Amicus Curiae Brief — Thomas C. Alexander, in His Official Capacity as President of the South Carolina Senate, et al., Appellants v. The South Carolina State Conference of the NAACP, et al.

No. 22-807

In the

Supreme Court of the United States

──────────────────────────

THOMAS C. ALEXANDER, IN HIS OFFICIAL CAPACITY AS

PRESIDENT OF THE SOUTH CAROLINA SENATE, ET AL.,

Appellants,

v.

THE SOUTH CAROLINA STATE CONFERENCE OF THE

NAACP, ET AL.,

Appellees.

──────────────────────────

On Appeal from the United States District Court for the

District of South Carolina

──────────────────────────

BRIEF OF ALABAMA AND 15 OTHER STATES AS

AMICI CURIAE IN SUPPORT OF APPELLANTS

──────────────────────────

OFFICE OF ALABAMA

ATTORNEY GENERAL

501 Washington Avenue

Montgomery, AL 36130

(334) 242-7300

edmund.lacour@

AlabamaAG.gov

STEVE MARSHALL

Alabama Attorney General

Edmund G. LaCour Jr.

Solicitor General

Counsel of Record

Bethany C. Lee

Assistant Solicitor General

Counsel for Amicus Curiae

State of Alabama

i

QUESTION PRESENTED

The district court found that the South Carolina

General Assembly sought to alter the partisan tilt of

District 1 in the State’s congressional redistricting

plan and had the partisan data needed to do so. The

General Assembly thus had no reason to use race to

achieve that partisan goal. Did the district court err

in ignoring the presumption of legislative good faith

and concluding that the General Assembly used race

to draw District 1?

ii

TABLE OF CONTENTS

Question Presented ..................................................... i

Table of Contents ....................................................... ii

Table of Authorities................................................... iii

Interest of Amici Curiae..............................................1

Summary of Argument................................................3

Argument .....................................................................5

I.

The District Court Ignored The

Presumption Of Legislative Good Faith ........5

II. The District Court Invented And

Ascribed To The General Assembly An

Irrational Racial Intent ................................ 10

III. The District Court Effectively Created A

Disparate-Impact

Regime

For

Redistricting ................................................. 13

Conclusion ................................................................. 21

iii

TABLE OF AUTHORITIES

Cases

Abbott v. Perez,

138 S. Ct. 2305 (2018) .............. 3, 6, 8, 9, 10, 11, 19

Ala. Legis. Black Caucus v. Alabama,

575 U.S. 254 (2015) .............................................. 11

Allen v. Milligan,

No. 21-1086 (filed Jan. 28, 2022) ......................... 16

Ashcroft v. Iqbal,

556 U.S. 662 (2009) ................................................7

Brnovich v. Democratic Nat’l Comm.,

141 S. Ct. 2321 (2021) .......................................... 18

Bush v. Vera,

517 U.S. 952 (1996) .................. 7, 11, 12, 14, 18, 21

Cooper v. Harris,

581 U.S. 285 (2017) ................ 6, 7, 9, 11, 13, 14, 19

Easley v. Cromartie, (Cromartie II)

532 U.S. 234 (2001) ........................ 8, 10, 13, 14, 20

Edwards v. Aguillard,

482 U.S. 578 (1987) ................................................6

Fletcher v. Peck,

10 U.S. 87 (1810) ....................................................1

iv

Gomillion v. Lightfoot,

364 U.S. 339 (1960) .......................................... 7, 21

Hunt v. Cromartie, (Cromartie I)

526 U.S. 541 (1999) .......................................... 7, 18

Hunter v. Underwood,

471 U.S. 222 (1985) ................................................7

League of Women Voters of Fla., Inc. v. Fla. Sec’y

of State,

32 F.4th 1363 (11th Cir. 2022) ..............................9

McCleskey v. Kemp,

481 U.S. 279 (1987) ............................................ 1, 7

Miller v. Johnson,

515 U.S. 900 (1995) .............. 1, 3, 4, 5, 6, 14, 17, 21

Milligan v. Merrill,

Case No. 2:21-cv-1530-AMM (N.D. Ala. filed

Dec. 15, 2021) (ECF No. 69) ................................ 16

Pers. Adm’r of Mass. v. Feeney,

442 U.S. 256 (1979) .............................. 6, 14, 17, 19

Rucho v. Common Cause,

139 S. Ct. 2484 (2019) .......................................... 12

Shaw v. Reno,

509 U.S. 630 (1993) .............................................. 18

v

Vill. of Arlington Heights v. Metro. Hous. Dev.

Corp.,

429 U.S. 252, 266 (1977) ...................................... 21

Washington v. Davis,

426 U.S. 229 (1976) .............................................. 14

Other Authorities

Andrew Gelman & Eric Loken, The Statistical

Crisis in Science, 102 AMERICAN SCIENTIST

460 (2014) ............................................................. 15

Erin M. Kirkham & Edward M. Weaver, A

Review of Multiple Hypothesis Testing in

Otolaryngology Literature, 125

LARYNGOSCOPE 599, 599-603 (2015) .................... 15

Denes Szucs & John P.A. Ioannidis, When Null

Hypothesis Significance Testing Is Unsuitable

for Research: A Reassessment, FRONTIERS

HUM. NEUROSCIENCE, Aug. 3, 2017 ..................... 15

1

INTEREST OF AMICI CURIAE1

The States of Alabama, Alaska, Arkansas, Florida,

Georgia, Indiana, Iowa, Louisiana, Mississippi, Montana, Nebraska, South Carolina, Tennessee, Texas,

Utah, and West Virginia respectfully submit this brief

as amici curiae in support of the South Carolina Appellants. Amici are States seeking to ensure that “the

good faith of a state legislature” continues to be “presumed.” Miller v. Johnson, 515 U.S. 900, 915 (1995).

Federal courts should never be eager to find a hidden,

unlawful purpose lurking behind a facially valid state

law. As Chief Justice Marshall declared, “it is not on

slight implication and vague conjecture that the legislature is to be pronounced to have transcended its

powers, and its acts to be considered as void.” Fletcher

v. Peck, 10 U.S. 87, 128 (1810). Instead, “[t]he opposition between the constitution and the law” must be

“clear.” Id. And in the redistricting context, courts

should be especially sure to tread lightly, as “[f]ederalcourt review of districting legislation represents a serious intrusion on the most vital of local functions.”

Miller, 515 U.S. at 915. Thus, when there are “legitimate reasons” for a legislature to enact a particular

law, courts should “not infer a discriminatory purpose

on the part of the State.” McCleskey v. Kemp, 481 U.S.

279, 298-99 (1987).

The district court tossed all that aside. It not only

ignored the presumption of legislative good faith but

flipped it on its head. The court recognized that the

1 More than 10 days before this brief was due, Alabama ensured

that counsel of record for all parties received notice of Alabama’s

intent to file this brief. See S. Ct. R. 37.2.

2

General Assembly sought to change the partisan tilt

of District 1 and that partisan legislators had the partisan data needed to accomplish their partisan goal.

Yet the court concluded, without any direct evidence,

that the racial effect of the resulting map was proof of

a racial target. By creating this racial target out of

thin air, the court effectively imposed a disparate impact regime for redistricting.

This presumption of bad faith represents a serious

threat to self-government and our federalist system.

Amici States have a strong interest in ensuring that

this Court emphatically rejects the district court’s approach and makes clear that applying the presumption of legislative good faith is not optional. The Court

should reverse the district court’s order.

3

SUMMARY OF ARGUMENT

“Redistricting is primarily the duty and responsibility of the State,” Abbott v. Perez, 138 S. Ct. 2305,

2324 (2018) (cleaned up), and federal courts must “exercise extraordinary caution in adjudicating claims

that a State has drawn district lines on the basis of

race,” Miller, 515 U.S. at 916. Thus, until plaintiffs

can meet their high burden to show that a legislature

“acted with invidious intent,” “the good faith of the

state legislature must be presumed.” Abbott, 138 S.

Ct. at 2324 (cleaned up). But the district court did precisely what the Court in Abbott prohibited: It “reversed the burden of proof,” holding that the General

Assembly had a racial intent simply because its partisan goal had a racial effect. Id. at 2325. The court declined to mention, much less apply, a presumption of

good faith.

The district court instead presumed bad faith, creating a racial target from whole cloth and then using

that target to impute racial intent to the South Carolina General Assembly. The court accepted that

“[w]hen the South Carolina House and Senate began

considering congressional reapportionment in 2021,

the Republican majorities in both bodies sought to create a stronger Republican tilt to Congressional District No. 1.” App.21a. The court then cited an expert

report and a closing statement demonstrative when

noting that “[a]nalyses of partisan voting patterns

within Congressional District No. 1 provided by both

Plaintiffs and Defendants indicated that a district in

the range of 17% African American produced a Republican tilt, a district in the range of 20% produced a

4

‘toss up district,’ and a plan in the 21-24% range produced a Democratic tilt.” App.22a-23a.

So far, so good. But then things took a bizarre turn.

The court purported to “find[] that this data demonstrating the need to limit the African American population to a certain level to produce the desired partisan tilt resulted in a target of 17% African American

population for Congressional District No. 1.” App.23a.

This leap in logic—from a “desired partisan tilt” to “a

target of 17% African American population,” id.—is as

inexplicable as it was unexplained. The court never

assessed whether the General Assembly, when pursuing its “desired partisan tilt,” id., acted “because of,

not merely in spite of,” the racial impact. Miller, 515

U.S. at 916 (cleaned up). Indeed, the direct evidence

consistently showed that the General Assembly was

motivated by a partisan target rather than a racial

one and that it had the partisan data needed to accomplish its non-racial goal.2

After inventing a racial target and projecting it

onto South Carolina, the court proceeded to conflate

awareness of race with intentional racial sorting,

2

Moreover, the data relied on by the district court cannot

“demonstrat[e] the need to limit the African American population

to a certain level.” App.23a (emphasis added). The data show

only that a handful of proposed plans featuring different percentages of African American voters in District 1 would have produced different partisan tilts. Those limited data points do not

rule out other potential iterations of District 1 with both a higher

percentage of African American voters and a Republican tilt. The

court thus clearly erred in finding that District 1 could not lean

Republican unless the percentage of African American voters in

the district was “in the range of 17%.” App.22a-23a.

5

reject partisan actors’ partisan goals as implausible

explanations for partisan actions, and presume that

any ostensible oddities in the plan were explainable

only by racial discrimination. The court even

strangely suggested that the General Assembly used

“partisanship as a proxy for race,” App.33a, never explaining why partisans would do that in a case where

it was unnecessary to use race to accomplish partisan

goals.

The district court’s standard stacks the deck

against the States. Redistricting in particular is a difficult subject for legislatures, with a “complex interplay of forces that enter a legislature’s redistricting

calculus.” Miller, 515 U.S. at 915-16. But for the district court, the math here was simple—District 1

ended up with a black population of 17%, so the General Assembly must have set “a target of 17%” from

the start. App.23a. The district court thus not only

failed to apply a presumption of good faith but did the

opposite, equating the racial effects of a partisan decision with intentional racial discrimination.

The district court’s approach is essentially a disparate-impact regime for redistricting, which would

lead to a flood of new lawsuits for state and local governments. This Court should note probable jurisdiction and reverse the decision below.

ARGUMENT

I. The District Court Ignored The Presumption

Of Legislative Good Faith.

A. The presumption of legislative good faith requires courts to presume that a legislature acted for

6

legitimate reasons unless there is unmistakable evidence to the contrary. The presumption reflects the

different roles that legislatures and courts occupy in

our federal system. Every time legislatures act, they

must “exercise the political judgment necessary to balance competing interests.” Miller, 515 U.S. at 915.

Disputes about whether laws are “undemocratic and

unwise” should remain in the statehouse, not the

courthouse. Pers. Adm’r of Mass. v. Feeney, 442 U.S.

256, 280 (1979). Moreover, “discerning the subjective

motivation of those enacting [a] statute is … almost

always an impossible task.” Edwards v. Aguillard,

482 U.S. 578, 636 (1987) (Scalia, J., dissenting). The

presumption of legislative good faith thus safeguards

the separation of powers between the States and the

federal government, recognizes the near-impossibility

of divining the intent of a multi-member legislative

body, and steers federal courts away from the temptation to ascribe bad motives whenever a judge views a

legislature’s work as bad policy.

The presumption “takes on special significance in

districting cases,” where “federal-court review of districting legislation represents a serious intrusion on

the most vital of local functions.” Abbott, 138 S. Ct. at

2324 (cleaned up). Courts “must be sensitive to the

complex interplay of forces that enter a legislature’s

redistricting calculus.” Miller, 515 U.S. at 915-16. Any

other approach would “invite losers in the redistricting process to seek to obtain in court what they could

not achieve in the political arena.” Cooper v. Harris,

581 U.S. 285, 335 (2017) (Alito, J., concurring in part).

To be sure, the presumption of good faith can be

overcome in certain circumstances, such as when the

7

State’s conceded “aim” is to “disenfranchis[e] practically all of” one racial group, Hunter v. Underwood,

471 U.S. 222, 230 (1985), or when the State asserts an

explicit desire to target a racial percentage in a district, Cooper, 581 U.S. at 311. There have also been a

few “rare cases in which a statistical pattern of discriminatory impact demonstrated a constitutional violation,” but those cases involved “statistical disparities” so stark that they were “‘tantamount for all practical purposes to a mathematical demonstration’ that

the State acted with a discriminatory purpose.” McCleskey v. Kemp, 481 U.S. 279, 294 n.12

(1987) (quoting Gomillion v. Lightfoot, 364 U.S.

339, 341 (1960)). In all other cases, where there are

nonracial, “legitimate reasons” for a law, courts are

not to “infer a discriminatory purpose.” Id. at 299.

Proving an illicit purpose is thus no mean feat even

where the decisionmaker is a single government official. See Ashcroft v. Iqbal, 556 U.S. 662, 680-83 (2009).

And plaintiffs face even greater “difficulties” where

the decisionmaker is a legislative body as large as a

state legislature. Hunter, 471 U.S. at 228. These difficulties are further compounded for plaintiffs alleging

a racial gerrymander when “political and racial reasons are capable of yielding similar oddities in a district’s boundaries.” Cooper, 581 U.S. at 308. Plaintiffs

must “disentangle race from politics and prove that

the former drove a district’s lines,” id. (citing Hunt v.

Cromartie, 526 U.S. 541, 546 (1999) (Cromartie I)),

which requires proof that “political considerations

were subordinated to racial classification,” Bush v.

Vera, 517 U.S. 952, 970-71 (1996).

8

Because of the strength of the good-faith presumption, weak circumstantial evidence is “plainly insufficient to prove … intentional discrimination” when a

legislature’s stated purpose is “reasonable” and “legitimate” on its face. Abbott, 138 S. Ct. at 2327. Indeed,

where “racial identification is highly correlated with

political affiliation,” even evidence of “the district’s

shape, its splitting of towns and counties, and its high

African-American voting population” is insufficient,

“as a matter of law,” to justify a finding of racial intent. Easley v. Cromartie, 532 U.S. 234, 243 (2001)

(Cromartie II) (emphasis added). In Abbott, for example, the Court faulted the lower court for imputing bad

faith to the legislature based on wholly circumstantial

evidence, such as recent discriminatory intent in prior

redistricting and the “willful ignorance” of the legislature toward deficiencies in the new plan. Abbott, 138

S. Ct. at 2327-29 (cleaned up). Rejecting this reliance

on circumstantial evidence, the Court emphasized

that “[t]he only direct evidence … suggest[ed] that the

2013 Legislature’s intent was legitimate” and that the

district court improperly “discounted this direct evidence.” Id. at 2327. After all, the legislature’s expressed intent was “entirely reasonable and certainly

legitimate,” so the circumstantial evidence was

“plainly insufficient to prove that the 2013 Legislature acted in bad faith and engaged in intentional discrimination.” Id.

B. Yet despite Defendants’ repeated references to

Abbott and its required good-faith presumption, see,

e.g., Doc. 323 at 1, 29, 33, the district court here didn’t

even mention, much less apply, the presumption of

legislative good faith. That unexplained failure to

9

apply the correct legal standard is grounds enough to

reverse. See Abbott, 138 S. Ct. at 2326 (“[W]hen a finding of fact is based on the application of an incorrect

burden of proof, the finding cannot stand.”); see also

League of Women Voters of Fla., Inc. v. Fla. Sec’y of

State, 32 F.4th 1363, 1373 (11th Cir. 2022) (per curiam) (staying district court injunction pending appeal because “it does not appear to us that the district

court here meaningfully accounted for the presumption” of legislative good faith).

Worse, the court applied the very analysis denounced in Abbott by cross-examining the legislative

aide who drew the map, ascribing racial motivations

to his actions, and then assigning those motivations to

the State itself. In doing so, the district court improperly shifted the burden of proof to the State, holding

that the mapdrawer “failed to provide the Court with

any plausible explanation for the abandonment of his

‘least change’ approach’ … or the subordination of traditional districting principles.” App.29a. The court did

not attempt to explain why the mapdrawer’s testimony that he targeted a “partisan lean” was not a

“plausible explanation.” App.24a. This was not a credibility determination between conflicting testimony,

cf. Cooper, 581 U.S. at 309-10, but instead a complete

rejection of the idea that a good-faith partisan explanation could be plausible.

Indeed, the district court recognized yet then “discounted … direct evidence” that universally indicated

legitimate political motives. Abbott, 138 S. Ct. at 2327.

First, “Republican majorities in both” the House and

Senate “sought to create a stronger Republican tilt to

Congressional District No. 1.” App.21a. Second, “the

10

lead proponent of what would become the enacted congressional district plan … explained at trial that he

was seeking to include” particular “counties in the reconfigured Congressional District No. 1 to give the

district a stronger Republican lean.” App.22a. Third,

the mapdrawer “testified that he relied ‘one hundred

percent’ on data regarding ‘the partisan lean of the

district.’” App.24a (quoting Roberts Tr. 1558:13-19).

And fourth, the State admitted its goal to preserve

“the 6-1 Republican-to-Democratic split in House

seats.” Doc. 323 at 16. Yet, citing even less evidence

than the lower court in Abbott—which could at least

point to the Texas Legislature’s recent history of discriminatory map drawing—the district court inexplicably “discounted this direct evidence” and assumed a

secret racial motivation. Abbott, 138 S. Ct. at 2327.

When “[t]he only direct evidence brought to [the

Court’s] attention suggests that the … Legislature’s

intent was legitimate,” the district court’s contrary

finding cannot stand. Id. The facts relied on by the

district court “cannot, as a matter of law, support the

District Court’s judgment.” Cromartie II, 532 U.S. at

243.

II. The District Court Invented And Ascribed To

The General Assembly An Irrational Racial

Intent.

The district court didn’t just assume that Republican majorities in the General Assembly were discriminatory; the court deemed them to be dimwitted too.

How else to explain the court’s conclusion that these

partisan actors with partisan data on hand decided to

imperil their plan by needlessly injecting race into the

redistricting process? Occam’s razor suggests that a

11

rational, partisan legislature would simply use partisan data to alter the partisan lean of District 1. But

the district court found that legislators instead constructed and aimed first at a racial “target of 17% African American population for Congressional District

No. 1” (App.23a) as a bank shot means of hitting the

partisan target they formed when they “began considering congressional reapportionment in 2021.”

App.21a. That makes no sense.

And this lack of any plausible racial motive sets

this case apart from the racial gerrymandering cases

the Court has considered over the last several decades. This case, for example, does not involve a State

relying on the Voting Rights Act to try to justify racebased lines. See, e.g., Abbott, 138 S. Ct. at 2334

(“Texas does not dispute that race was the predominant factor in the design of HD90, but it argues that

this was permissible because it had ‘good reasons to

believe’ that this was necessary to satisfy § 2 of the

Voting Rights Act.”); Cooper, 581 U.S. at 299 (legislators advocated drawing a district “to be majority-minority, so as to comply with the VRA”); Ala. Legis.

Black Caucus v. Alabama, 575 U.S. 254, 275 (2015)

(attempts to comply with VRA led to racial gerrymander); Bush, 517 U.S. at 979 (State relying on VRA to

defend “bizarrely shaped” districts). Nor did the district court suggest that “legislators use[d] race as”

cover for partisanship, “thinking that a proposed district is more ‘sellable’ as a race-based VRA compliance

measure than as a political gerrymander.” Cooper,

581 U.S. at 308 n.7.

Neither is this a situation like that faced by the

mapdrawers in the early 1990s in Bush v. Vera, who

12

had access to racial data that was more detailed than

then-available partisan data, which created an incentive to use race as a proxy for partisanship. 517 U.S.

at 961-62. As the Court recounted, mapdrawers had

access to “unprecedented” “block-by-block racial

data,” and this “uniquely detailed racial data … enabled districters to make more intricate refinements on

the basis of race than on the basis of other demographic information.” Id. The result was unmistakable. Districts’ borders “change[d] from block to block”

in a “nearly perfect” emulation of “racial data at the

block-by-block level.” Id. at 961-62. Notably, “other

data, such as party registration and past voting statistics, were only available at the level of voter tabulation districts.” Id. at 961. Thus, because only racial

data were available with such granularity, only race

could explain the district’s “change from block to

block, from one side of the street to the other, … in

seemingly arbitrary fashion.” Id. at 962.

Unlike thirty years ago, today’s mapdrawers have

no need to use race as a proxy for partisanship because they “now have access to more granular data

about party preference and voting behavior than ever

before,” including “city-block-level data.” Rucho v.

Common Cause, 139 S. Ct. 2484, 2513 (2019) (Kagan,

J., dissenting). Thus, even if the district court wasn’t

willing to presume the South Carolina General Assembly’s good faith, it should have at least assumed

that the General Assembly would not needlessly (and

recklessly) sort voters based on race instead of partisanship when trying to “create a stronger Republican

tilt to Congressional District No. 1.” App.21a. “A legislature trying to secure a safe [Republican] seat is

13

interested in [Republican] voting behavior.”

Cromartie II, 532 U.S. at 245. Because that admitted

partisan aim explains this legislature’s actions, Plaintiffs failed “to disentangle race from politics and prove

that the former drove a district’s lines.” Cooper, 581

U.S. at 308.

In addition to the district court’s apparent theory

that the General Assembly used racial data as a proxy

for partisanship data it already possessed, the district

court appears to have also found that the General Assembly “use[d] partisanship as a proxy for race.”

App.33a. It is not clear exactly what the district court

meant by that. Why would a legislature ever use partisanship as a proxy for race, when detailed racial

data could be used instead? Even Plaintiffs never

claimed in their complaint that partisanship was used

as a proxy for race. See Doc. 267. And more to the

point, why would a legislature set on “creat[ing] a

stronger Republican tilt to Congressional District No.

1,” App.21a, limit its options for doing so by incorporating a racial target? The decision below underscores

the dangers of abandoning the presumption of legislative good faith and the need for the Court to reaffirm

the presumption.

III. The District Court Effectively Created A Disparate-Impact Regime For Redistricting.

A. As this Court has repeatedly recognized, nearly

every policy has the potential for some disparate racial impact. A legislative decision cannot be held discriminatory simply because “in practice it benefits or

burdens one race more than another”; such a rule

“would be far-reaching and would raise serious

14

questions about, and perhaps invalidate, a whole

range of tax, welfare, public service, regulatory, and

licensing statutes that may be more burdensome to”

some racial groups than to others. Washington v. Davis, 426 U.S. 229, 248 (1976).

This observation rings particularly true in redistricting, where “racial identification is highly correlated with political affiliation.” Cromartie II, 532 U.S.

at 243. “If district lines merely correlate with race because they are drawn on the basis of political affiliation, which correlates with race, there is no racial classification to justify, just as racial disproportions in the

level of prosecutions for a particular crime may be unobjectionable if they merely reflect racial disproportions in the commission of that crime.” Bush, 517 U.S.

at 968. At the same time, this correlation means that

courts must “‘exercise extraordinary caution’ in distinguishing race-based redistricting from politics-based

redistricting,” lest “the federal courts … be transformed into weapons of political warfare.” Cooper, 581

U.S. at 335 (Alito, J., concurring in part) (quoting Miller, 515 U.S. at 916). Thus, plaintiffs bringing racial

gerrymandering claims must prove “more than intent

as volition or intent as awareness of consequences.”

Miller, 515 U.S. at 916 (quoting Feeney, 442 U.S. at

279). When courts fail to hold plaintiffs to this high

burden, “they … invite the losers in the redistricting

process to seek to obtain in court what they could not

achieve in the political arena.” Cooper, 581 U.S. at 335

(Alito, J., concurring in part).

Conflating racial correlation with racial motives is

particularly dangerous because plaintiffs can almost

always

find

a

suspicious-looking

statistical

15

correlation. As scientists have routinely emphasized,

“statistical significance can obviously be obtained

even from pure noise by the simple means of repeatedly performing comparisons, excluding data in different ways, examining different interactions, controlling for different predictors, and so forth.” Andrew

Gelman & Eric Loken, The Statistical Crisis in Science, 102 AMERICAN SCIENTIST 460, 460-65 (2014).

Thus, if “a highly unscrupulous researcher … perform[ed] test after test in a search for statistical significance,” a statistically significant result “could almost certainly be found.” Id. As one study explains,

“[w]hen rolling one die, the chance of a six is 1/6, or

17%. When ten dice are rolled, the chance of at least

one landing on six is 84%. Similarly, when multiple

hypotheses are tested, each at a significance level of

0.05, the chance of obtaining at least one false positive

rises precipitously with the number of hypotheses

tested.” Erin M. Kirkham & Edward M. Weaver, A Review of Multiple Hypothesis Testing in Otolaryngology

Literature, 125 LARYNGOSCOPE 599, 599-603 (2015);

see also Denes Szucs & John P.A. Ioannidis, When

Null Hypothesis Significance Testing Is Unsuitable for

Research: A Reassessment, FRONTIERS HUM. NEUROSCIENCE, Aug. 3, 2017, at 1, 11 (discussing the many

ways in which this phenomenon makes “unjustified

inference too easy”). Thus, if one metric for a redistricting plan does not suit the plaintiffs’ needs, they

can always try another. And if controlling for certain

traditional redistricting criteria explains the racial

breakdown of a district, an expert can simply ignore

those criteria and pin the distribution instead on race.

16

Plaintiffs in redistricting cases generally know this

phenomenon to be true, which is why they sometimes

disavow their own statistical analyses. For example,

in the ongoing litigation over Alabama’s congressional

districts, the plaintiffs—represented by many of the

same counsel representing Plaintiffs in this case—

used Dr. Kosuke Imai to perform a “race-blind simulation analysis” that would purportedly reveal the

challenged map to be race predominant. See App.30a;

Allen v. Milligan, No. 21-1086 (filed Jan. 28, 2022),

Supp. Jt. App. 52. They argued that his race-neutral

analysis “alone show[ed] that [Alabama’s] HB1 used

race as a predominant factor to crowd Black voters

into District 7” because of the statistical differences

between his simulations and the enacted plan. Mot.

for Prelim. Inj., Milligan v. Merrill, Case No. 2:21-cv1530-AMM (N.D. Ala. filed Dec. 15, 2021) (ECF No.

69); Milligan, No. 21-1086, Supp. Jt. App. 62. The

problem with this analysis was obvious from the start:

Alabama’s enacted plan, as in many States, followed

existing district lines; Dr. Imai’s simulations did not.

He drew on a blank slate. His analysis thus could not

reveal that race predominated. But when Alabama

noted that Dr. Imai’s analysis could be a race-neutral

benchmark to show that race predominated in other

plans offered by plaintiffs in that litigation—allowing

an apples-to-apples comparison because all plaintiffs’

plans were drawn on a blank slate—plaintiffs tossed

their expert to the curb. In their words, “[s]imulations

that do not match what states actually do in redistricting are neither useful nor relevant.” Appellees’ Br. at

50, Milligan, No. 21-1086 (filed July 11, 2022). In

other words, simulations are good so long as they are

faulting the State, but they can’t vindicate it. Yet

17

plaintiffs keep offering courts cherrypicked correlations, and too often, courts mistake these for causation.

B. The decision below illustrates this problem well.

The district court based its finding of racial motivation almost entirely on correlations it found suspicious. Indeed, by focusing on the statistical effects of

the new map, the district court appeared to define

“discriminatory purpose” as mere volition or awareness of consequences—the very analysis this Court

has repeatedly warned against. See, e.g., Miller, 515

U.S. at 916; Feeney, 442 U.S. at 279. According to the

district court, the General Assembly chose a map with

a partisan tilt (volition), knowing that the partisan tilt

would result in “a district in the range of 17% African

American” (awareness of consequences). App.23a.

From this alone, the court reasoned that the General

Assembly had a predetermined purpose to achieve “a

target of 17% African American population” in the district. Id. The district court never considered that the

General Assembly may have acted “in spite of,” not

“because of,” racial effects. Miller, 515 U.S. at 916

(quoting Feeney, 442 U.S. at 279). Indeed, none of the

evidence cited by the court was incompatible with the

State’s asserted purpose of creating a partisan tilt in

District 1. Yet without any evidence showing that the

General Assembly chose the map because of its effects

on race, the court was left with a singular focus on the

General Assembly’s decision to choose a map while

aware of its racial effects. This is materially indistinguishable from a disparate-impact regime for redistricting.

18

Other elements of the opinion bear this out. The

district court emphasized the mapdrawer’s awareness

of race, reasoning that his “in-depth knowledge of the

racial demographics of South Carolina” belied “his

claim that he did not consider race in drawing Congressional District No. 1.” App.29a-30a. There are at

least two problems here. First, the mapdrawer is not

the General Assembly, so his purported consideration

of race (conscious or otherwise) cannot be imputed to

it. See Brnovich v. Democratic Nat’l Comm., 141 S. Ct.

2321, 2350 (2021) (“The ‘cat’s paw’ theory has no application to legislative bodies.”). Second, “the legislature always is aware of race when it draws district

lines, just as it is aware of age, economic status, religious and political persuasion, and a variety of other

demographic factors. That sort of race consciousness

does not lead inevitably to impermissible race discrimination.” Shaw v. Reno, 509 U.S. 630, 646 (1993).

Thus, “a jurisdiction may engage in constitutional political gerrymandering, even if it so happens that the

most loyal Democrats happen to be black Democrats

and even if those responsible for drawing the district

are conscious of that fact.” Cromartie I, 526 U.S. at

542. “If the State’s goal is otherwise constitutional political gerrymandering, it is free … to achieve that

goal regardless of its awareness of its racial implications.” Bush, 517 U.S. at 968.

Apart from awareness of race, the court’s only

other “striking evidence” of racial intent was the mapdrawer’s purported “subordination of traditional districting principles” and “abandonment of his ‘least

19

change’ approach.”3 App.29a. But each of these pieces

of evidence is more readily explained by the State’s

declared partisan motivations. The State never hid

the ball: It wanted to preserve “the 6-1 Republican-toDemocratic split in House seats,” a target that “animated the General Assembly’s line-drawing decisions” and ultimately manifested in the final plan.

Doc. 323 at 16. The court never explained why this

could not be a plausible explanation for any of the evidence it found suspicious. Because political motivations could explain all the “oddities in [the] district’s

boundaries,” Cooper, 581 U.S. at 308, the State’s express political motivation—which was “entirely reasonable and certainly legitimate”—should have received a presumption of good faith, not cursory dismissal. Abbott, 138 S. Ct. at 2327. Indeed, “the legitimate

noninvidious purposes of [the] law cannot be missed.”

Feeney, 442 U.S. at 275.

The district court’s other observations similarly go

only to the effect, not the intent, of the law. For example, the court noted that the mapdrawer “acknowledged … that if there was a target for the district of

17%, the inclusion of a VTD that was 35% African

American would adversely impact the 17% objective.”

App.28a (emphasis added). But this unremarkable

3 The court also discussed “South Carolina’s legal and political

history” but did not state whether that history supported a finding of discriminatory intent. App.18a-20a. In any case, “[p]ast

discrimination cannot, in the manner of original sin, condemn

governmental action that is not itself unlawful,” and “[t]he allocation of the burden of proof and the presumption of legislative

good faith are not changed by a finding of past discrimination.”

Abbott, 138 S. Ct. at 2324.

20

(and question-begging) tautology merely acknowledged the racial effect of the selected map.

The district court next noted that 2020 census data

applied to the 2011 lines for District 1 resulted in an

African-American percentage of 17.8%, the same percentage in District 1 under the 2022 plan enacted by

the General Assembly. App.29a. In the court’s view,

this “was more than a coincidence and was accomplished only by the stark racial gerrymander.” Id. Not

only was this an odd statement (normally, a lack of

change would be the least suspicious outcome), but the

court was again merely observing an effect of the plan

and ascribing to it a racial intent. Neither numerology

nor speculation are bases for invalidating a facially

neutral law.

The district court also relied on the analysis of Dr.

Jordan Ragusa, who concluded that the racial composition of a voter tabulation district (VTD) was a better

predictor than political composition of whether the

VTD would be moved from the prior version District 1

to another district. App.31a-32a. But Dr. Ragusa’s

analysis suffered the very flaw this Court criticized in

Cromartie II: It failed to “specify whether the excluded

white-reliably-Democratic precincts were located near

enough to [the district’s] boundaries or each other for

the legislature as a practical matter to have drawn

[the district’s] boundaries to have included them,

without sacrificing other important political goals.”

532 U.S. at 247; see Doc. 323-29. Indeed, Dr. Ragusa

did not consider compactness at all. Doc. 323-30 at 3.

An analysis that suffers this fundamental flaw “offers

little insight into the legislature’s true motive,”

Cromartie II, 532 U.S. at 248, because it cannot show

21

“that the legislature could have achieved its legitimate political objectives in alternative ways that are

comparably consistent with traditional districting

principles,” id. at 258. And it certainly cannot show

that a plan is “unexplainable in terms other than

race.” Bush, 517 U.S. at 972.

Finally, the court cited the testimony of Dr. Imai,

who concluded that the 2022 plan “splits Charleston

County by placing a disproportionately large number

of black voters into District 6, while assigning relatively few voters to District 1.” App.30a (quoting PX0032, Expert Report of Kosuke Imai at 13). But Dr.

Imai admitted that he never considered partisan information in his analysis, much less controlled for it.

Doc. 386-3 at 12.

In sum, the district court’s observations about the

effect of the plan fall far short of showing that the

State acted “because of, not merely in spite of,” racial

impact. Miller, 515 U.S. at 916 (cleaned up). “Absent

a pattern as stark as that in Gomillion or Yick Wo,

impact alone is not determinative, and the Court must

look to other evidence.” Vill. of Arlington Heights v.

Metro. Hous. Dev. Corp., 429 U.S. 252, 266 (1977). Because the record lacked such additional evidence, and

this case is no Gomillion, the presumption of legislative good faith must carry the day.

CONCLUSION

The Court should note probable jurisdiction and

reverse the decision below.

22

Respectfully submitted,

STEVE MARSHALL

Alabama Attorney General

Edmund G. LaCour Jr.

Solicitor General

Counsel of Record

Bethany C. Lee

Assistant Solicitor General

OFFICE OF ALA. ATT’Y GEN.

501 Washington Avenue

Montgomery, AL 36130

(334) 242-7300

edmund.lacour@AlabamaAG.gov

MARCH 29, 2023

(Additional Counsel listed below)

23

ADDITIONAL COUNSEL

TREG TAYLOR

Alaska Attorney General

TIM GRIFFIN

Arkansas Attorney General

ASHLEY MOODY

Florida Attorney General

CHRIS CARR

Georgia Attorney General

THEODORE E. ROKITA

Indiana Attorney General

BRENNA BIRD

Iowa Attorney General

JEFF LANDRY

Louisiana Attorney General

LYNN FITCH

Mississippi Attorney General

AUSTIN KNUDSEN

Montana Attorney General

MICHAEL T. HILGERS

Nebraska Attorney General

ALAN WILSON

South Carolina Attorney General

24

JONATHAN SKRMETTI

Tennessee Attorney General

KEN PAXTON

Texas Attorney General

SEAN D. REYES

Utah Attorney General

PATRICK MORRISEY

West Virginia Attorney General