitution. Id. at 185-6, 97 S.Ct. 996 (internal quotation marks omitted).
The drafters in this case took essentially the same approach as the State in UJO. Apparently believing that § 5 required their actions, the drafters adopted particular minority percentages for each majority-black district, and mechanically adhered to those figures whenever it was possible to do so. In UJO, the figure was 65% across the board, while in this case each majority-black district had its own figure, based on the black population at the time of reapportionment. But the result is the same in either ease under Shaw: when redistricting is driven by “rigid adherence *1329to quotas,” UJO, 430 U.S. at 186, 97 S.Ct. 996, strict scrutiny applies.
iv.
The majority states that the drafters’ need to pursue certain racial percentages for the majority-black districts was not a “bright-line rule” and that it gave way where “necessary to achieve other objectives.” Ante at 1294.
I am not quite sure what the majority means by saying that there was no bright-line rule. If the majority means that significance should be drawn from the fact that the drafters did not succeed in securing the sought-after percentage of black residents in each and every majority-black district, I have no qualm in noting that significance. Perhaps, for those districts where the drafters fell short, factors other than race can explain resulting percentages, and I am willing to engage the majority in a determination of whether the plaintiffs should prevail as to those districts. With this dissent, I am not saying that the plaintiffs should prevail as to all the districts. What I am saying is two things: First, there must be an individual assessment for each district as whether race was a predominant factor. See Miller, 515 U.S. at 916, 115 S.Ct. 2475 (question is whether “race was the predominant factor motivating the legislature’s decision to place a significant number of voters within or without a particular district”). Second, the fact that the drafters failed to achieve their sought-aftér percentage in one district does not detract one iota from the fact that they did achieve it in another. The racial quota, and nothing else, explains why SD 26 is 75% black. And the same ’is true for the fact that SD 24 and SD 23 are 63% and 64% black, respectively, and that HD 55 and HD 67 are 73% and 69% black, respectively, and so on. If the drafters relied on a racial quota in drawing even one district, that decision is subject to strict scrutiny.
In any event, what is most striking is the extent to which the drafters did succeed in matching the black percentage of the majority-black districts: the black percentage of the population in 13 House Districts23 and three Senate Districts24 is within one percentage point of the stated goal; in other words, the drafters effectively hit their quotas in those districts. Seven House Districts25 and three Senate Districts26 have an even higher percentage of black residents than under the old plan.27 Overall, the drafters either effec*1330tively achieved or surpassed their quotas in 75% of the majority-black districts.
Moreover, the majority points to no evidence that the drafters’ quotas ever actually did give way to any “other objectives.” Ante at 1294. While the percentage was lowered in some districts,28 the record contains essentially no evidence to explain why. In fact, the only objective Hinaman ever cited to explain lowering .the black percentage of a majority-black district was the creation of another majority-black district near HD 19, namely the displaced HD 53. Tr. Vol. III at 161-2. Maintaining the same number of majority-minority districts was part of the drafters’ understanding of what § 5 required; thus this explanation cannot support the conclusion that factors other than race trumped the drafters’ quotas. Hinaman never testified that he lowered the black percentage in any district for any other reason.
In fact, based on this record, the most likely explanation for the lower black percentage in some districts is that there were simply not enough black people nearby to maintain the already high black population percentages in some districts. It appears in some cases even extreme racial gerrymandering was not enough to find all the black people the drafters sought. But the fact that the drafters ultimately could not find enough black people to fill their quotas certainly does not mean that they did not try; and sorting people by race in the process of trying to achieve racial quotas is quite enough to trigger strict scrutiny.
Looking to where the drafters fell short is a distraction from the important point, which is where they succeeded. In most of the districts, the drafters of these plans either surpassed their quotas or effectively achieved them (to within a percentage point). In some cases, the precision with which the drafters refilled districts with the exact number of black individuals they sought is breathtaking. The most extreme example is HD 52. There, the quota was an additional 1,145 black people; the drafters added 1,143. See note 3, supra; APX 6. Out of a total population of 45,083, this represents racial sifting down to the finest level, a racial exactitude that would be admirable in its skill if it were not illegal.
In any event, if, with the observation that the drafters were not using a bright-line rule, the majority is suggesting that the drafters were pursuing ‘goals,’ or some synonym of that term, then “[tjhis semantic distinction is beside the point.” Bakke, 438 U.S. at 289, 98 S.Ct. 2733 (Opinion of Powell, J.). “Whether this limitation is described as a quota or a goal, it is a line drawn on the basis of race.” Id. In this case, the drafters have described setting a specific percentage of black population to achieve in each majority-black district. Thus, semantics aside, strict scrutiny applies.
v.
Even if the racial quotas, standing alone, were not enough to require strict scrutiny in this case, there is ample circumstantial evidence to establish that such scrutiny applies. That evidence shows that, time after time, the drafters subordinated various other districting factors to the goal of achieving their racial quotas.
Filling those quotas posed an enormous challenge to the drafters. In order to maintain the black percentage in the majority-black districts while repopulating the districts up to compliance with the 2% rule, the drafters needed to add over 120,-000 additional black people to the majori*1331ty-black House Districts. See note 3, supra. This amounted to 19.7% of total black population in the State not already living in a majority-black House District. See note 5, supra. When one considers that many of the black people in Alabama but not already living in a majority-black district were likely dispersed around the rest of the State, the chance of finding those 120,000 in areas contiguous to the majority-black districts is even smaller. The same is true in the Senate: the drafters needed to find over 106,000 additional black people in order to achieve their twin goals. See note 7, supra. That amounts to some 15.8% of the black population not already living in a majority-black Senate District. See note 8, supra.29
The challenge of meeting those quotas explains why the drafters drew these plans in the way they did; indeed, seeking to achieve the racial quotas drove everything. An examination of the steps the drafters took in seeking to maintain the previous black population percentages offers compelling circumstantial evidence that race predominated, further supporting the direct evidence already discussed.
For example, the racial quotas trumped the drafters’ stated goal of accommodating the preferences of incumbents. Dial rejected Keahey’s numerous suggestions, despite the fact that the legislators from nearby majority-black districts agreed to those suggestions. Of course, Dial was under no legal obligation to accept those suggestions; but his reason for doing so was Keahey’s failure to achieve the correct quota. Keahey had sought to avoid Dial’s idea of retrogression, but had mistakenly used the 2000 data instead of the 2010 data. The resulting discrepancy, that the black population in the nearby districts was lower than the quota by a percentage point or two, was enough to disqualify Keahey’s suggestions. Like New York’s rejection of a proposal which would have lowered the minority population just slightly in UJO, Dial’s rejection of Keahey’s proposals shows that the drafters rigidly adhered to their quotas.
Similarly, the quotas led Hinaman to abolish HD 53 and 73 in order to distribute their mostly-black populations among the surrounding majority-black districts. In those districts, the racial quotas trumped the stated goals of both maintaining the core of districts and avoiding conflicts between incumbents. Indeed, Hinaman testified that this latter priority was a “nice goal,” but one that “[djidn’t always work out.” Tr. Vol. Ill at 161; as the evidence establishes, that goal did not work out because'it came into conflict with achieving the racial quotas.
The quotas also led Hinaman to “reach [ ] out” to find majority-black precincts to add to majority-black districts. Tr. Vol. Ill p. 141-2. And, when precincts with enough black people were not available at hand, it led him to split “massive” numbers of precincts, Tr. Vol. II at 105, some 25% across the State, largely along racial lines.
Hinaman’s racial methodology in splitting precincts shows how far the drafters went to reach the target percentages of black people. Maptitude, the computer program he utilized, contained racial data at the census block level, but not political data. This means that when he split that *1332“massive” number of precincts, Tr. Vol. II at 105, he could not have done so based on how many Democrats or Republicans lived in each census block. Rather, it was racial data to which Hinaman looked in splitting precincts. And, indeed, Hinaman testified that he would split precincts in order to avoid what he considered retrogression. Tr. Vol. Ill p. 143. In addition, splitting a precinct by blocks required extra work, extra “clicking.” Tr. Vol. Ill p. 166. Each split was an affirmative choice, and the data on which Hinaman relied in making those choices were racial.
The Supreme Court has found this kind of evidence of racial methodology particularly compelling. In Vera, the plurality described a strikingly similar computer system to the one used here:
“REDAPPL permitted redistricters to manipulate district lines on computer maps, on which racial and other socioeconomic data were superimposed. At each change in configuration of the district lines being drafted, REDAPPL displayed updated racial composition statistics for the district as drawn. REDAPPL contained racial data at the block-by-block level, whereas other data, such as party registration and past voting statistics, were only available at the level of voter tabulation districts (which approximate election precincts).”
517 U.S. at 961, 116 S.Ct. 1941 (plurality opinion); see also Hinaman Dep., APX 75, at 123-4 (describing his use of racial data in Maptitude). The Vera plurality found that “the direct evidence of racial considerations, coupled with the fact that the computer program used was significantly more sophisticated with respect to race than with respect to other demographic data, provides substantial evidence that it was race that led to the neglect of traditional districting criteria here.” 517 U.S. at 963, 116 S.Ct. 1941 (emphasis omitted). In particular, since only racial data were available at the sub-precinct level, evidence of split precincts along racial lines suggested that “racial criteria predominated.” Id. at 970-71, 116 S.Ct. 1941. The same is true here. As in Vera, Hinaman’s race-based methodology is powerful evidence that race predominanted, particularly in combination with the direct evidence of racial quotas.
The majority in this case concludes that “at least some of the precinct splits” were attributable to the 2% rule. Ante at 1300. I agree this is probably true; Hinaman cited population deviation as the other reason to split precincts, along with compliance with the VRA. But the evidence shows that many if not most of the splits were made based on racial data. Cooper testified that, “If the only concerns were maintaining 27 majority black districts and achieving a plus or minus 1 percent deviation, you wouldn’t need to split anywhere near that many precincts.” Tr. Vol. II at 105. And Arrington noted that, as in SD 26, the splits were mostly along racial lines statewide; if Hinaman were primarily splitting precincts to equalize population, there is no reason he would need to separate black residents from white ones in this way. The plaintiffs certainly do not need to show that every precinct split was racially motivated to establish that the drafters went to great lengths to achieve their racial quotas. The circumstantial evidence that Hinaman relied on the race of voters in deciding how to split many precincts, along with the other circumstantial evidence and the direct evidence of racial quotas, amply establishes that race was the predominant factor.
vi.
The majority finds that race cannot have been predominant because there is a factor, namely the 2% rule, that was not subordinated to race. The majority also *1333points out that the drafters considered other factors as well. While I readily concede that the drafters abided by the 2% rule, and that they considered other factors, I must respectfully disagree that this allows their use of racial quotas to escape strict scrutiny.
The fact that a Shaw claim is a “mixed motive suit” does not mean that no racial gerrymander exists. Vera, 517 U.S. at 959, 116 S.Ct. 1941. On the contrary, in Vera the plurality, after noting, as the majority does here, that “The record does not reflect a history of purely race-based districting revisions,” Vera, 517 U.S. at 959, 116 S.Ct. 1941 (emphasis and internal quotation marks omitted), went on to strike down that plan under Shaw. The question there, as here, was whether race predominated over other factors as to any individual districting decision.
But in considering that question, the majority misapprehends the appropriate analysis. It appears the majority believes that race cannot predominate as long as there is some factor which is not subordinated to race. But this is wrong. The fact that the drafters pursued “multiple objectives,” Vera, 517 U.S. at 972, 116 S.Ct. 1941, does not preclude a finding of racial gerrymandering; again, that was the case in Vera, and the plan in that case was struck down. The existence of some factor which is not subordinated to race cannot defeat a Shaw claim.
For example, contiguity of a district is a traditional districting factor; the Miller Court cited it as a factor that, if subordinated to race, could establish that race predominated. 515 U.S. at 916, 115 S.Ct. 2475. Does that mean that contiguity must always be subordinated to race in order to prevail on a Shaw claim? On the majority’s view, it would appear so: unless a district was non-contiguous, for example split into two unconnected sections on different sides of the State, then race would not predominate. But, of course, that is not the law; for example, in Miller the Court struck down a district despite the fact that every part of it was connected to every other part. See id. at Appendix B.
The majority views the question of race predominating as a sort of ranking of factors as to the overall plan: since the 2% deviation rule is above the racial quotas in the drafters’ hierarchy overall, no amount of sorting people by the color of their skin can trigger strict scrutiny. In other words, the majority believes that once some race-neutral factor is established as the highest priority for the plan as a whole, that means that no Shaw claim can succeed as to any part of that plan. But this is not the Supreme Court’s analysis.
Instead, the Supreme Court has established that the harm of racial gerrymandering is a local one: the court must scrutinize each and every individual district to see whether race was the predominant factor. In Vera, for example, the plaintiffs initially challenged 24 of Texas’ 30 congressional districts; the district court found Shaw violations in three of those districts, and the Supreme Court upheld that finding as to those districts. 517 U.S. at 957, 116 S.Ct. 1941 (plurality opinion). The analysis was not what factors were predominant as to the plan as a whole, or even as to all 24 challenged districts considered together, but whether race was the predominant factor as to any one district individually. ■
Furthermore, a plaintiff need not even show that race was the predominant factor as to an entire district. In Miller, the Court stated that the plaintiffs burden in a Shaw case was to show “that race was the predominant factor motivating the legislature’s decision to place a significant number of voters within or without a particular district.” 515 U.S. at 916, 115 S.Ct. *13342475 (emphasis added). The plaintiffs have made just this showing, establishing that racial quotas led the drafters to place very significant numbers of people in the majority-black districts because they were black.
From this perspective, it is clear the 2% rule cannot explain why all these districts were drawn as they were. The drafters’ quotas for SD 26 called for that district to have 72.75% black population after reapportionment; the district is over 75% black under the new plan. How does the 2% rule explain why black people ended up on one side of the district line and white people ended up on the other? How can it explain why just 36 out of 15,785 new residents of SD 26 were white, despite the racially mixed demographics of the areas from which those people were drawn? The answer is clear: it does not.
In fact, it is clear that one factor and one factor alone explains why SD 26 is 75% black: race. The drafters had a quota for that district, which they believed was required under § 5, and they reached and exceeded that quota. Nothing else explains that percentage. The same is true of SD 23, with a quota of 64.79% black and an eventual population of 64.81% black. And the same is true of HD 55, with a quota of 73.54% black and an eventual population of 73.6% black. And the same is true of HD 67, with a quota of 69.14% black and an eventual population of 69.2% black. The 2% deviation rule simply does not explain away this clear reliance on, and achievement of, racial quotas.
But the Supreme Court’s cases establish that, when confronted with compelling evidence of this sort that district lines were motivated by race, a State seeking to avoid strict scrutiny must show that another factor explains away the apparent reliance on race. That is, the Supreme Court’s cases establish that a State may seek to show that “correlations between racial demographics and district lines may be explicable in terms of nonracial motivations.” Vera, 517 U.S. at 964, 116 S.Ct. 1941.
In Vera, the alternative the State offered was incumbency protection. The State argued that the direct and circumstantial evidence that race predominated was rebutted because another factor, protection of incumbents, actually explained the apparently racial divisions of voters. The plurality rejected that argument on the facts, but acknowledged that such a showing would undermine the ease for strict scrutiny. Id. at 964-5, 116 S.Ct. 1941.
Similarly, in Easley v. Cromarbie, the Court considered the argument that an apparent racial gerrymander was actually better explained as a partisan gerrymander. 532 U.S. 234, 257, 121 S.Ct. 1452, 149 L.Ed.2d 430 (2001) (“The basic question is whether the legislature drew District 12’s boundaries because of race rather than because of political behavior”) (emphasis in original). The Court reversed the district court and found the evidence in that case insufficient to establish that the apparently racial district boundaries were not in reality motivated by another factor. Id.
But the majority does not contend that the 2% deviation rather than the drafters’ goal of achieving racial quotas can explain the racialized boundaries of the majority-black districts. Nor could it, for there is no evidence to support that contention. Thus, the majority’s observation that the 2% rule never gave way to race is beside the point. The plaintiffs have come forward with compelling direct and circumstantial evidence that the drafters of these plans relied on a system of racial quotas to determine who would be added to the majority-black districts and who would not. The State’s adherence to the 2% rule simply does not rebut that evidence.
*1335Indeed, by and large the 2% rule served to increase the impact of the drafters’ racial quotas. While most of the majority-black districts were under-populated even using a more traditional 10% deviation rule, the 2% rule dramatically increased the number of additional black residents the drafters needed to find in order to achieve the quotas. This led to the sorting of individuals by race on a vast scale across the State in order to achieve racial quotas. Far from absolving the State of its liability under Shaw, it appears that in this case the 2% rule further aggravated the constitutional harm.30
Thus, there is no legitimate basis for rejecting the conclusion that race predominated in this case. The State did consider other factors, but the evidence is clear: race was the predominant factor in drawing the majority-black districts. Incumbency protection was a factor; but when Hinaman determined that he needed additional black residents for the under-populated districts in Montgomery and Jefferson Counties, he abolished HD 53 and HD 73, leaving their incumbents in another legislator’s district. Preserving the core of districts was a factor, but again one that gave way to race, in the cases of HD 53 and HD 73, which were abolished and redrawn elsewhere.. Respecting political subdivisions was a factor; but, in order to sift the black people from the white, Hinaman split massive numbers of precincts, depositing their black residents in the already heavily-black districts and their white residents in the adjoining majority-white districts.31 Compactness was a factor; but when Hinaman made up for SD 26’s under-population with new residents that were overwhelmingly black (and 99.8% minority), he did so by creating a bizarre district that wraps around the white portipns . of Montgomery.32 Honoring the wishes of incumbents was .a factor; but, as with Keahey’s nearly ten proposed alternative maps, those wishes were ignored if they came into conflict with the drafters’ rigid quotas. Preserving communities of interest was apparently a factor; but ultimately the boundaries of the . majority-black districts were predominantly drawn in order to achieve the racial quotas for each _ district. These plans were a racial gerrymander.
*1336C. Narrow Tailoring
Such a finding does not, of course, end the analysis. The State may save these plans by showing that they are narrowly tailored to achieve a compelling government interest. Miller, 515 U.S. at 904, 115 S.Ct. 2475. The majority concludes that compliance with the VRA is a compelling state interest, and I agree.
The Supreme Court has made clear, though, that to qualify as narrowly tailored, the district as drawn must be “required by a correct reading of § 5.” Shaw II, 517 U.S. at 911, 116 S.Ct. 1894 (emphasis added); see also Miller, 515 U.S. at 921, 115 S.Ct. 2475 (district must be “required by the substantive provisions of the Act”); ante, at 1306 (“we conclude that a plan will be narrowly tailored to achieve that interest when the race-based action taken was reasonably necessary under a constitutional reading and application of the Act”). And the legislature must have had a “strong basis in evidence” that its action was “needed in order not to violate” the VRA. Shaw II, 517 U.S. at 915, 116 S.Ct. 1894. As I will explain, these plans must fall because they are not required by any correct reading of § 5; because the drafters had no strong basis in evidence "to believe they were required by ,§ 5; and because in any event § 5 can no longer justify a racial gerrymander after Shelby County. ' .
, i.
The State has made a number of arguments about why its racial quotas were narrowly tailored to achieve the compelling purpose of compliance with § 5. Those arguments are all without merit.
The drafters of the proposed plans have all described their understanding of what was necessary to obtain preclearance in the same terms: they needed to maintain the same overall number of majority-minority districts and, within those districts, they needed to get as close as possible to maintaining the black percentage of the population calculated with the 2010 census data imposed on the 2001 redistricting plan. As I have explained, this amounted to imposing a racial quota on each such district.
All of the drafters expressed concern that doing less might expose them to denial of preclearance by the Justice Department. See Tr. Vol. Ill at 145 (Hinaman believed that “if I was significantly below [those percentages], I was concerned about that being retrogression that would be looked upon unfavorably by the Justice Department under Section 5”); Tr. Vol. I at 42 (Dial believed “our job was to get a plan ... that would meet Justice”); Tr. Vol. Ill at 220-1 (McClendon’s goal was Justice Department approval, and he was not aware of any hard numbers in terms of percentages that would be “okay”). The State has argued that this understanding was “not unreasonable.” Dfs. Proposed Findings of Fact and Conclusions of Law (Doc. No. 196), at 85. The State argues that erring on the side of caution is appropriate, particularly because the Justice Department review process is so “opaque.” Id. at 30; see also Tr. Vol. I at 12 (The State, in opening statement, noting that “To the extent the Department of Justice says anything, it’s pretty well general. Not too many African Americans in a district, not too few, but there’s no specifics.”).
Whether the State’s understanding was unreasonable is not the appropriate question under Miller and Shaw II. Nor is the question whether the Justice Department would approve or “look favorably” on the plans, or whether the drafters could accurately predict how the Justice Department would proceed. In Miller, the Court rejected the idea that narrow tailoring is *1337satisfied by actions taken in order to obtain preclearanee as a practical matter. 515 U.S. at 921, 115 S.Ct. 2475 (“It is, therefore, safe to say that the congressional plan enacted in the end was required in order to obtain preclearance. It does not follow, however, that the plan was required by the substantive provisions of the Act”). In that case, the Justice Department had demanded that the State draw certain districts as part of its preclearance review; the Court found that this was not sufficient to establish that those districts were narrowly tailored. Miller, 515 U.S. at 922, 115 S.Ct. 2475 (“We do not accept the contention that the State has a compelling interest in complying with whatever preclearance mandates the Justice Department issues.”). Rather, the only way to survive strict scrutiny is to show the plans were actually required by the statute. Id. at 921, 115 S.Ct. 2475.
On this point, the State argues that, “Given the fact that the State’s plans have been precleared, the State’s reading of Section 5 cannot be said to be incorrect.” Dfs. Proposed Findings of Fact and Conclusions of Law (Doc. No. 196), at 83; see also id. at 85 (arguing the drafters’ understanding was not “demonstrably incorrect”). This, again, is wrong. First, under strict scrutiny it is the State’s burden to establish that its action was required under a correct reading of the statute, not the plaintiffs’ burden to show the drafters’ understanding was demonstrably incorrect. Miller, 515 U.S. at 920, 115 S.Ct. 2475; Fisher, 133 S.Ct. at 2419 (“Strict scrutiny is a searching examination, and it is the government that bears the burden
Second, the fact that the Justice Department precleared the plans does not determine one way or the other whether the State’s actions were actually mandated by the substantive statute. This would be so even if the drafters had correctly interpreted the Justice Department’s commands. “Where a State relies on the Department’s determination that race-based districting is necessary to comply with the Act, the judiciary retains an independent obligation in adjudicating consequent equal protection challenges to ensure that the State’s actions are narrowly tailored to achieve a compelling interest.” Miller, 515 U.S. at 922, 115 S.Ct. 2475. Here, however, the Justice Department never commanded the State to adopt its quotas; the drafters merely inferred, or believed, or guessed that such a step would smooth the preclearance process. That is insufficient to establish that the drafters’ actions were narrowly tailored.
In reality, the drafters’ understanding of § 5 was woefully incorrect, and as a result their' solution is not narrowly tailored. Nothing in § 5, or in the cases interpreting it, required the State to adopt and adhere to these quotas. In Beer v. United States, the Supreme Court noted that “the purpose of [§ ] 5 has always been to insure 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.” 425 U.S. 130, 141, 96 S.Ct. 1357, 47 L.Ed.2d 629 (1976). Thus, § 5 as properly interpreted requires a State to determine whether an action would reducé ' minority voters’ effective ability to elect candidates of choice; it does not command the State to match the preexisting level of minority population.
The State relies on Texas v. United States, a recent three-judge-court § 5 case, as establishing that “‘A district with a minority voting majority of sixty-five percent (or more) essentially guarantees that, despite changes in voter turnout, registration, and other factors that affect participation at the polls, a cohesive minority *1338group will be able to elect its candidate of choice.’ ” Dfs. Proposed Findings of Fact and Conclusions of Law (Doc. No. 196) at 86 (quoting Texas v. United States, 831 F.Supp.2d 244, 263 & n. 22 (D.D.C.2011)). In the State’s view, Texas establishes that the State’s decision to add black people to majority-black districts as it did was required under § 5. The State is incorrect.
In the relevant portion of its opinion on summary judgment, the Texas court established that a majority-minority population of 65% percent “essentially guarantee^]” ability to elect in that ease. 831 F.Supp.2d at 263. Texas was a § 5 case, in which the issue was whether certain districts the State had drawn violated § 5 by retrogressing minority voting power. In establishing its per-se 65% rule, the court was making an evidentiary ruling: when it examined whether a given district the legislature had drawn was likely to elect a candidate of minority voters’ choice, a 65% minority population was sufficient evidence standing alone. This amounts to a common-sense observation: at some point a State may put so many minorities in a district that ‘the numbers speak for themselves’ when it comes to the ability of that minority group to win elections in the district.
In this case, the question is not whether certain districts violated § 5 (for example by containing a minority population that is too low), but whether § 5 required the drafters to adopt the quotas as they did. Therefore, the court’s observation in Texas that 65% minority populations are essentially guaranteed to be able to elect candidates of choice is not relevant here; the same is true, of course, of 75%, or 85%, or 100% minority districts. That tells one nothing about whether § 5 requires such high percentages. Thus, in sum, the State offers no reason to believe that its racial quotas were actually required by § 5.
ii.
The majority agrees with the State that these plans were justified by § 5. But, while the majority’s interpretation of § 5 is different from the State’s, it is no less mistaken. In the majority’s view, the drafters’ conduct was narrowly tailored because the 2006 amendments to the VRA altered the standard for assessing retrogression. In those amendments, Congress expressly noted its intention to overturn the Supreme Court’s decision in Georgia v. Ashcroft, 539 U.S. 461, 123 S.Ct. 2498, 156 L.Ed.2d 428 (2003). See Fannie Lou Hamer, Rosa Parks, and Coretta Scott King Voting Rights Act Reauthorization and Amendments Act, Pub.L. No. 109-246, § 2, 120 Stat. 577, § 2(b)(6) (2006). The majority concludes that the amendments mean that “any diminishment in a minority’s ability to elect its preferred candidates violates section 5.” Ante at 1309. The majority further concludes that this, in turn, required the State, “where feasible,” to “not substantially reduce the relative percentages of black voters” in majority-black districts. Ante at 1310. In other words, as the majority reads the amended statute, it required the drafters to do just what they did: adopt the previous black percentages as racial quotas for each district.33
*1339The interpretation of the amended § 5 which the majority adopts is the wrong one. In the majority’s view, the 2006 amendments mean that any reduction in the black percentage of a majority-black district is per-se retrogressive (except where that reduction is unavoidable). The problem is that this interpretation is contrary to the intent of Congress; has been rejected by both entities primarily responsible for administering § 5; and would create serious, if not fatal, constitutional concerns.
In order to explain why the majority’s reading is wrong, I must first explain how the majority arrives at its conclusion, and where we part ways. The majority first finds that the 2006 amendments altered the retrogression analysis under § 5 to make it more stringent, and I agree. The majority also concludes that the amendment to the language of § 5 served, in relevant part, to overturn the Supreme Court’s decision in Georgia, and to restore the standard articulated in Beer, 425 U.S. at 141, 96 S.Ct. 1357 (“the purpose of [§ ]5 has always been to insure 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”). Again I agree.
The Georgia decision introduced a “totality of the circumstances” approach to determining whether a change would be retrogressive under § 5. The Court found that the ability of a minority group to elect a candidate of choice was important, but was not the only relevant factor. In addition, the Court held, retrogression analysis must take account of the minority group’s ability to participate in the political process. In particular, the Court found that “influence” districts, in which the minority group cannot elect a candidate of choice but can “play a substantial ... role in the electoral process,” could compensate for a reduction in the number of districts in which minorities could elect candidates of choice. 539 U.S. at 482, 123 S.Ct. 2498. Also, the Court found that representatives of the minority group holding positions of “legislative leadership, influence, and power” was a factor suggesting that a new plan was not retrogressive. Id. at 483, 123 S.Ct. 2498. Because the Court determined that the district court had focused too narrowly on ability to elect, it remanded the case for analysis under the totality of the circumstances test. Id. at 485, 123 S.Ct. 2498.
The majority finds that, in rejecting Georgia, Congress commanded that Alabama could not reduce “the percentages of black voters in the majority-black districts because to do so would be to diminish black voters’ ability to elect their preferred candidates.” Ante at 1309. That is, the majority believes that, after the 2006 amendments, any reduction in a minority group’s percentage of the population in a given majority-minority district reduces *1340the ability to elect, and is per-se retrogressive. I will explain why this is incorrect.
First, though, I pause to observe just how implausible this reading of the statute is. On the majority’s view, if a district is 99% black, the legislature is prohibited by federal law from reducing the black population to a mere 98%. Read in this way, § 5 would become a one-way ratchet: the black population of a district could go up, either through demographic shifts or redistricting plans (like this one) that raise the percentage of black people in some majority-black districts. But the legislature could never lower the black percentage, at least so long as it was “feasible” to avoid it. Ante at 1310-11. Why? Because any reduction in the black population of a district would “by definition ... diminish black voters’ ability to elect their preferred candidates.” Id. at 1309. With respect, this result cannot be.
It is also not what Congress intended. As amended, § 5 provides in relevant part that a voting change is prohibited if it “will have the effect of diminishing the ability of any citizens of the United States on account of race or color ... to elect their preferred candidates of choice.” 42 U.S.C. § 1973c(b). Congress specified that the purpose of the above-quoted language “is to protect the ability of such citizens to elect their preferred candidates of choice.” 42 U.S.C. § 1973e(c). It is clear from this language that “ ‘ability to elect’ is the statutory watchword.” Texas, 831 F.Supp.2d at 260.
The congressional findings, and the legislative history, make it clear that the goal of this new language was to overturn Georgia. In the majority’s view, this change means that now any reduction in a minority group’s proportional share of the population in a district is retrogressive. The better reading of Congress’ intent is that, in emphasizing “ability to elect,” Congress sought only to overturn the aspect of Georgia that so many found disturbing: namely the prospect that States would trade away districts where minority voters had actual ability to elect in exchange for amorphous influence districts or apparently politically powerful jobs for minority representatives. The House Committee Report described the problem in this way:
“Under its ‘new' analysis, the Supreme Court would allow the minority community’s own choice of preferred candidates to be trumped by political deals struck by State legislators purporting to give ‘influence’ to the minority community while removing that community’s ability to elect candidates. Permitting these trade-offs is inconsistent with the original and current purpose of Section 5.”
H.R.Rep. No. 109-478, at 69. Congress rejected this idea, endorsing instead the position of the dissent in Georgia. See id. at 68 n. 183 (“The dissent in [the] Georgia v. Ashcroft case correctly pointed out that a ‘totality of the circumstances’ under Section 5 is hopelessly unadministrable by the Department of Justice because such a concept does not retain ‘the anchoring reference to electing a candidate of choice.’ ”) (quoting Georgia, 539 U.S. at 493, 123 S.Ct. 2498 (Souter, J., dissenting)). Rather than the extreme interpretation embraced by the majority in this case, it is clear that what Congress intended when it sought to overturn Georgia was to legislatively adopt the position of Justice Souter’s dissent in Georgia.
But Justice Souter’s dissent did not interpret § 5 in the way the majority does in this case. On the contrary, Justice Souter agreed with the majority in Georgia that reducing the percentage of black population in a majority-black district would not necessarily be retrogressive. “The District Court began with the acknowledgment (to which we would all assent) that *1341the simple fact of a decrease in black voting age population (BVAP) in some districts is not alone dispositive about whether a proposed plan is retrogressive.” Georgia, 539 U.S. at 498, 123 S.Ct. 2498 (Souter, J., dissenting) (emphasis added); see also id. at 504, 123 S.Ct. 2498 (Souter, J., dissenting) (“nonretrogression does not necessarily require maintenance of existing supermajority minority districts”).
Justice Souter’s view on this issue was hardly lost on Congress. Most of the debate surrounding the changes to the retrogression standard focused on whether or not “coalition” districts (in which a minority group does not constitute a majority but can routinely elect candidates of choice in coalition with other racial groups) could constitute “ability to elect” districts for § 5 purposes. That question is not presented in this case. The question that is presented here — whether a minority percentage that is lower than the benchmark34 plan is always retrogressive — was not widely debated. But the two discussions of it in the legislative history firmly reject the majority’s view.
Representative Watt, a leading proponent of the bill in the House and chair of the Congressional Black Caucus at the time, specifically noted and endorsed the Georgia Court’s unanimous position on this issue during a key hearing on the effect of Georgia on the retrogression standard:
“Nine justices agreed, as do I, that section 5 does not prohibit the reduction of super majority minority voting age population percentages from that in a benchmark plan. Where the majority in Georgia v. Ashcroft strayed, however, losing four justices in the process, was in its failure to enunciate an articulable standard under which the opportunities to elect are preserved.”
Voting Rights Act: The Judicial Evolution of the Retrogression Standard: Hearing Before the Subcomm. On the alteration Constitution of the H. Comm. On the Judiciary, 109th Cong. 109-74 (2005) at 5 (emphasis added).
The principal sponsors of the amendments in the Senate agreed. During floor debate, some senators had suggested that coalition districts would not be protected by the retrogression standard. Senator Leahy responded by entering into the record a statement reflecting “my understanding of the purpose and scope of [the relevant] provisions as an original and lead sponsor.” 152 Cong. Rec. 96, S8004 (2006). That statement provided:
“This change to Section 5 makes clear that Congress rejects the Supreme Court’s Ashcroft decision and reestablishes that a covered state’s redistricting plan cannot eliminate ‘ability to elect’ districts and replace them with ‘influence districts’ ... The amendment to Section 5 does not, however, freeze into place the current minority voter percentages in any given district. As stated by the dissenters in Georgia v. Ashcroft, as well as by Professor Arrington and Professor Persily at the Committee hearings, reducing the number of minorities in a district is perfectly consistent with the pre-Ashcroft understanding of Section 5 as long as other factors demonstrate that minorities retain their ability to elect their preferred candidates.”
Id. (emphasis added); see also id. at S8009 (Sen. Feingold, “[a]s ranking member of the Judiciary Committee’s Subcommittee on the Constitution, Civil Rights, and *1342Property Rights,” concurring with Sen. Leahy’s understanding).
Equally striking is the fact that no one contested this understanding. While there was a concerted effort by some in the Senate to establish that the retrogression standard would not lock in coalition districts, no one ever suggested that Congress was adopting the novel and implausible standard the majority posits in this case. Indeed, there is nothing in the text, nothing in the legislative history, and nothing in the dissent in Georgia which would support the majority’s view.35
That view has also been rejected by the U.S. District Court for the District of Columbia, which is entrusted with the primary responsibility for enforcing § 5. The D.C. District Court’s most extensive application of § 5 after the 2006 amendments came in the Texas case. In the opinion after trial in that case, the three-judge court rejected the idea that lowering the minority percentage of a supermajority district is per-se retrogressive. In considering the changes to Texas’ House District 41, the court noted that the Hispanic citizen voting-age population had been reduced from 77.5% in the benchmark plan to 72.1% in the new plan. Texas v. United States, 887 F.Supp.2d 133, 169 (D.D.C.2012), vacated and remanded on other grounds, — U.S. -, 133 S.Ct. 2885, 186 L.Ed.2d 930 (2013). Under the test adopted 'by the majority in this case, that information by itself would establish retrogression. But the Texas court rejected a claim that this change was retrogressive, finding that even with a lower percentage of the population, Hispanic voters still had the ability to elect candidates of choice. Id.
Instead of the majority’s test, which looks solely to whether a minority group’s percentage of the population is lower than it had been under the benchmark plan, the Texas court adopted a “functional” approach. Rejecting the State’s argument that the court should look only to population demographics, the court found that it was necessary to examine a number of factors to determine whether a minority group has the ability to elect candidates of choice. “A single-factor inquiry, such as the test Texas proposed relying on racial and ethnic population statistics alone, is inconsistent with precedent and too limited to provide an accurate picture of the on-the-ground realities of voting power.” Id. at 140. Rather, the court established at summary judgment that “Section 5 analysis must go beyond mere population data to include factors such as minority voter registration, minority voter turnout, election history, and minority/majority voting behaviors.” Texas, 831 F.Supp.2d at 263.36
This is substantially the same interpretation of the amended § 5 as that adopted by the Justice Department, the other primary adjudicator of preclearance. In its *1343updated guidance, released in 2011, the Justice Department, like the D.C. District Court, applies a functional, multi-factor test. See Guidance Concerning Redistricting Under Section 5 of the Voting Rights Act, 76 Fed.Reg. 27, 7471 (Feb. 9, 2011). As the Justice Department interprets § 5, the analysis of retrogressive effect “starts with a basic comparison of the benchmark and proposed plans at issue, using updated census data in each.” Id. (emphasis added). But it does not end there:
“In determining whether the ability to elect exists in the benchmark plan and whether it continues in the proposed plan, the Attorney General does not rely on any predetermined or fixed demographic percentages at any point in the assessment. Rather, in the Department’s view, this determination requires a functional analysis of the electoral behavior within the particular jurisdiction or election district. As noted above, census data alone may not provide sufficient indicia of electoral behavior to make the requisite determination.”
Id. In other words, both the D.C. District Court and the Justice Department have explicitly rejected the majority’s interpretation.
And with good cause. The majority’s interpretation of the amended § 5 would raise serious, if not fatal, constitutional concerns. There is an inherent tension between the race consciousness of the VRA, and in particular § 5, and the protections of the Fourteenth Amendment. See Vera, 517 U.S. at 995, 116 S.Ct. 1941 (O’Conner, J., concurring) (“The VRA requires the States and the courts to take action to remedy the reality of racial inequality in our political system, sometimes necessitating race-based action, while the Fourteenth Amendment requires us to look with suspicion on the excessive use of racial considerations by the government”).
Yet the majority urges an interpretation of § 5 that would require States to engage in hugely racialized redistricting; indeed, an interpretation that would require States to redistrict in compliance with racial quotas. Under the majority’s rule, a State faced with a 90% minority district has no choice: it must find nine minority individuals for every 10 needed to repopulate the district. Racial gerrymandering would become unavoidable, essentially required by a federal statute. “When [an] interpretation of the [VRA] compels race-based districting, it by definition raises a serious constitutional question.” Miller, 515 U.S. at 923, 115 S.Ct. 2475.
UJO, discussed above, places these constitutional questions in stark contrast. Chief Justice Burger’s dissent, which applied the Shaw reasoning later adopted by the Court, rejected the defendants’ rigid adherence to a specific minority percentage, 65%, in seeking to comply with § 5. He observed that there was “no indication whatever that use of this rigid figure was in any way related much less necessary to fulfilling the State’s obligation under the Voting Rights Act as defined in Beer.” 430 U.S. at 183, 97 S.Ct. 996. Rather, he would have found this unjustified “rigid adherence to quotas” unconstitutional. Id. at 185-6, 97 S.Ct. 996 (internal quotation marks omitted). The interpretation the majority adopts is no less rigid; it too equates ability to elect with a certain predetermined percentage of the population. It raises the same constitutional questions that Chief Justice Burger identified.
But facing those constitutional questions is simply unnecessary. Congress did not seek to impose racial quotas on States, nor permanently to freeze in place minority supermajorities, long after minority groups’ need for those supermajorities in *1344order to elect candidates of choice has passed. The purpose of the VRA is to help minority groups achieve equality, not to lock them into legislative ghettos. Congress intended no such thing. The majority’s interpretation of the amended § 5 is in error.
iii.
Applying instead the functional test articulated in Texas, I think it is clear that even substantial reductions in the black percentage of many of the majority-black districts would be permissible under § 5. As such, in seeking out so many black people to satisfy their unjustified racial quotas, the drafters “went beyond what was reasonably necessary to avoid retrogression.” Vera, 517 U.S. at 983, 116 S.Ct. 1941 (plurality opinion).
The Texas court’s functional analysis requires the court to look-to a variety of factors, including the mobilization of the minority group in question. In Texas, the court was concerned that many of the relevant factors meant that the minority group at issue in that case, Latinos, would require substantially more than 50% of the population to effectively elect candidates of choice. Evidence and congressional findings of low Latino rates of registration and turnout “underscore^] why Texas’ reliance on a bare majority-minority district [could not] be used to determine an ability district under Section 5.” Texas, 831 F.Supp.2d at 264. That is, Texas held that, considering the particular circumstances of Latinos in Texas, § 5 required substantially more than 50% minority population in majority-minority districts.
In this case, there is significant evidence that, in light of much-improved black voter mobilization and near-universal citizenship, the black voting population in Alabama can elect candidates of their choice at significantly lower levels of population than the Texas court deemed necessary in that case. The evidence suggests that in Alabama black voters need to be only about 50% of a given district to be able to elect representatives of choice. See Arrington Report, NPX 323, at 17; Lichtman Report, NPX 324, at 21-2. If that is so, then even if the legislature substantially reduced the percentage of black residents of, for example, HD 55 (73% black), black residents would still have the ability to elect candidates of choice there. The point is not that the State was required to lower the black percentage of HD 55. Rather, it is that § 5 did not prohibit the State from doing so. And, that being the case, the State here cannot claim that the VRA required it to maintain HD 55 with 73% black people. Therefore, the drafters’ conduct was not narrowly tailored.
The majority has found that much of the evidence that black voters can elect candidates of choice with little more than a bare majority is not credible, and therefore concluded that the record can support no conclusion about the minimum level of black population necessary to allow black voters to elect candidates of choice. I disagree with those factual determinations; in particular, I can discern no legitimate basis in the record for the majority to find Arrington’s testimony not credible. Compare ante at 1302-03 (rejecting Arrington’s testimony) with Tr. Vol. III at 81-2; Arrington Report, NPX 323, at 19; id. at 17; Tr. Vol. III at 64-5 (Arrington giving reasonable and unrebutted explanations for supposed inconsistencies). I would credit Arrington’s testimony on this issue.
But, more importantly, even if one accepts the majority’s conclusion that the record supports no determination one way or the other regarding the level of black population necessary to elect candidates of choice, see ante at 1279, in the context of racial gerrymandering that conclusion can only harm the State’s case. The burden is *1345on it to establish that it had a “strong basis in evidence”, Miller, 515 U.S. at 922, 115 S.Ct. 2475, for the need for their purported solution, namely striving to fill racial quotas. If it has not shown a strong basis in evidence, because the record can support no conclusion one way or the other, then the racial gerrymander is unconstitutional.
The drafters’ failure to take any steps to examine what § 5 actually required in this case underscores that these plans are not narrowly tailored. Hinaman testified that he did not review any studies of black voter participation in Alabama, did not look at variations among black communities, and did not use the political data he had available to examine effectiveness of majority-black districts. Tr. Vol. Ill at 148-150.37 Dial testified that he did not inquire at all into what level of black population would be necessary to avoid retrogression in a given district:
“Q. So your testimony is that you really didn’t look into the behavior of individual districts. Instead, you simply went by the black — the number of black people, the black percentage in the district. And what you did was try and at least maintain that or increase it. Is that your — fair statement of your testimony?
“A. That’s fair, yes.”
Tr. Vol. I at 133-4. Had any of the drafters analyzed the available data, they might (or might not) have had a “strong basis in evidence,” Miller, 515 U.S. at 922, 115 S.Ct. 2475, to conclude that § 5 required them to maintain the high percentages of black population; as they did nothing of the sort, they had nothing but guesses. And that is not enough to justify the use of racial quotas in drawing legislative districts.
The question here is whether the State was required by the VRA to seek out black people to add to the already heavily black majority-minority districts in order to achieve their racial quotas. And the clear answer is no. There was an available alternative: not to sift individuals by race at all, or only to do so to the extent actually required by the VRA, and instead to use other districting principles to draw those lines. They could have been guided by protecting incumbents, following natural and political boundaries, keeping districts compact, etc. Instead, the drafters reached out and grabbed as many black people as possible to achieve their racial quotas even as the total population of those districts grew. The conclusion is as clear as day: the drafters’ action was not required under any correct reading of the statute, and so cannot survive as narrowly tailored.
iv.
Even if the drafters’, racial quotas were somehow required by § 5, that would not be enough to save these plans, because Alabama is no longer subject to the preclearance requirements of § 5. The Alabama Legislative Black Caucus plaintiffs argue that the State cannot now rely on § 5 to justify its racial gerrymander because of the Supreme Court’s intervening decision in Shelby County, 133 S.Ct. 2612, which was handed down after this case was filed but before trial. The majority responds that Shelby County struck down *1346only § 4 of the VRA, 42 U.S.C. § 1973b, the formula for determining whether a jurisdiction is covered by § 5, but left § 5 itself undisturbed. However, without § 4, and absent further action by Congress, Alabama is no longer a covered jurisdiction subject to § 5 and need not obtain preclearance. See Shelby County, 133 S.Ct. at 2632 n. 1 (Ginsburg, J., dissenting) (noting that “without th[e] formula, § 5 is immobilized”).38
The majority then concludes that, even if compliance with § 5 is not now a compelling interest, the State’s actions should be evaluated based on the circumstances when the plans were enacted, not those of the time of judgment. I disagree. These plans have not yet gone into effect, and “changed circumstances may ... transform a compelling interest into a less than compelling one.” United States v. Antoine, 318 F.3d 919, 921 (9th Cir.2003). Indeed, when it comes to racial classifications, the solution offered must last no longer than the compelling interest on which the State relies. See Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 238, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995) (plurality opinion) (part of narrow tailoring analysis is whether race-based solution “was appropriately limited such that it ‘will not last longer than the [problem] it is designed to [address]’ ”) (quoting Fullilove v. Klutznick, 448 U.S. 448, 513, 100 S.Ct. 2758, 65 L.Ed.2d 902 (1980) (Powell, J., concurring)). Here, relying on the fact that § 5 was still applicable at the time the drafters designed the plans, the State asks us to approve a race-based solution that has not only already outlived its problem, but also one that will be in effect into the next decade, through the 2020 census.39 But the question in strict-scrutiny analysis is not whether the drafters acted in “good faith” when they enacted these plans, see Fisher, 133 S.Ct. at 2421, nor in strict scrutiny do we grant the kind of deference to which States are often entitled in other areas of law. See id. at 2420. In the absence of an actual compelling interest at the time of judgment, the court cannot approve a racial gerrymander.
v.
There is perhaps one last unarticulated premise to confront. One might think that the plaintiffs here, who are mostly black legislators and voters, should lose on their Shaw claims because the majority-black districts were drawn for their benefit. The plaintiffs in Shaw and its progeny were, after all, white voters who objected to the creation of majority-minority districts. It may be thought that there is some incongruity to black voters bringing *1347the same charge against districts in which they are the majority.
The Supreme Court’s equal protection cases teach that it is sometimes difficult to discern when a race-conscious policy inures to the benefit of a minority group and when it covertly prejudices them. See Vera, 517 U.S. at 984, 116 S.Ct. 1941 (plurality opinion)(“we subject racial classifications to strict scrutiny precisely because that scrutiny is necessary to determine whether they are benign”); Parents Involved, 551 U.S. at 742, 127 S.Ct. 2738 (plurality opinion of Roberts, C.J.)(“History should teach greater humility” than to assume one can differentiate good intentions from bad)(internal quotation marks omitted). Indeed, as Justice Thomas recently observed, “The worst forms of racial discrimination in this Nation have always been accompanied by straight-faced representations that discrimination helped minorities.” Fisher, 133 S.Ct. at 2429 (Thomas, J., concurring).
In this case, there is a deep dispute regarding the legislative purpose behind these plans. According to the drafters, they sought nothing more than to comply with their legal duties and honor their colleagues’ wishes as far as that was possible. According to the plaintiffs, these redistricting plans are part of a scheme to eliminate all white Democrats in the State and thereby establish the Republican Party as the natural home for all white Alabamians, leaving the Democratic Party comprised of only black voters and legislators. In furtherance of that scheme, the plaintiffs claim, the drafters packed as many black people as possible into the majority-black districts, thereby eliminating their influence anywhere else. All this, the plaintiffs claim, was done under the pretext of seeking to comply with § 5, while in reality the drafters were motivated by invidious racial discrimination. Apparently for this reason, no black legislator voted in favor of these plans.
In my view, we need not resolve the question of the drafters’ ultimate purpose, nor need we reach the plaintiffs’ other claims. For, again, to me this case is simple. In drawing the majority-black districts, Hinaman and the others were driven by an overriding consideration: the race of those individuals who would be included in or excluded from those districts. They adopted racial quotas for each district, and they went to extraordinary lengths to achieve those quotas. Whether they did so in a good-faith belief that the quotas were required by § 5, or for some invidious purpose, is ultimately of no consequence for the Shaw claims. But that they did so is as clear as day. Because the State has offered no sufficient justification for the use of racial quotas, the plans are unconstitutional, and I would so hold.
There is a cruel irony to these cases. Earlier this year, the State of Alabama passionately argued to the Supreme Court that it should be free from the VRA requirements of preclearance. See Br. of State of Alabama as Amicus Curiae, Shelby County, available at 2013 WL 98691. The Court agreed, effectively removing the preclearance requirement for covered jurisdictions nationwide. Noting that “Our country has changed,” the Court found that Congress’s remedy for racial discrimination in voting failed to “speak [] to current conditions.” Shelby County, 133 S.Ct. at 2631.
The evidence here is overwhelming that the State has intentionally singled out individuals based on race and cabined them into district after district. The drafting of majority-black districts was driven by naked racial quotas; that alone is enough to condemn these plans. But Alabama *1348argues that these percentages were justified by, of all things, § 5. Even as it was asking the Supreme Court to strike down the requirement of preclearance for failure to speak to current conditions, the State of Alabama was relying on racial quotas with absolutely no evidence that they had anything to do with current conditions, and seeking to justify those quotas with the very provision it was helping to render inert.
To be sure, conditions 30 years ago or 20 years ago or even a decade ago (in or around 2001) may have justified requiring high percentages of black population in majority-black districts. Indeed, as I now consider Alabama’s and the majority’s argument that the record justifies these high racial percentages, I feel as if I were in a time warp carried back into the past, with the arguments being the same but with the parties having switched sides. But, again, the issue here is, What are the conditions today? Not, what they were back then.
As a nation, we must continue to strive towards “the goal of a political system in which race no longer matters.” Shaw, 509 U.S. at 657, 113 S.Ct. 2816. But plans like the ones the Alabama legislature has adopted take us in the wrong direction; they continue to “balkanize us into competing racial factions,” id., “carving [us] into racial blocs.” Miller, 515 U.S. at 927, 115 S.Ct. 2475. The problem is not that these plans consider race, for some consideration of race is permissible and even required by the VRA The problem is that these plans adopt severe racial quotas — seeking to match numbers as high as 78% black— with no evidence or even real argument that their extreme reliance on race is necessary. The Supreme Court’s jurisprudence of race condemns them.
Therefore, just as the Supreme Court, in applying principles of federalism, found in Shelby County that Congress’s remedy for racial discrimination in voting failed to “speak [] to current conditions,” Shelby County, 133 S.Ct. at 2631, this court, applying strict scrutiny, should likewise find that the Alabama Legislature’s racially based redistricting plans fail to speak to current conditions. And just as the Supreme Court sent Congress back to the drawing board, this court should send the Alabama Legislature back as well.40
Moreover, because these plans have not gone into effect, there is ample time for the Alabama Legislature to come up with plans that accede to the request made by all of Alabama’s black legislators, a request that is not only a legitimate and laudable one but is, in fact, the only legitimate request that can be made absent current conditions reflecting otherwise: to carry out its decennial reapportionment, as required by one-person-one vote, based more on traditional districting factors (such as respect for political subdivisions and precincts, compactness, contiguity, and incumbency) and based far less on race.
Fashioning and implementing such a remedy would not be difficult. Without a doubt, if, following the 2010 census, the Alabama Legislature had not used these naked racial quotas in redistricting for the House and Senate, the plans it would have adopted would not be the ones before us today. Therefore, my command to the State in redrawing the plans would be simple and direct: Get rid of racial percentages, that is, the naked racial quotas, *1349that the State incorrectly believed § 5 to require. In other words, while race and racial issues may be valid considerations, and may even be required under § 2 of the VRA, naked racial quotas (that is, racial line-drawing not rooted in and compelled by a sensitive assessment of current conditions) are unconstitutional.
Therefore, because the plans before this court rely on quotas to cabin individuals into districts based on the race of those individuals in an intentional, unjustified, and thus illegal manner, I cannot give the plans my imprimatur. I respectfully dissent.
. The majority appears to suggest that the 2% deviation rule was required by the one-person-one-vote principle, as applied in Larios v. Cox, 300 F.Supp.2d 1320 (N.D.Ga.2004) (three-judge court), aff'd by Cox v. Larios, 542 U.S. 947, 124 S.Ct. 2806, 159 L.Ed.2d 831 (2004). while I do not believe we must reach this issue, I feel obliged to set the record straight on this important question. Alabama's 2% rule is not constitutionally required; rather, it is well established that for state legislative redistricting any deviation up to 10% is presumptively constitutional. Brown v. Thomson, 462 U.S. 835, 852-53, 103 S.Ct. 2690, 77 L.Ed.2d 214 (1983). Therefore, the 2% rule is simply a state policy (not even a state statute or regulation) and must give way to the VRA when the two come into conflict. Thus, contrary to the majority’s view, the State's 2% rule cannot restrict a § 2 plaintiff's ability to present an alternative map with a greater population deviation. “Larios in no way mandates that plaintiffs in a § 2 case bear a greater burden than simply presenting a plan with a population deviation under 10%.” Georgia State Conference of NAACP v. Fayette Cnty. Bd. of Comm’rs, 950 F.Supp.2d 1294, 1311 (N.D.Ga.2013) (Batten, J.).
.To calculate the additional black population the drafters needed to add to each majority-black House District in order to comply with their rules, the total additional population needed by each majority-black district was multiplied by the black quota percentage used by the drafters for that district. The total additional population needed was calculated by subtracting the pre-reapportionment population, APX 6, from 99% of ideal population (45,065). Because some districts in this plan were 1% over-populated, 99% was used so *1316that the total deviation would be no more than 2%. The total number of additional residents needed to comply with the 2% rule for each district was then multiplied by the quota percentage, which is the percentage of black population in each district before apportionment based on 2010 census data. APX 6. The resulting number of additional black individuals required for each district was rounded to the lowest integer. The sum of this calculation for each majority-black House District is 120,825 additional black people necessary across all districts to meet the requirements of the drafters' rules.
. Mr. Hinaman looked to total black population, not voting-age black population, in implementing his understanding of non-retrogression. Tr. Vol. Ill at 118. I do likewise throughout this opinion.
. According to the 2010 census, the total population of Alabama that identifies as any part black is 1,281,118. 2010 Demographic Profile Data, NPX 325. There were 669,134 black individuals living in majority-black House Districts at the time of the 2010 census. APX 6. This leaves 611,984 black people not yet living in a majority-black House District. The 120,825 additional black people needed to achieve the drafters' quotas represented 19.7% of those not already living in a majority-black district.
. Of course, there was likely overlap. The calculations for the House and Senate are independent, not cumulative.
. The process for calculating this number is the same as described in note 3, supra. Ninety-nine percent of the ideal Senate District is 135,198 people. The population and black percentage figures for the Senate were drawn from APX 7. The sum of this calculation for all the majority-black Senate Districts is 106,-946 additional black people.
. See note 5, supra. The total population of Alabama that identifies as any part black is 1,281,118. 2010 Demographic Profile Data, NPX 325. There were 603,978 black individuals living in majority-black Senate Districts at the time of the 2010 census. APX 7. This leaves 677,140 black people not yet living in a majority-black Senate District. Thus the 106,946 additional black people needed to achieve the drafters quotas, see note 7, supra, represented 15.8% of those not already living in a majority-black district.
. There is one additional district, HD 85, that is majority black under the new plan in terms of total black population. APX 6. HD 85 was 48.37% total black population under the 2001 plan with 2010 census; it has a bare majority of total black population under the new plan. Id. However, it remains only plurality black in terms of voting-age population. Id.
. The majority refers to HD 73 as a majority-white district. That was true at the time of the 2001 reapportionment. With 2010 census data, HD 73 was 48.44% black and 44.07% white. Population Summary Report, SDX 406, at 6.
.As the majority notes, Newton has since passed away, and Hubbard has moved. Obviously the drafters were not aware of these circumstances at the time; thus these unforeseen later events do not bear on the question of whether race predominated over incumbency protection.
. If every additional resident of SD 26 under the new plan had been white, it would be 64.3% black. APX 7.
. Plaintiffs submitted a map illustrating the precinct splits statewide. See Map, NPX 357. There was no testimony to further explain this map.
. Hinaman testified that the only block-level political data he had available were generated by Maptitude on a strictly proportional basis from the precinct political data. That is, if a precinct voted 60% Republican in the last election, Maptitude indicates that each census block in that precinct voted 60% Republican in the last election. But of course in reality 100% of any particular census block might have voted for the Republican, or 0%, or anywhere in between. Hinaman acknowledged that these data were not accurate. Hinaman Dep., APX 75, at 113-4.
. The margin between Keahey's and Dial’s interpretation was even narrower for the other nearby majority-black district that Keahey discussed. Using the 2000 census data, as Keahey did, SD 24 was 62.41% black. APX 4 at 5. Using the 2010 census data, as Dial required, that district was 62.68% black. APX 7.
. HD 32 ( + .68%); HD 52 ( + .01%); HD 53 ( + .49%) (transplanted to Madison County); *1321HD 54 (+.13%); HD 55 (+.06%); HD 56 ( + .04%); HD 57 ( + .01%); HD 60 ( + .27%); HD 67 ( + .06%); HD 69 ( + .09%); HD 70 ( + .31%); HD 83 (+.67%); and HD 97 ( + .07%).
. SD 18 (-.81%); SD 23 ( + .02%); and SD 24 ( + .48%).
. HD 59 (+9.76%); HD 68 ( + 2.1%); HD 71 ( + 2.62%); HD 72 ( + 4.38%); HD 76 ( + 4.34%); HD 82 ( + 5.02%); and HD 84 (+1.73%).
. SD 26 ( + 2.47%); SD 28 ( + 8.91%); SD 33 ( + 6.82%).
. For the calculations underlying the figures in this paragraph, see notes 3 & 7, supra; APX 6 & 7.
. The majority finds that the plaintiff Alabama Legislative Black Caucus has standing to challenge these plans as racial gerrymanders, and I agree. I understand the Caucus to challenge each individual majority-black district in addition to the plans in their entirety, and find that it has standing to do so as well. I would also find three individual plaintiffs have standing to bring racial gerrymandering claims. Bobby Singleton is the Senator for majority-black SD 24, and so has standing for the reasons stated in the majority opinion. In addition, Alabama Democratic Conference plaintiffs Lynn Pettway and Stacey Stallworth have standing to challenge the abolition and movement of HD 73. The stipulations and trial testimony establish that both are residents of current HD 73. The plaintiffs claim that the drafters racially gerrymandered HD 73 out of existence. Thus, whichever surrounding district these plaintiffs ended up in under the new House plan, they claim that they were placed there predominantly because of race. This is sufficient for standing under United States v. Hays, 515 U.S. 737, 744-5, 115 S.Ct. 2431, 132 L.Ed.2d 635 (1995).
. I believe that the standard articulated in Miller, namely that a plaintiff must show that race was the predominant factor motivating a districting decision, is appropriate in cases like Shaw and Miller. But this is a different case. Here, black legislators and black voters are challenging the State's decision to place them in majority-black districts. Whether that same predominant-factor standard should apply in a case like this one, where the class of individuals seeking protection from a racial classification are members of a group historically and currently subject to invidious racial discrimination, is a serious open question. It may be that under these circumstances, the principles of general Fourteenth Amendment jurisprudence should control such a case. However, because the plaintiffs in this case are entitled to relief even under a predominant-factor analysis, I will assume for the purposes of this dissent that the Miller analysis is applicable to this case.
. HD 32 ( + .68%); HD 52 (+.01%); HD 53 (+.49%) (transplanted to Madison County); HD 54 (+.13%); HD 55 ( + .06%); HD'56 ( + .04%); HD 57 ( + .01%); HD 60 ( + .27%); HD 67 (+.06%); HD 69 (+.09%); HD 70 ( + .31%); HD 83 ( + .67%); HD 97 (+.07%).
. SD 18 (-.81%); SD 23 ( + .02%); SD 24 ( + .48%).
. HD 59 ( + 9.76%); HD 68 ( + 2.1%); HD 71 ( + 2.62%); HD 72 ( + 4.38%); HD 76 ( + 4.34%); HD 82 (+5.02%); HD 84 (+1.73%).
. SD 26 ( + 2.47%); SD 28 ( + 8.91%); SD 33 ( + 6.82%).
. The majority states that of the majority-black House Districts there are five under the new plan -with less than 60% black population, while there were only two such districts under the 2001 plan. Ante at 1301. I am puzzled by this observation. Using the 2010 census data and the 2001 district lines, as Hinaman did in seeking to achieve his quotas, there were actually six districts under 60% black (HD 32, 53, 54, 82, 83, 84) in addition to HD 85, which was under 50% total black population under the 2001 plan. APX 6. How many districts were over 60% black under the 2001 plan with 2000 census data, the figure on which the majority apparently relies, is not relevant to the consideration of the drafters’ success in achieving their quotas, which were defined by the 2001 districts with 2010 census data.
. HD 19 (-8.54%); HD 58 (-5.08%); HD 77 (-6.58%); HD 78 (-4.14%); HD 98 (-5.23%); HD 99 (-7.75%); HD 103 (-4.6%); SD 19 (- 6.26%); SD 20 (-14.58%).
. The majority argues that, during the 2001 redistricting, the legislature, then controlled by Democrats, also moved many black individuals into majority-black districts. The majority fails to point to any evidence in the record to support the conclusion that the State did so to achieve quotas, or that it subordinated any other districting principles in the process. But, most importantly, even if the State’s conduct in 2001 were unconstitutional, that would not excuse the State's constitutional violations in this case.
. While I reject the notion that the 2% rule explains why all the majority-black districts have the black percentages they have, I should not be understood to say 'that the rule could not have had some determinative line-drawing role as to a particular district. One of life's lessons (which, unfortunately, I have not always been able to abide) is to avoid speaking in absolutes. Thus, if there is a majority-black district for which the 2% rule explains, even in part, why its lines are as they are, I am willing to engage the majority in a determination of whether the plaintiffs should prevail as to that district. With this dissent, as I have stated, I am not saying that the plaintiffs should prevail as to all the districts. What I am saying is that there must be an individual , assessment for each district as whether race was a predominate factor.
. The same is true of counties. At his deposition, Hinaman testified that "... there would be county splits potentially based on the Voting Rights Act and not retrogressing a Majority/Minority district.” Hinaman Dep., APX 75, at 34.
.The majority emphasizes that the districts in this case are not as bizarre as those rejected in Shaw, 509 U.S. at 644, 113 S.Ct. 2816, or Vera, 517 U.S. at 973, 116 S.Ct. 1941. The Supreme Court has made it clear that bizarre district lines may be evidence of a Shaw violation, but are not a necessary part of such a claim. Miller, 515 U.S. at 913, 115 S.Ct. 2475 ("Shape is relevant not because bizarreness is a necessary element of the constitutional wrong or a threshold requirement of proof, but because it 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 it district lines.”).
. As I understand the majority’s test, any reduction in the black percentage, other than an unavoidable reduction, constitutes retrogression. At some points in its discussion, though, the majority qualifies this test: only “significant” or “substantial” reductions would be retrogressive. The majority does not explain what constitutes a significant or substantial reduction, or how a State is supposed to determine what is significant or substantial. I must conclude that these qualifiers are rhetorical rather than substantive. For if § 5 actually permitted some reduction of the black percentage on the majority’s view, that *1339view could not save these plans. For example, imagine any reduction up to 5% counts as insignificant. If the drafters hit their quotas of 65% black in a particular district, then even on the majority’s view § 5 only would have required 60% black population. The additional 5% black population under the new plan would have been included by racial gerrymandering without a narrowly tailored justification, and so the plans would be struck down. See Miller, 515 U.S. at 916, 115 S.Ct. 2475 (a racial gerrymander exists when "race was the predominant factor motivating the legislature's decision to place a significant number of voters within or without a particular district”) (emphasis added). Thus, to justify the establishment and accomplishment of racial quotas in this case, the majority’s view of § 5 must be that it required the drafters to hit their marks where possible, without any carve-out for ‘insignificant’ or less-than-substantial reductions.
. In § 5 analysis, the benchmark plan refers to the last districting plan in place before the challenged alteration.
. The majority also relies on a law review article suggesting a possible interpretation of the 2006 amendments. It is noteworthy that the actual conclusion of that article is a rejection of the majority’s view as well: ”[G]iv-en that the statute will be in place for twenty-five years, the standard ought to be flexible enough to adapt to changing political realities. An interpretation of the standard that would freeze the current minority percentages in all covered districts, for example, ignores the realistic possibility that the percentages required for minorities to elect their preferred candidates will likely change over time.” Nathaniel Persily, The Promise and Pitfalls of the New Voting Rights Act, 117 Yale L.J. 174, 218 (2007).
. Justice Souter’s dissent in Georgia suggested a similar approach: "percentages tell us nothing in isolation, and ... without contextual evidence the raw facts about population levels” cannot show retrogression or lack of retrogression. 539 U.S. at 505, 123 S.Ct. 2498.
. Specifically, with regard to his decision to abolish and relocate HD 53, Hinaman testified at his deposition that if he had maintained nine majority-black districts in Jefferson County their black populations might have been lowered. "They may have gone from 60 percent to 51 or something like that, and I didn't think that was — I thought that would potentially create preclearance issues.” APX 75, at 61. He stated that he never tried to draw nine such districts. Id. He believed it would be possible to do so. Id. at 86.
. A jurisdiction may still be required to obtain preclearance of redistricting plans, even after Shelby County, under the "bail-in” provision of § 3 of the VRA. See 42 U.S.C. § 1973a(c); Shelby Cnty., Ala. v. Holder, 679 F.3d 848, 855 (D.C.Cir. 2012), rev’d on other grounds - U.S. -, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013). That provision "authorizes courts to impose preclearance in response to violations of the Fourteenth and Fifteenth Amendments.” Travis Crum, Note, The Voting Rights Act's Secret Weapon: Pocket Trigger Litigation and Dynamic Preclearance, 119 Yale L.J. 1992, 2006 (2010). The state of the VRA is in flux at the moment, and it is unclear to what extent this provision will be utilized to fill the void left after Shelby County. See http://www.justice.gov/iso/opa/ag/ speeches/2013/ag-speech-l 30725 .html (Attorney General noting that he will seek a court order subjecting Texas to preclearance after Shelby County). However, Alabama has not been "bailed-in” and is therefore currently not subject to any preclearance requirement.
. Indeed, because this plan will continue to be in effect for years, I would find that it was not narrowly tailored even if it had already gone into effect; in strict scrutiny, we simply cannot allow unjustified racial classifications to continue.
. Interestingly, the majority observes that “Governor Wallace and segregation are long gone, and Alabama has virtually eliminated any racial gap in voter registration or participation,” citing to the State’s evidence submitted in the Shelby County. Ante at 1312. But this, if true, is exactly my point too. And my pointed question remains, Why these high racial percentages today? Why these racial quotas today?