Christian Ministerial Alliance, et al., Plaintiffs, v. John Thurston, Defendant. (2025)

Order

Multiple Arkansas citizens challenge how the General Assembly redrew the state's congressional district lines. Although their allegations were plausible enough to survive a motion to dismiss [Docs. 35, 42], the evidence does not back up their claims of racial discrimination. For that reason, we grant summary judgment to Secretary of State John Thurston.

I.

First came the 2020 census, then redistricting. [Doc. 63, at 1.] The census showed that some parts of the state had increased in population, while others had declined. To achieve “equal representation for equal numbers of people,” Wesberry v. Sanders, 376 U.S. 1, 18 (1964), legislators had to move voters out of the second district, which had grown more than the first and the fourth. [Doc. 63, at 1.]

Working with the Bureau of Legislative Research, the General Assembly considered dozens of possibilities. [Doc. 58-11, at 25, 78–80; Doc. 62-15.] The one it eventually selected cut into the second district's southeast corner, which bordered both the first and fourth districts. The old map had traced around the edge of Pulaski County, leaving it completely intact within the second district, with the first to the east and the fourth to the south. [Doc. 58-3, at 25 fig. 12.] The new map split it among all three. [Doc. 58-3, at 35 fig.21.] Little Rock, Arkansas's largest city, was no longer in a single congressional district. [Doc. 58-11, at 5.]

Tabular or graphical material not displayable at this time.

[Doc. 58-3, at 25 fig.12, 35 fig.21.]

“[L]ike clockwork,” litigation followed. Ark. State Conf. NAACP v. Ark. Bd. of Apportionment, 86 F.4th 1204, 1207 (8th Cir. 2023). See generally Simpson v. Hutchinson, 636 F. Supp. 3d 951 (E.D. Ark. 2022). The problem, according to the plaintiffs, was that the southeast corner of Pulaski County is home to a high concentration of black voters. [Doc. 42, at 2–3 (citing Simpson, 636 F. Supp. 3d at 955).] The result was “crack[ing]” them into districts where their votes carried less weight. [Doc. 20 ¶¶ 7, 130, 215–25; Doc. 58-3, at 20, 22.]

The complaint alleged two constitutional claims: one for racial gerrymandering under the Fourteenth Amendment and the other for vote dilution under the Fourteenth and Fifteenth Amendments. [Doc. 20 ¶¶ 215–25.] Although the two theories are “analytically distinct,” both require proof that impacting black voters was the point, not just the effect, of what the General Assembly did. Miller v. Johnson, 515 U.S. 900, 911 (1995) (citation omitted).

At the pleading stage, the plaintiffs established a plausible inference that it was, based on their allegation that “black and white voters ‘with the same party preferences ․ were sorted differently among the relevant districts.’ ” [Doc. 42, at 6 (quoting Doc. 20 ¶ 22).] Without it, the complaint would have fallen short of the plausibility line between racial discrimination, which is actionable, and old-fashioned “partisan politics,” which is not. [Doc. 42, at 6.] See Rucho v. Common Cause, 588 U.S. 684, 718 (2019); see also Ashcroft v. Iqbal, 556 U.S. 662, 682 (2009).

Even then, we warned that “[a]ctually proving” racial discrimination could “turn out to be a challenge.” [Doc. 42, at 6.] It was possible, after all, that “geography rather than race played the predominant role in the General Assembly's decision.” [Doc. 42, at 7.] Not to mention that the plaintiffs would have to “disentangle race from politics,” which “poses special challenges” when both can explain “similar oddities in a district's boundaries.” Cooper v. Harris, 581 U.S. 285, 308 (2017).

II.

As in many redistricting cases, the parties vigorously dispute whether race or politics motivated the decision. In one recent example, South Carolina State Conference of the NAACP v. Alexander, a three-judge district court found, following a trial, that South Carolina had discriminated against black voters in drawing district lines. 649 F. Supp. 3d 177, 197 (D.S.C. 2023). The Supreme Court reversed, but not before explaining how to resolve these types of disputes:

First, a party challenging a map's constitutionality must disentangle race and politics if it wishes to prove that the legislature was motivated by race as opposed to partisanship. Second, in assessing a legislature's work, we start with a presumption that the legislature acted in good faith.

Alexander v. S.C. State Conference of the NAACP, 602 U.S. 1, 6 (2024) (emphasis added). Our case, by contrast, arises in a summary-judgment posture, but we must still account for both points in analyzing the plaintiffs' racial-gerrymandering and vote-dilution claims.

The question surrounding any summary-judgment motion is whether a “genuine dispute [exists] as to any material fact.” Fed. R. Civ. P. 56(a). Ultimately, it comes down to whether a “reasonable [factfinder] could return a verdict for the nonmoving party.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). In a redistricting case like this one, it “makes no sense” to try to answer that question without considering “the quality and quantity of evidence required” to establish racial discrimination. Id. at 254; see id. at 255 (explaining that we “must view the evidence presented through the prism of the substantive evidentiary burden”). Alexander's discussion of the presumption of legislative good faith sheds light on what is needed.

The burden it places on a map's challengers is, in the Supreme Court's words, “especially stringent.” Alexander, 602 U.S. at 11. The “starting” point is to assume that the legislature had a legitimate, nonracial motivation whenever the evidence can “plausibly support multiple conclusions.” Id. at 10. Only by “rul[ing] out the possibility that [nonracial considerations] drove the districting process” can plaintiffs “overcome” it. Id. at 20, 24 (emphasis added); see Abbott v. Perez, 585 U.S. 579, 605 (2018) (“[I]t [is] the plaintiffs' burden to overcome the presumption of legislative good faith and show that the ․ [l]egislature acted with invidious intent.”); Miller, 515 U.S. at 915 (“Although race-based decisionmaking is inherently suspect, until a claimant makes a showing sufficient to support that allegation the good faith of a state legislature must be presumed.” (citation omitted)). Requiring any less would be “clear factual error.” Alexander, 602 U.S. at 17.

At summary judgment, the presumption separates the “justifiable inferences” we can make from those we cannot. Anderson, 477 U.S. at 255. For example, an inference of race discrimination is unjustifiable when partisanship or geography “provides an entirely reasonable explanation” for the changes from a previous map. Alexander, 602 U.S. at 21.

It also sets the ground rules for the parties. Although we cannot rewrite history, see Bethune-Hill v. Va. State Bd. of Elections, 580 U.S. 178, 189–90 (2017) (rejecting “post hoc justifications”),¹ if the plaintiffs' evidence fails to overcome the presumption, the state can win simply by “pointing out to the district court[ ] that there is an absence of evidence to support [their] case,” Celotex Corp. v. Catrett, 477 U.S. 317, 325 (1986). Evidence that “plausibly supports] multiple conclusions,” Alexander, 602 U.S. at 10, creates a “failure of proof concerning an essential element” that relieves the state of having to present evidence of its own “negating” the other side's case, Celotex, 477 U.S. at 323. It becomes a steep hill for a map's challengers to climb.

III.

Steep enough, in fact, that the plaintiffs' evidence of racial gerrymandering in this case is “[in]sufficient to support an inference that can overcome” it. Alexander, 602 U.S. at 19–20. The one they need a reasonable factfinder to draw is “that race was the predominant factor motivating the [General Assembly's] decision” to split Pulaski County into three congressional districts. Miller, 515 U.S. at 916. Awareness, or even acceptance, of a racially disparate impact is not enough. See id. (acknowledging that legislators are “almost always․ aware of racial demographics”). The proof, construing all reasonable inferences in their favor, must show “that the legislature subordinated traditional race-neutral districting principles ․ to racial considerations.” Id.; see Cooper, 581 U.S. at 291 n.1 (noting that it does not matter whether the legislature prioritized race for its own sake or as a proxy for “other goals, including political ones”).

Sometimes an explicit racial target or “smoking[-]gun” paper trail makes predominance easier to prove, Cooper, 581 U.S. at 299–300, 310–11, 318, but neither exists here. Rather, the plaintiffs face the “much more difficult” task of proving racial discrimination circumstantially. Alexander, 602 U.S. at 8.

A.

One way to do so is through “an alternative map.” Id. at 34. In theory, if the General Assembly could have drawn the district lines to achieve both “legitimate political objectives” and “significantly greater racial balance,” without sacrificing “traditional districting principles,” it would go a long way toward proving intentional discrimination. Easley v. Cromartie, 532 U.S. 234, 258 (2001). Such a map would leave us wondering what, other than a desire to target black voters, could explain the decision to pick a different one instead. See Alexander, 602 U.S. at 35 (emphasizing “[t]he evidentiary force of an alternative map”); Cooper, 581 U.S. at 317, 319 (describing it as “key evidence,” “highly persuasive,” and “sometimes ne[cessary]․ as a practical matter”).

The inference only works, however, if the alternative map still “accomplish[es] all [the legislature's] partisan goals.” Cooper, 581 U.S. at 317. If it does not, then it just highlights how the pursuit of a nonracial aim—like retention, partisanship, or geography—could have led to an unintended racial disparity. All three of the plaintiffs' alternatives fall short in exactly this way.

Tabular or graphical material not displayable at this time.

[Doc. 58-3, at 25 fig.12, 38 fig.23.]

Each comes from William Cooper, an expert on “redistricting and demographics” who has testified in dozens of voting-rights cases. [Doc. 58-3, at 2.] The first is supposedly a “least[-]change plan” that seeks to keep each district's population together, a goal known as “core retention.” [Doc. 58-3, at 37 n.13, 38.] See Alexander, 602 U.S. at 27 (noting that legislatures “usually begin with the existing map” and only change it as necessary). It splits White instead of Pulaski County, and moves Van Buren from the second to the fourth congressional district. [Doc. 58-3, at 25 fig.12, 38 & fig.23.]

Despite purportedly “prioritiz[ing] core retention,” this alternative does worse on that measure than the one the General Assembly adopted. [Doc. 58-3, at 38–39.] The adopted map keeps 92.16% of the population in the same district as before, whereas Cooper's alternative manages only 87.53%. [Doc, 58-3, at 40 fig.25.] The numbers are even worse for legitimate considerations the alternative map does not “prioritize[ ],” like politics. [Doc. 58-3, at 38.] According to the state's expert on “demography, census data, and advanced analytics,” it would have been worse for Republicans than either the map it chose or the benchmark map from 2011. [Doc. 58-1, at 7; Doc. 58-7, at 30 tbls.VII.A.1 & VII.A.2.] These problems plausibly explain why it was not the Republican-controlled General Assembly's choice. See Alexander, 602 U.S. at 6 (explaining that “a legislature may [constitutionally] pursue partisan ends when it engages in redistricting”).

Tabular or graphical material not displayable at this time.

[Doc. 58-3, at 25 fig. 12, 41 fig.26.]

Cooper's second alternative suffers from some of the same problems. It leaves all of Pulaski County in the second district, just like the first alternative, but shuffles around several others instead, [Doc. 58-3, at 41 fig.26.] It is not as weak politically as the first, but it is still worse for Republicans than the adopted map, no matter which measure is used. [Doc. 58-3, at 43 fig.28; Doc. 58-7, at 30 tbls.VII.A.1 & VII.A.2.] Cooper admits, in fact, that it would have led to a two-percentage-point swing toward President Biden in the second district had it been in place in 2020. [Doc. 58-3, at 43 fig.28.] It also does even worse in terms of core retention, keeping just 80.31% of the population in place. [Doc. 58-3, at 42 fig.28.] The adopted map's 92.16% retention rate provides a reason why “a rational legislature, driven only by its professed mapmaking criteria,” could have made the choice it did. Alexander, 602 U.S. at 34; see id. at 26–27 (explaining that the “fail[ure] to consider ‘core district retention” ’ is a “serious flaw” that “betrays a blinkered view of the redistricting process”).

Tabular or graphical material not displayable at this time.

[Doc. 58-3, at 25 fig.12; Doc. 62-9, at 8 fig.l.]

Cooper's final alternative redraws district lines even more extensively. It chops off the second district's western side and stretches it north to the Missouri border. [Doc, 62-9, at 8 fig.1.] No surprise that it would have had the most dramatic effect on core retention by reducing it to 70.66%. [Doc. 62-9, at 10 fig.3.] It also would have split two additional cities, broken up three more school districts, and tripled the deviation from one-person-one-vote proportionality. [Doc. 62-9, at 10 fig.3.] There were plenty of reasons to reject it, in other words, without factoring in race. See Alexander, 602 U.S. at 10, 26–27.

The plaintiffs argue that we should not assume that Arkansas's legislators were sticklers about these metrics. They point to language from Alexander describing “alternative[s] ․ that [we]re comparably consistent with traditional districting principles.” Id. at 10 (emphasis added) (quoting Cromartie, 532 U.S. at 258). In their view, comparable means close, so we should cut them some slack at summary judgment. But cf. Cooper, 581 U.S. at 317 (explaining that an alternative map gets its “persuasive” power from “showf[ing] that the legislature had the capacity to accomplish all its partisan goals without moving so many members of a minority group” (emphasis added)).

Whatever wiggle room the term “comparabl[e]” leaves for borderline cases, Alexander, 602 U.S. at 10 (quoting Cromartie, 532 U.S. at 258), this is not one of them. The reason, again, is the presumption of legislative good faith. Even if the General Assembly could have thought the alternate maps were close enough to be interchangeable on some nonracial factors, the discrepancies in core retention and partisan outcomes were still large enough to be disqualifying. See Cromartie, 532 U.S. at 257 (emphasizing “the sensitivity ․ that district courts must show to avoid treading on legislative prerogatives”). In these circumstances, when only one possible hypothesis would lead to bad faith, we must reject “the less charitable conclusion that the legislature's real aim was racial.” Alexander, 602 U.S. at 22; see id. at 20 (“In light of the presumption of legislative good faith, [a] possibility is dispositive.”).

For the same reason, the plaintiffs get the law backwards when they insist that the alternative maps only look worse than the adopted one because we are jumping to conclusions about what the General Assembly stressed in the selection process. The overall idea is that maybe the areas where their maps underperform meant little to legislators. If the plaintiffs had independent evidence to that effect, then perhaps we could ignore the apparent shortcomings on those metrics and use their maps to rebut other innocent explanations. Cf. Cooper, 581 U.S. at 317–18 (explaining that a strong enough combination of direct and circumstantial evidence can work “sans any map”). But without such proof, maps like these, which look better from some angles and worse from others, invite too much guesswork. And in the face of a good-faith presumption, guesswork is simply not enough. See Cromartie, 532 U.S. at 249 (explaining that it does not matter if “other ways ․ exist” to accomplish “some” of the legislature's goals, “unless the evidence also shows that these hypothetical alternative districts would have better satisfied the legislature's other nonracial political goals as well as traditional nonracial districting principles” (second emphasis added)). Even when we are talking about summary judgment.

B.

If a comparably performing map without the racial effects existed, we can safely assume the plaintiffs would have found it. See Alexander, 602 U.S. at 36–37. Now that district maps are so “easy to produce,” failing to identify one suggests that it “cannot be created.” Id. (noting that the challengers introduced “tens of thousands of other maps into the record,” but none performed as well as the one the legislature picked). The “adverse inference” is so strong, in fact, that “[i]t would be clear error․ to overlook [it],” and only “extraordinarily powerful circumstantial evidence” can overcome it. Id. at 35, 37. In this case, the plaintiffs offer little to counterbalance the absence of a comparable map. See id.; Anderson, 477 U.S. at 254 (explaining that, when a party's evidence “is of insufficient caliber or quantity to allow a rational finder of fact to find [in its favor],” summary judgment is appropriate).

1.

Start with the expert report from Dr. Baodong Liu, who studies “the relationship between election systems and the ability of minority voters to participate equally in the political process.” [Doc. 62-17, at 3.] His conclusions are that “race was a statistically significant factor in the design of the” second district and “consideration of voters' partisan preferences is not a statistically supportable alternative explanation” for the map the General Assembly chose. [Doc. 62-17, at 32.]

He relied on two statistical analyses of the second district. The first involved drawing a line around an area a bit larger than its current boundaries and assuming anyone within this “envelope” had an equally likely chance to end up as a second-district voter. [Doc. 62-17, at 8.] The demographic mismatch he observed between the hypothetical envelope and the real-life district was, in his view, a “sign that race [was] empirically related to how the [lines were] drawn.” [Doc. 62-17, at 8–9, 16 & tbl. 1.] Same went for his other method, which compared who was added, removed, and left in place when the boundary lines changed in 2021. [Doc. 62-17, at 9–10.] Based on the assumption that each voter had the same odds of being in any of those categories, the fact that black voters were disproportionately removed was itself evidence of race discrimination. [Doc. 62-17, at 10–11, 17 & tb1.2.] Dr. Liu, in other words, inferred causation from correlation.

Missing from both approaches, however, was a control for where voters of different races lived. As the Supreme Court pointed out when Dr. Liu performed a similar analysis in Alexander, his conclusions rest on “highly unrealistic” assumptions. 602 U.S. at 32.

For one thing, the General Assembly, unlike Dr. Liu, could not “treat[ ] each voter as an independent unit” and “include or exclude” them regardless of what happened to their neighbors. Id.; see id. at 31 (explaining that Dr. Liu “failed to account for contiguity and compactness”). No real-world mapmaker could, for example, “assign a black Republican to one district while moving a black Democrat who lives in the same apartment building to another.” Id. at 32. Closely related is the geographic limitation that voters living in the middle of the district were much less likely to be reassigned than those living right on the border. See id. at 29 (identifying the same basic mistake in an analysis that focused on precincts, rather than individual voters); Cromartie, 532 U.S. at 247 (rejecting expert testimony that failed to consider “whether the excluded ․ precincts were located near enough to [the district's] boundaries or each other for the legislature as a practical matter to have drawn [its] boundaries to have included them”). The correlation to race he found could just reflect those realities.

It is true that “weighing ․ the evidence” must usually await trial. Anderson, 477 U.S. at 255. The problem for the plaintiffs, however, is that there is nothing to weigh here because the Supreme Court has already told us that Dr. Liu's “defect[ive]” methodology “preclude[s] reliance on [his] report.” Alexander, 602 U.S. at 32. Had he taken steps to fix those issues, perhaps there could have been a genuine fact dispute. But not here, faced with the same basic “mistake[s]” as before. Id. at 30 (explaining it would be clearly erroneous to do otherwise); see Anderson, 477 U.S. at 252 (emphasizing that “evidence on which the [factfinder] could reasonably find for the plaintiff” is necessary to defeat summary judgment).

2.

The remaining evidence is the legislative record itself and a report analyzing it. Dr. Traci Burch, an “expert on political behavior, barriers to voting, and political participation,” thinks the General Assembly left enough clues for us to infer racial discrimination. [Doc. 58-11, at 3.] For support, she points to the last-minute decision to split Pulaski County, which required rushing the proposal through the relevant committees and then scheduling expedited votes by both the Senate and the House. [Doc. 58-11, at 6–25.] After some legislators objected to the lack of notice and discussion, the map's supporters explained it was necessary to avoid “another argument[,] ․ another thing[,]․ another fight.” [Doc. 58-11, at 24.] One of those fights was the impact on black voters, which came up during the hearings and debates that took place. [Doc. 58-11, at 41–45.] The response, to the extent the map's supporters had one, was to ignore racial effects altogether, despite having been told by staff from the Bureau of Legislative Research that they could take them into account if the goal was compliance with the Voting Rights Act.² [Doc. 58-11, at 39–40, 47–50.] See Cooper, 581 U.S. at 292 (noting that the “Court has long assumed” this point).

Instead of race, the record shows that the General Assembly's focus was on keeping counties, cities, and other communities of interest intact, while equalizing each district's population to the extent it could. [Doc. 58-11, at 25–38.] It is true, as the plaintiffs point out, that references to party politics during the hearings were “few and far between” and racial data was onscreen while staffers redrew the maps, sometimes with legislators in the room. [Doc. 58-11, at 45–46, 51.] But so were measures reflecting traditional redistricting criteria, even if partisan information was not. [Doc. 58-11, at 45–46, 51; Doc. 62-4, at 26–28.]

This evidence may be consistent with a racially discriminatory motive but, like everything else so far, fails to “overcome the presumption of legislative good faith.” Alexander, 602 U.S. at 20. Even before Alexander, we noted that “the procedural irregularities ․ f[a]ll short” of establishing that race motivated the General Assembly. [Doc. 42, at 6.] We reach the same conclusion today, with a full summary-judgment record. “[T]he brevity of the legislative process” does not automatically “give rise to an inference of bad faith—and certainly not an inference that is strong enough to overcome the presumption.” Abbott, 585 U.S. at 610; see, e.g., id. at 590, 610–11 (considering court-created maps that had been “pushed ․ through quickly in a special session” (citation omitted)).

Neither does the availability of racial data as the maps were being drawn. The lines splitting Pulaski County are not so “intricate[ly] refine[d]” that the “use of block-by-block racial data” in the map-drawing software is the only way to account for them. Bush v. Vera, 517 U.S. 952, 961–62 (1996) (plurality opinion). Add in former-Senator Jason Rapert's unrebutted testimony that a website, “Dave's Redistricting,” provided “partisan information” that “tons of people ․ tr[ied] to utilize” [Doc. 58-10, at 6–7], and the record as a whole “could plausibly support” a finding that factors like politics, retention, and geography, rather than race, explain the decision to pick a map that split Pulaski County. Alexander, 602 U.S. at 10; see id. at 6 (noting that “[l]egislators are almost always aware of the political ramifications of the maps they adopt”).

No one is saying that legislators and staffers were “[un]aware[ ] of the State's racial demographics.” Id. at 37 (explaining that “there is nothing nefarious” about such an “awareness”). Nor denying that they downplayed the role of politics. Those same issues, however, were also present in Alexander. “[K]ey State witnesses” repeatedly “den[ied] partisan motives” before eventually owning up to them. Id. at 79 (Kagan, J., dissenting); see id. (noting that “politicians don't like admitting to partisan gerrymanders”). In the end, it did not matter, because the presumption of good faith applied anyway. See id. at 36–37 (majority opinion).

Dr. Burch, for her part, came out the opposite way, but only after inferring racial motives from ambiguous evidence. At times, she even treated the fact that legislators said one thing as proof they meant something else. When a legislator admitted that “[t]his is about politics,” for example, she interpreted it as “a shield to deflect concerns about racial motivations.” [Doc. 58-11, at 52.] This sort of reading between the lines, if credited, becomes nothing more than a way to get around the presumption of good faith. See Alexander, 602 U.S. at 6, 22 (explaining that when a “stated partisan goal can easily explain [a] decision,” it is “err[or] [to] credit[ ] the less charitable conclusion that [its] real aim was racial”); Miller, 515 U.S. at 916 (stating that “the sensitive nature of redistricting and the presumption of good faith that must be accorded legislative enactments[ ] require[ ] courts to exercise extraordinary caution in adjudicating claims that a State has drawn district lines on the basis of race”). The expert would just be drawing the forbidden inferences for us. See Cromartie, 532 U.S. at 249 (“[A] conclusion is no stronger than the evidence that underlies it.”); cf. Alexander, 602 U.S. at 24-33 (holding it was clear error to rely on experts who ignored potential nonracial explanations).

* * *

Absent direct evidence of racial discrimination and with only weak circumstantial evidence supporting the plaintiffs' case, the presumption of legislative good faith tips the balance. Throw in the fact that no alternative map achieves the General Assembly's legitimate goals with “significantly greater racial balance,” Cromartie, 532 U.S. at 258, and we “could [not] reasonably find” on this record “that the plaintiff[s] proved” enough for their racial-gerrymandering claim to survive summary judgment. Anderson, 477 U.S. at 254.

IV.

The same goes for their vote-dilution claim.³ The theory behind it is that Arkansas drew the map “as a purposeful device ‘to minimize or cancel out the voting potential of” black voters. Miller, 515 U.S. at 911 (quoting City of Mobile v. Bolden, 446 U.S. 55, 66 (1980) (plurality opinion)). It is analytically distinct from the racial-gerrymandering claim in a couple of ways, but ultimately it ends up in the same place.

We start with the differences. Rather than a general race-based motivation, a vote-dilution claim requires the more specific showing that “the purpose” of a new map was to “dilut[e] the minority vote.” Alexander, 602 U.S. at 39 (citation omitted). The focus may be narrower but establishing it might be easier. If we treat it like other equal-protection claims, it requires the dilutive purpose to just be a “motivating factor” behind the map, not the predominant one. Village of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 265–66 (1977); cf. Alexander, 602 U.S. at 38–39 (suggesting that “[a] plaintiff pressing a vote-dilution claim” might need to “show[ ] that race played a predominant role,” plus a dilutive effect).

Regardless of these differences, however, the primary obstacle remains the same: the presumption of legislative good faith. See Abbott, 585 U.S. at 591–92, 603 (applying the presumption to an “intentional vote dilution” claim and explaining that it kicks in whenever “discriminatory intent” is an element); see also Alexander, 602 U.S. at 38 (vacating a vote-dilution finding in part because it failed to account for the presumption). Even assuming that dilutive intent must only be a motivating factor in the decision, the plaintiffs do not have enough evidence to get there.

The record, viewed most favorably to them, shows that the new map “bears more heavily on” black voters by spreading them out among the first, second, and fourth districts. Arlington Heights, 429 U.S. at 266 (citation omitted). The effect is not so “stark,” however, that “impact alone is․ determinative,” so getting past summary judgment requires “other evidence” of legislative intent. Id. Most of what the plaintiffs offer are the materials we have already discussed: maps, statistics, and legislative history, none of which are enough to infer a racial motivation. There is little by way of “[s]ubstantive departures [from] ․ factors usually considered important,” “contemporary statements by members of the decisionmaking body,” or other “legislative or administrative history” suggesting a dilutive intent. Id. at 267–68.

The only evidence we have not already considered is a report from Ryan Smith, a Ph.D. candidate studying “the long struggle for political and civil equality in the American South.” [Doc. 62-1, at 3.] He describes Arkansas's “long history” of racism and resistance to black voting, much of which predates the passage of the Voting Rights Act. [Doc. 62-1, at 4, 15–24.] Even if he identifies a few scattered examples since then, none are “reasonably contemporaneous with the challenged decision,” giving us little insight into what the General Assembly may have been thinking four years ago. McCleskey v. Kemp, 481 U.S. 279, 298 n.20 (1987); see Bolden, 446 U.S. at 74 (plurality opinion) (“[P]ast discrimination cannot, in the manner of original sin, condemn governmental action that is not itself unlawful.”).

V.

We accordingly GRANT the motion for summary judgment.

IT IS SO ORDERED this 6th day of June, 2025.

FOOTNOTES

1.   Make no mistake, the question is what drove the legislature “in reality,” not what “could have ․ but․ did not.” Bethune-Hill, 580 U.S. at 189–90. Compare id. at 189 (emphasizing that only the “actual considerations” matter), with FCC v. Beach Commc'ns, Inc., 508 U.S. 307, 315 (1993) (explaining that, for rational-basis review, “it is entirely irrelevant․ whether the conceived reason․ actually motivated the legislature”). If the evidence points directly to a race-based motivation, the presumption is no reason to pretend otherwise, even if it would have been possible for the legislature to reach the same result in good faith. See Cooper, 581 U.S. at 301 n.3; cf. Alexander, 602 U.S. at 20 (explaining that the presumption kicks in when “nothing rules out th[e] possibility” of a nonracial explanation).

2.   The plaintiffs, however, have never claimed that the new map violates the Voting Rights Act, let alone that explicit consideration of race was necessary to avoid doing so.

3.   The plaintiffs ground this claim in both the Fourteenth and Fifteenth Amendments. We still “disagree about whether Fifteenth Amendment vote-dilution claims exist,” but again “assume without deciding” that they do. [Doc. 42, at 8 (citation omitted).]

STRAS, Circuit Judge.