Reply Brief — Mark Brnovich, Attorney General of Arizona, et al., Petitioners v. Democratic National Committee, et al.

Supreme Court briefFeb 12, 2021

Ask Donna

What actually matters in this document.

Text

Nos. 19-1257, 1258

In the

Supreme Court of the United States

MARK BRNOVICH, IN HIS OFFICIAL CAPACITY AS

ARIZONA ATTORNEY GENERAL, ET AL.,

Petitioners,

DEMOCRATIC NATIONAL COMMITTEE, ET AL.,

Respondents.

On Writs of Certiorari to

the United States Court of Appeals for the Ninth Circuit

REPLY BRIEF FOR STATE PETITIONERS

WILLIAM S. CONSOVOY

TYLER R. GREEN

CONSOVOY MCCARTHY PLLC

1600 Wilson Blvd., Ste. 700

Arlington, VA 22209

(703) 243-9423

MARK BRNOVICH

Attorney General

JOSEPH A. KANEFIELD

Chief Deputy and

Chief of Staff

Counsel of Record

BRUNN W. ROYSDEN III

Solicitor General

DREW C. ENSIGN

Deputy Solicitor General

KATE B. SAWYER

Assistant Solicitor General

OFFICE OF THE ARIZONA

ATTORNEY GENERAL

2005 N. Central Ave.

Phoenix, AZ 85004

(602) 542-5025

joe.kanefield@azag.gov

Counsel for State Petitioners

TABLE OF CONTENTS

TABLE OF AUTHORITIES....................................... ii

INTRODUCTION ........................................................1

ARGUMENT ...............................................................3

I. Section 2 requires plaintiffs to prove that

a challenged law causes a substantial

disparity in opportunities for members of

a protected class to participate in the

political process and to affect electoral

outcomes. ...........................................................3

A. Respondents’

proposed

standard

breaks from §2’s text and precedent..........3

B. Respondents’ view of §2 raises

serious constitutional concerns. .............. 11

II. Arizona’s out-of-precinct policy and

ballot-collection law do not violate §2. ...........13

A. Arizona’s out-of-precinct policy does

not violate §2. ........................................... 13

B. The ballot-collection law does not

violate §2................................................... 17

III. The

Ninth

Circuit’s

intentionaldiscrimination holding is untenable. .............20

IV. The State has standing to seek review of

the out-of-precinct policy. ...............................22

V. The Court should reverse the judgment

without

remanding

for

further

proceedings. .....................................................23

CONCLUSION .......................................................... 24

TABLE OF AUTHORITIES

CASES

Anderson v. City of Bessemer City,

470 U.S. 564 (1985) ............................................... 21

Arizonans for Official English v. Arizona,

520 U.S. 43 (1997) ................................................. 22

Bartlett v. Strickland,

556 U.S. 1 (2009) ................................................... 11

Cooper v. Harris,

137 S. Ct. 1455 (2017) ........................................... 21

Crawford v. Marion County Election Board,

533 U.S. 181 (2008) ............................................... 22

Frank v. Walker,

768 F.3d 744 (7th Cir. 2014) ..................... 1, 4, 7, 14

Greater Birmingham Ministries v. Secretary of State

for Alabama,

966 F.3d 1202 (11th Cir. 2020) ........................... 1, 7

Hunt v. Cromartie,

526 U.S. 541 (1999) ............................................... 21

League of United Latin American Citizens v. Perry,

548 U.S. 399 (2006) .................................................2

Lee v. Virginia State Board of Elections,

843 F.3d 592 (4th Cir. 2016) ............................... 4, 7

Luft v. Evers,

963 F.3d 665 (7th Cir. 2020) ............................... 1, 7

Miller v. Johnson,

515 U.S. 900 (1995) ............................................... 13

Northeast Ohio Coalition for the Homeless v. Husted,

837 F.3d 612 (6th Cir. 2016) ............................... 1, 4

Northwest Austin Municipal Utility District

No. 1 v. Holder,

557 U.S. 193 (2009) ........................................... 2, 11

iii

TABLE OF AUTHORITIES—Continued

Ohio Democratic Party v. Husted,

834 F.3d 620 (6th Cir. 2016) ...................................7

Santa Rita Mining Company v. Department of

Property Valuation,

530 P.2d 360 (Ariz. 1975) ...................................... 22

Shelby County. v. Holder,

570 U.S. 529 (2013) ........................................... 2, 11

Thornburg v. Gingles,

478 U.S. 30 (1986) ............................................. 4, 10

Veasey v. Abbott,

830 F.3d 216 (5th Cir. 2016) ............................. 1, 16

Virginia House of Delegates v. Bethune-Hill,

139 S. Ct. 1945 (2019) ........................................... 22

STATUTES

52 U.S.C. §10301 ............................................... passim

A.R.S. §16-411(B)(4) .................................................. 15

A.R.S. §16-452(B) ...................................................... 23

A.R.S. §41-193(A)(3) .................................................. 22

OTHER AUTHORITIES

Arizona Population and Vital Statistics, Arizona’s

Economy, https://www.azeconomy.org/arizonapopulation/ ............................................................. 16

Carter Center Statement On Voting By Mail For 2020

U.S. Elections, Carter Center (May 6, 2020),

https://www.cartercenter.org/news/pr/2020/

united-states-050620.html .................................... 19

City Reminds Voters that City Elections Staff Do Not

Pick Up Ballots at Residences or Ask How

Residents Voted, City of Phoenix (Aug. 12, 2015),

https://www.phoenix.gov/news/ cityclerk/900....... 19

TABLE OF AUTHORITIES—Continued

Nicholas O. Stephanopoulos, Disparate Impact,

Unified Law,

128 Yale L.J. 1566 (2019) ........................................6

Urban Percentage of the Population for States,

Historical, Iowa Community Indicators Program,

https://www.icip.iastate.edu/tables/population/urb

an-pct-states .......................................................... 16

INTRODUCTION

Section 2 of the Voting Rights Act requires votedenial plaintiffs to prove a substantial disparate impact caused by the challenged state law. Nothing

about that interpretation is novel; it is the law in at

least three circuits. See Frank v. Walker, 768 F.3d

744 (7th Cir. 2014) (Easterbrook, J.); Luft v. Evers,

963 F.3d 665 (7th Cir. 2020) (Easterbrook, J.); Ne.

Ohio Coal. for the Homeless v. Husted, 837 F.3d 612,

628-629 (6th Cir. 2016); Greater Birmingham Ministries v. Sec’y of State for Ala., 966 F.3d 1202 (11th

Cir. 2020); see also Veasey v. Abbott, 830 F.3d 216,

310-312 (5th Cir. 2016) (en banc) (Jones, J., concurring in part and dissenting in part).

Rather than answering those arguments, Respondents ignore them. The DNC twice contends that

“[n]o court has ever required plaintiffs to make a

threshold showing of ‘substantial’ disparity in a §2

vote-denial case.” DNC Br. 26; accord DNC Br. 20

(“Neither this Court nor any other has ever required

this[.]”). No court, that is, except those cited above—

which the State Petitioners cited so frequently in

their opening brief that those cases appear as passim

cites. Far from “invent[ing]” those requirements

“from whole cloth” (Hobbs Br. 16), Arizona cuts from

the §2 tapestry the courts of appeals have woven.

As to §2’s causation requirement, Secretary Hobbs

admits (at 13) that “Section 2 requires … but-for

causation,” but ignores that the Ninth Circuit required no such showing. The DNC, in turn, ignores

the concept of but-for causation and instead substitutes the “objective Senate Factors.” DNC Br. 32.

That follows the discredited approach of legislative

history trumping text—indeed, what even Secretary

Hobbs admits §2’s text demands.

Respondents also discount the severe constitutional

concerns that their interpretations raise. Indeed—

despite this Court’s teachings in Northwest Austin

and Shelby County—the DNC and Secretary Hobbs

both disclaim any requirements of congruence and

proportionality for legislation enacted under §2 of the

Fifteenth Amendment, even though those requirements apply to legislation enacted under §5 of the

Fourteenth Amendment. But because both sections

have virtually identical text, history, and context,

that standard governs legislation enacted under each

section. Beyond that, Respondents’ interpretations

necessarily require overtly race-conscious decisionmaking for election laws, resulting in §2 forcing

state legislatures to persist in the “sordid business …

[of] divvying us up by race.” League of United Latin

Am. Citizens v. Perry, 548 U.S. 399, 511 (2006) (Roberts, C.J., concurring in part). Section 2’s text does

not require any of this.

Applying §2 as Congress enacted it, Plaintiffs’

claims fail. After a ten-day trial, Plaintiffs’ evidence

established that the out-of-precinct policy had no impact on 99.9% of non-minority voters or 99.8% of minority voters. See State Br. 35. And Plaintiffs’ evidence of the ballot-collection law’s impact was “circumstantial and anecdotal.” JA 324. Neither establishes a substantial disparate impact sufficient to

trigger §2 liability. What’s more, Plaintiffs’ evidence

did not establish that the out-of-precinct policy or the

ballot-collection law were but-for causes of any impact. Plaintiffs’ failures of proof on both showings for

both claims require reversing the en banc majority’s

judgment and directing judgment for Defendants.

ARGUMENT

I. Section 2 requires plaintiffs to prove that a

challenged law causes a substantial disparity in opportunities for members of a protected class to participate in the political

process and to affect electoral outcomes.

The parties appear to agree that a §2 vote-denial

claim should be evaluated under a two-step inquiry.

They also agree, in broad strokes, that step one requires a showing of a disparate impact and that step

two requires proof of causation.

But that is where the agreement ends. Respondents’ interpretation allows federal courts to rewrite

state election laws under §2 based on anything more

than a de minimis disparate impact linked to social

or historical conditions that the State itself did not

cause. This view stretches §2 beyond its textual

bounds and raises grave constitutional concerns.

The Court should reject it and adopt the State Petitioners’ text-based interpretation instead.

Respondents’ proposed standard breaks

from §2’s text and precedent.

1. As the State Petitioners explained (at 19-23),

§2’s first step imposes liability only for substantial

disparate impacts. That requirement arises from

what §2(a) forbids—state laws that “den[y] or

abridg[e]” the right to vote, 52 U.S.C. §10301(a)—

and from what §2(b) mandates: equal “opportunity”

for all voters, regardless of race, to “participate in the

political process and to elect representatives of their

choice,” id. §10301(b). Indeed, as the State Petitioners showed, no fewer than three circuits have recognized the substantiality requirement this text imposes. In the Sixth Circuit, “[a] law cannot disparately

impact minority voters if its impact is insignificant

to begin with.” Ne. Ohio Coal., 837 F.3d at 628 (emphasis added). In the Fourth Circuit, a law that

“does not impose a substantial burden” does not trigger §2, since it is an “unjustified leap from the disparate inconveniences that voters face when voting to

the denial or abridgement of the right to vote.” Lee v.

Va. State Bd. of Elections, 843 F.3d 592, 600-601 (4th

Cir. 2016). And the Seventh Circuit does not allow

insubstantial differentials—such as “if whites are 2%

more likely to register than are blacks”—to state a §2

claim because that would “sweep[] away almost all

registration and voting rules.” Frank, 768 F.3d at

754; see also id. at 752 n.3 (explaining that a §2

claim would fail “if 99.9% of whites had photo IDs,

and 99.7% of blacks did”). This all comports with

this Court’s teaching that a disparate impact establishes “less opportunity … to participate in the political process,” §10301(b), only when the disparity between “minority group members” and other members

of the electorate is “substantial.” Thornburg v. Gingles, 478 U.S. 30, 48 n.15 (1986).

Respondents answer those arguments by ignoring

them. The DNC contends (at 26) that “[n]o court has

ever required plaintiffs to make a threshold showing

of ‘substantial’ disparity in a §2 vote-denial case” and

repeats (at 20) that “[n]either this Court nor any other has ever required this[.]” (Emphasis added).

Never mind that Northeast Ohio Coalition, Lee, and

Frank did just that. The DNC, however, never cites

Northeast Ohio Coalition, and cites Lee and Frank

only for their outcomes without answering their reasoning. Yet that is a step closer than Secretary

Hobbs, who does not cite any of those cases at all.

The DNC also tries to limit (at 26 n.5) Gingles’s

statement that a cognizable burden must be “substantial” to the context of “multimember district vote

dilution cases.” But the DNC’s central premise is

that a single “test … applie[s] to vote-denial and

vote-dilution cases alike.” DNC Br. 22; accord DNC

Br. 5-6, 33. Taking the DNC at its word requires applying Gingles’s substantiality test to every §2 claim

“alike.”

Nor does the DNC or Secretary Hobbs meaningfully engage with the State Petitioners’ argument (at

22) that §2 requires an impact to minority voters’

“opportunity … to elect representatives of their

choice.” Instead, the DNC argues (at 27) that §2 liability should attach even when “the number of voters

affected is too small to affect the outcome of an election[.]” But Congress’s use of the conjunctive “and”

requires impact both to opportunity to participate

and to ability “to elect representatives of their

choice.” §10301. The DNC’s arguments ignore that

conjunctive construction. See DNC Br. 27-28.

Secretary Hobbs, in turn, suggests (at 30) that “the

size of a disparity may bear on whether causation

can be established, or whether the totality of circumstances shows a discriminatory result.” (Emphasis

added). Secretary Hobbs’s acknowledgment that the

size of a disparity can “bear” on whether there is a

“discriminatory result”—and presumably answer

that question in the negative where it is insubstantial—buttresses the State Petitioners’ (and the

Fourth, Sixth, and Seventh Circuits’) position that §2

liability arises only for substantial disparate im-

pacts. 1 In any event, Secretary Hobbs supplies no

textual basis for considering the size of the disparity

only at step two (causation) and not also at step one

(impact).

Scholars have already confirmed the dangers of Respondents’ diluted step-one standard: In courts that

have adopted it, the step-one outcome is a “nearperfect” predictor of ultimate outcome. See Nicholas

O. Stephanopoulos, Disparate Impact, Unified Law,

128 Yale L.J. 1566, 1592 (2019). Except for a single

district court case involving “idiosyncratic preferences,” in “every other case, if a court discerned a

disparate impact, it also managed to link that impact

to past and present discrimination, as illuminated by

the Senate factors.” Id. at 1591-1592.

This empirical data belies Respondents’ contentions that their proposed step-two standard will address the practical and constitutional problems inherent in their approach. Instead, the data confirms

that Respondents’ second step is toothless in theory

and in practice.

2. The State Petitioners contend (at 20-22) that §2

guarantees what it says—equality of “opportunity …

to participate in the political process.” §10301(b); see

also Ohio, et al., Amicus 11-15; Pacific Legal Amicus

11. Respondents do not quibble with those statutory

The United States similarly argues (at 29) that “[a] single

polling-place clerk violates Section 2 by turning away only minority voters whether or not their votes would swing the election.” No doubt that conduct would violate §2—and the Fifteenth Amendment—as intentional discrimination. But it

would not support a claim under §2’s results test absent proof of

actual impact to “opportunity … to elect representatives of their

choice.” §10301(b).

words, but they interpret them to require States to

adopt processes guaranteeing voters of all races

equal outcomes. Put differently, Respondents argue

that a §2 claim arises when minority voters use equal

opportunities at differential rates than non-minority

voters.

That argument does not state a cognizable §2 claim

under the tests recognized by the Fourth, Sixth, Seventh, and Eleventh Circuits. See Greater Birmingham Ministries, 966 F.3d at 1232-1235; Lee, 843 F.3d

at 600-601; Frank, 768 F.3d at 753-755; Luft, 963

F.3d at 672; Ohio Democratic Party v. Husted, 834

F.3d 620, 637-638 (6th Cir. 2016). Respondents have

no response to any of these cases, largely or entirely

ignoring them.

3. By omitting any substantiality requirement and

permitting violations based on disparate use, Respondents’ step one would cripple States’ ability to

manage their election codes. See, e.g., Haahr, et al.,

Amicus; Wis. Legislators Amicus. Respondents’ stepone standard invites §2 liability for the most minute

disparity; it “would mean that every polling place

would need to be precisely located such that no group

had to spend more time traveling to vote than did

any other.” Lee, 843 F.3d at 601.

Plaintiffs do not answer this concern. Indeed, they

do not even acknowledge Lee’s reasoning. And this

danger is particularly acute as both Respondents

would forbid courts resolving §2 challenges to consider other existing ways voters can participate in

the process, contending that “a state cannot wave

away a closed door by pointing to an open window.”

DNC Br. 29 n.6; accord Hobbs Br. 22-23. So much

for the “totality of circumstances” inquiry Congress

mandates. §10301(b).

It is not hard to envision how Respondents’ onlysome-of-the-circumstances standard will hamstring

State election officials. Suppose a law shifting polling hours by one hour increased Hispanic voter turnout by 2%, decreased African-American voter turnout

by 2%, and left white voter turnout flat. That result

would violate step one of Respondents’ test because

of the disparate (though not substantial) burden on

African-American voters relative to white voters.

But reverting to the prior law would (under Respondents’ theory) raise a separate §2 claim because

of the now-known corresponding disparate impact on

Hispanic voters. A State would potentially face inescapable §2 liability whenever a law benefits one minority group but hinders another—a scenario that

will be far more common if non-significant disparities sufficed.

Even policies increasing opportunity for all would

potentially violate §2. Suppose, for example, a legislature adopts no-excuse voting by mail, which increases African-American, Hispanic, and white voter

turnout by 6% and Native American voter turnout by

3%. That too would violate §2 under Respondents’

theory because of the disparate results for Native

American voters.

4. No party disagrees with the State Petitioners’

view (at 23-24)—consistent with every circuit decision addressing the question—that §2 plaintiffs must

prove causation. But Respondents’ approaches to

proving causation do violence to §2’s text.

Start with Secretary Hobbs. She admits (at 18-19)

that §2 “requires … but-for causation.” But Secre-

tary Hobbs never shows how the en banc Ninth Circuit applied any but-for causation test. That’s because it did not: The en banc majority allowed plaintiffs to prove that the disparate impact was either

“caused by or linked to ‘social and historical conditions[.]’” JA 659, 671 (emphasis added) (citation

omitted). The Ninth Circuit thus allowed an impact’s alleged link to “social and historical conditions” to substitute for the but-for causation that

Secretary Hobbs admits §2’s text demands.

Unlike Secretary Hobbs, who embraces but-for causation, the DNC ignores it by refusing even to address whether §2 requires but-for causation. Instead, the DNC follows the en banc majority’s lead

and latches onto the Gingles factors as the sole

method to determine causation. That approach accurately describes what happened below, but has little else to recommend it. Consider two problems it

raises. First, the DNC’s approach does not answer

whether §2’s text itself requires but-for causation—

what Secretary Hobbs calls (at 18) “the ‘default’ or

‘background’ rule[.]” Second, the DNC contends (at

26 n.5) that some aspects of Gingles should be “limited to multimember district vote dilution cases,” but

never explains why the Court should import Gingles’s apparent presumption of causation into the

vote-denial context. There is a straightforward answer why it should not: In vote-dilution cases, the

government itself draws district lines, so causation

will rarely be contestable. But a disparate-impact

analysis for vote-denial claims must account for a

host of voter- and group-specific factors—the “totality

of circumstances,” §10301(b)—untethered to State

action. That makes proof that the challenged law

itself caused the disparate impact an indispensable

textual requirement of vote-denial claims.

5. Two brief responses to the United States. First,

the United States contends (at 21-24) that §2 requires proof of proximate causation. In the State Petitioners’ view, this Court need not decide whether §2

requires proximate causation or but-for causation because Plaintiffs did not prove, and the Ninth Circuit

did not require, even the lower of those two showings. This Court thus could reserve that question for

a future case and reverse the judgment for Plaintiffs’

failure to prove even but-for causation. If the Court

decides to reach this question, however, the State Petitioners agree with the United States that §2 requires a showing of proximate causation and urge

the Court to adopt that standard for the reasons the

United States explains. See U.S. Br. 21-24.

Second, the State Petitioners agree with the United

States (at 24-25) that States must retain the ability

to present their governmental interests when defending against vote-denial claims. But whether that occurs as part of a “totality of circumstances” inquiry

or (as the United States proposes) as a separate third

step is a distinction without a difference. In either

event, plaintiffs must bear the burden at all phases

of §2 litigation, since §2 cannot be read to impose a

burden-shifting framework.

Nor does “tenuous[ness],” Gingles, 478 U.S. at 36-37, do any real

work if a state chooses to present its interests supporting a challenged law; that factor has become

merely an invitation for courts to substitute their

judgment for the judgments of legislatures. Supra

at 6 (discussing toothlessness of step two when based

on Gingles Factors).

Respondents’ view of §2 raises serious

constitutional concerns.

The State Petitioners explained (at 24-28) how the

en banc court’s §2 standard raises two serious constitutional concerns—it (1) makes §2 exceed Congress’s

powers to enforce the Reconstruction Amendments,

and (2) violates the Fourteenth Amendment by making race the predominant factor in fashioning election laws. See also Ohio, et al., Amicus 27-31. Respondents’ briefs do nothing to dispel those severe

concerns.

1. Respondents answer principally by fighting the

existence of meaningful constitutional constraints.

In their telling, Congress’s powers under §2 of the

Fifteenth Amendment—unlike under §5 of the Fourteenth Amendment—authorize any law that can be

deemed a rational use of that power. DNC Br. 23,

47; Hobbs Br. 34-35. The DNC further appears to

contend (at 49-50) that “religious-liberty” rights occupy a lower tier than voting rights.

Respondents’ premises fail. The text, history, and

context of the two enforcement sections are virtually

identical. Given those shared characteristics, Respondents offer no valid reason not to give “both

Amendments” the same readings. Shelby Cty. v.

Holder, 570 U.S. 529, 542 n.1 (2013); Nw. Austin

Mun. Util. Dist. No. 1 v. Holder, 557 U.S. 193, 204

(2009). And the Court already recognized in the context of §2 that race-based inquiries and decisionmaking “raise[] serious constitutional questions.”

Bartlett v. Strickland, 556 U.S. 1, 18, 21 (2009) (plurality op.). Respondents neither acknowledge nor

answer Strickland’s concerns. Nor does the DNC ex-

plain how the Constitution places religious-liberty

rights on a lower tier than voting rights.

Respondents also suggest that Congress has power

under the Elections Clause to give §2 the reach they

propose. Hobbs Br. 35-36; DNC Br. 53. But the

Elections Clause allows Congress to prescribe procedures only for federal elections, and §2 applies to

State procedures for all elections. So §2 cannot be

justified as an exercise of Congress’s Election Clause

powers. Nor is the Elections Clause a license for

Congress to require State voting processes that violate the equal treatment guaranteed by the first sections of the Fourteenth and Fifteenth Amendments.

2. Respondents also fight the premise that overtly

race-conscious decisions by the government raise

constitutional concerns. Secretary Hobbs (at 35)

even accuses “Petitioners” of “invent[ing]” those concerns. So much for this Court’s extensive precedents

also raising them. See State Br. 24-28. Secretary

Hobbs simply refuses to explain why her interpretation would not raise the very concerns this Court has

repeatedly identified but she ignores.

Similarly, the DNC seems to see the race-conscious

consequences of its interpretation as a feature rather

than a constitutional bug. It suggests (at 54, 56) that

“the Fourteenth Amendment does not command race

blindness in government decision-making” and “that

government need not be race-blind.” To be sure, the

Fourteenth Amendment sometimes permits racebased decision-making in extraordinary circumstances when the government satisfies strict scrutiny. But this Court’s precedents make clear that

race-based policies must be a last resort, not a first

choice. See, e.g., Miller v. Johnson, 515 U.S. 900,

904, 916 (1995). The ubiquitous governmental consideration of race resulting from the DNC’s arguments turns those principles on their heads.

3. Respondents end by invoking step two of their

test to assuage potential constitutional concerns.

Hobbs Br. 34; DNC Br. 58. But Respondents’ diluted

step-two causation analysis cannot bear the weight

they place on it. Instead, as explained, empirical evidence shows that—in all but a single case—step two

rubber stamps whatever the court found at step one.

Supra at 6. Respondents’ causation analysis thus

performs no meaningful screening function.

Because the supposed protections of Respondents’

second step are illusory, they cannot dispel the constitutional doubts here.

II.

Arizona’s out-of-precinct policy and ballotcollection law do not violate §2.

Arizona’s out-of-precinct policy does not

violate §2.

Plaintiffs’ claims fail both parts of the §2 inquiry.

1. Plaintiffs’ §2 challenge to the out-of-precinct policy fails at step one for three reasons. First, the outof-precinct policy is race neutral; it gives all voters

an equal opportunity to vote in their correct precinct

on Election Day, or otherwise to choose from Arizona’s robust set of early voting options, and it requires

discarding the votes of all out-of-precinct voters, irrespective of race. State Br. 16-17, 33-40; U.S. Br. 2526. Although Respondents point to differential outcomes in minority-group usage of in-person precincts,

they offer no evidence demonstrating unequal opportunity.

Second, Plaintiffs failed to prove a substantial disparity, particularly when reviewing Arizona’s elections system as a whole. In the 2016 general election, the out-of-precinct policy affected about 0.15%

of Arizona voters. State Br. 9, 17, 34-37. Though

Plaintiffs proved “minorities are over-represented

among the small number of voters casting [out-ofprecinct] ballots,” JA 332, the district court found

that “such a small and ever-decreasing fraction of the

overall votes cast” meant that Arizona’s policy “has

no meaningfully disparate impact.” JA 334. And as

the United States correctly contends, the Ninth Circuit erroneously fixated only on in-person, electionday votes to increase the disparity to 99% vs. 99.5%

(U.S. Br. 20-21, 25-27), which is still not substantial

for purposes of §2 in any event.

Beyond that, Respondents’ continued insistence

(DNC Br. 9; Hobbs Br. 39) that the out-of-precinct

policy affected twice as many minorities as nonminorities would be a statistically accurate description if the rates were 99.99% of non-minorities vs.

99.98% of minorities, or even 99.999% of nonminorities vs. 99.998% of minorities—“a misuse of

data” that confirms why comparing percentages

alone without considering absolute magnitudes cannot establish a substantial disparity. Frank, 768

F.3d at 752 n.3.

Third, the DNC’s repeated reliance (at 21, 31) on

Florida’s 537-vote margin in the 2000 presidential

race does nothing to satisfy their burden. For one

thing, the 2000 election could not have informed

what Congress meant when it passed a statute

eighteen years earlier in 1982. Even so, close elections existed before 2000, yet Congress still chose

language mandating impacts on electoral outcomes—

which even the DNC acknowledges requires “substantial” impacts for vote-dilution claims to be cognizable. Supra at 5-6. So too for vote-denial claims.

2. At step two, Plaintiffs failed to prove even butfor causation. Indeed, the DNC does not allege that

it did. Supra at 9. In any event, Plaintiffs never

proved that the out-of-precinct policy actually caused

voters of any race to vote in the wrong precinct.

State Br. 17, 37-38; see also JA 708 (O’Scannlain, J.,

dissenting); U.S. Br. 6, 26.

As a result, Plaintiffs’ challenge is really to the precinct system itself or to county-level officials’ actions

regarding polling places. But Plaintiffs did not bring

those challenges, which would focus on county defendants, involve different evidence, and require different remedies. Indeed, the DNC’s proposal to

count some number of races at the top of the ballot

suggests that the DNC seeks not to cure an alleged

§2 violation but only to benefit itself politically in

top-of-the-ticket races. In any event, that remedy—

counting an arbitrary number of races at the top of

the ballot—is not a cure for any violation based on

the location or movement of polling places. Secretary

Hobbs, in turn, suggests (at 44) ballot centers as the

remedy. But counties have the option of choosing

ballot centers under current statutory law. Ariz.

Rev. Stat. (“A.R.S.”) §16-411(B)(4).

Even if the DNC were correct that it could sidestep

ordinary causation by relying on the Gingles factors,

its arguments still fail. While the DNC hints (at 40

n.8) that the Ninth Circuit “recognized recent examples of discrimination,” it identifies no errors in the

State Petitioners’ demonstration (at 39) that those

recent examples were “scant and equivocal,” and ex-

clusively involved actions by non-parties. That the

en banc majority allowed that type of evidence to establish causation exemplifies how readily the Gingles

factors “lend themselves to manipulation.” Veasey,

830 F.3d at 327 (Elrod, J., concurring in part and

dissenting in part). 2

3. If the State’s justification bears on the challenged law’s validity, there is a strong justification

here. The United States (at 4, 27, 29-30) well summarizes the important state interests that a precinctbased system serves, including allowing ballots to

list the correct races and facilitating state efforts to

monitor votes and prevent election fraud. In fact,

Secretary Hobbs defended the out-of-precinct policy

below largely on those grounds and does not rebut

them here. Hobbs Br. 43 (discussing instead the

administrative feasibility of partially counting very

low numbers of out-of-precinct ballots). And while

she now (at 43-44) makes conclusory assertions

about extra-record conversations she had with unnamed county officials, none of that was evidence before the district court.

The DNC’s characterizations of the record are misleading in

any case. As an example, the DNC contends (at 9) that “Arizona is heavily rural.” But with more than 80% of its population

in just the Phoenix and Tucson metropolitan areas alone, see

Arizona Population and Vital Statistics, Arizona’s Economy,

https://www.azeconomy.org/arizona-population/ (last visited

Feb. 11, 2021), Arizona is actually among the ten least rural

states in the country, see Urban Percentage of the Population for

States, Historical, Iowa Community Indicators Program,

https://www.icip.iastate.edu/tables/population/urban-pct-states

(last visited Feb. 11, 2021).

The ballot-collection law does not violate

§2.

1. The en banc majority erred in reversing the district court’s finding that Plaintiffs had not satisfied

step one by showing any “meaningful inequality in

the electoral opportunities of minorities as compared

to nonminorities.” JA 331; see U.S. Br. 27. As the

district court correctly found—and Respondents fail

to contest with anything but conclusory assertions—

Plaintiffs presented “no quantitative or statistical

evidence comparing the proportion that is minority

versus nonminority.” JA 321. Even so, the district

court thoroughly considered the limited circumstantial and anecdotal evidence Plaintiffs introduced at

trial, JA 324, and still found insufficient evidence to

show a substantial disparate impact. JA 331.

Both Secretary Hobbs and the DNC admit that a

“bare statistical showing” is not sufficient. DNC Br.

27; Hobbs Br. 33. A case-in-chief lacking that bare

showing necessarily fails. And out of the sparse anecdotal evidence Plaintiffs offered, much of it pointed

to partisan, not racial, motivations. See JA 329-330

(“Within the last decade, ballot collection has become

a larger part of the Democratic Party’s [get-out-thevote] strategy … [while] the Republican Party has

not significantly engaged in ballot collection as a

[get-out-the-vote] strategy.”).

2. Plaintiffs also failed to show causation. Respondents do not even try to show that but-for causation was proven here. Instead, like the en banc majority, they rely (DNC Br. 32-34) only on a linkage to

“social and historical” conditions—their only option,

since “no individual voter testified that H.B. 2023’s

limitations on who may collect an early ballot would

make it significantly more difficult to vote.” JA 331.

The recent ballot-collection fraud that upended an

election in North Carolina only confirms the wisdom

of Arizona’s choice to adopt the ballot-collection law

as a prophylactic measure against fraud. See State

Br. 44. The DNC apparently views the State Petitioners’ desire to avoid a similar fate as unanswerable, since it offers no answer; it merely persists in

claiming that the State’s fraud concerns are invented. DNC Br. 23-24.

Secretary Hobbs, in contrast, contends (at 50) that

North Carolina evidence is irrelevant because such

fraud “would have been illegal under Arizona laws

independent of H.B. 2023.” But Secretary Hobbs’s

contention ignores the very purpose of prophylactic

measures. Consider another real-life example: The

Food and Drug Administration and the States limit

purchases of cold medicine containing pseudoephedrine not because they believe the medicines themselves will be abused, but instead because buyers can

use those medicines to produce methamphetamine.

And it is no answer to say that States and the Federal Government can outlaw manufacture and possession of methamphetamine (as they do). That is hard

to detect and enforce, so governments prophylactically regulate sales of cold medication. Arizona’s approach to ballot collection is no different, adopting

both a prophylactic ban on conduct that heightens

the risk of fraud and a ban on the underlying fraud

itself.

On that score, Arizona’s ballot-collection law is virtually identical to the Carter-Baker Commission’s

recommendations. State Br. 8. The very purpose of

the Carter-Baker Commission’s recommendation was

prophylactically to ban practices that risk such

fraud. Secretary Hobbs does not acknowledge that

the State’s law is virtually identical to what CarterBaker actually recommended, instead focusing (at

46-47) on what Secretary Hobbs thought the Commission “meant to” do. The DNC, in turn, reaches

outside the record (at 61 & n.19) to suggest that

President Carter effectively recanted his recommendations. Not so. President Carter actually said that

voting by mail works “where safeguards for ballot

integrity are in place.” 3 That is exactly what Arizona’s ballot-collection law is designed to accomplish.

This is not a hypothetical concern for Arizona.

Less than a year before the legislature passed the

ballot-collection law, the City of Phoenix had to issue

a warning following reports that “individuals have

come to [voters’] doors stating that they work for city

elections and that they are there to pick up the voter’s early ballot to return it.” 4 The ballot-collection

law is an appropriate way to help prevent that type

of actual misconduct from recurring and preserve the

integrity of the State’s elections.

3 The statement further elaborated that the “main recommendations on vote-by-mail and absentee voting” included “eliminat[ing] the practice of allowing candidates or party workers to

pick up and deliver absentee ballots.” See Carter Center Statement On Voting By Mail For 2020 U.S. Elections, Carter Center

(May 6, 2020), https://www.cartercenter.org/news/pr/2020/

united-states-050620.html.

City Reminds Voters that City Elections Staff Do Not Pick Up

Ballots at Residences or Ask How Residents Voted, City of

Phoenix (Aug. 12, 2015), https://www.phoenix.gov/news/

cityclerk/900.

III. The

Ninth

Circuit’s

intentionaldiscrimination holding is untenable.

1. The Ninth Circuit’s intentional-discrimination

holding cannot fairly be read to rest on anything but

the “cat’s paw” theory. The en banc majority called

“the majority” of legislators “sincere in their belief

that ballot collection increased the risk of early voting fraud, and that H.B. 2023 was a necessary

prophylactic measure to bring early mail ballot security in line with in-person voting.” JA 350; see also

JA 357. Those concededly sincere beliefs preclude

any finding of intentional discrimination without

some way to impute improper intent to those legislators. The Ninth Circuit’s way to do so was the cat’spaw theory, so its holding rises or falls on that theory’s applicability.

The cat’s-paw theory does not—and cannot—apply

here. Some type of “agency” relationship must exist

for the “cat’s paw” doctrine to do any work—and no

agency relationship exists among co-equal, independent legislators. State Br. 45-47; U.S. Br. 33-34.

Neither Respondent seriously disputes that view of

the cat’s-paw theory. Nor do they try to supply a

doctrinal or factual basis to apply the theory to coequal, independent legislators. Instead, Secretary

Hobbs posits (at 48) that the “roots” of the theory

“are irrelevant” because “legislative history was just

one of many reasons the en banc court deemed the

law racially motivated.” But the en banc majority

specifically said that a majority of legislators were

acting on “sincere … belief[s],” requiring the court

instead to rely almost entirely on then-Senator

Shooter’s legislative efforts under the cat’s-paw theory. JA 674-679. In any case, it is not the historical

“roots” of the cat’s-paw theory that are at issue but

rather its doctrinal coherence in this context—which

simply does not exist absent an agency relationship.

The DNC’s arguments are no more persuasive.

The DNC asserts (at 59) that applying the cat’s-paw

theory “was apt” because “racially-motivated allegations [were] peddled by influential actors [that]

tainted the whole process.” But Plaintiffs offered no

proof that those actors were influential; that is the

inquiry the Ninth Circuit used the cat’s-paw theory

to short-circuit. And Shooter’s lack of influence with

his colleagues is confirmed by the later overwhelming vote to expel him, State Br. 47, which neither Respondent mentions.

2. Any non-cat’s-paw-theory reasoning the Ninth

Circuit employed to support this holding is untenable. As the State Petitioners explained (at 49-51),

the en banc majority “overstep[ped] the bounds” of

clear-error review for pure factual findings, and this

Court can reverse on that basis alone. See Anderson

v. City of Bessemer City, 470 U.S. 564, 575 (1985).

The en banc majority also erred by conflating the

district court’s findings of partisan motives with racial motives. JA 717-718 (O’Scannlain, J., dissenting) (“The majority simply concludes that such finding shows racially discriminatory intent as a motivating factor.”). Respondents argue that this was

permissible because “‘racial identification is highly

correlated with political affiliation.’” Hobbs Br. 50

(citation omitted); accord DNC Br. 60. But this

Court has expressly rejected such conflation. See,

e.g., Cooper v. Harris, 137 S. Ct. 1455, 1473 (2017);

Hunt v. Cromartie, 526 U.S. 541, 551-552 (1999).

3. The en banc majority further erred by concluding that Arizona legislators must have acted with racially tainted motives when enacting prophylactic

measures because those legislators lacked evidence

of fraud in Arizona. See JA 718 (O’Scannlain, J., dissenting). But Crawford v. Marion County Election

Board recognized the legitimacy of preventive legislative enactments to protect election integrity. See

533 U.S. 181, 194-196 (2008). Nothing in Crawford’s

reasoning is inapplicable to §2.

IV. The State has standing to seek review of

the out-of-precinct policy.

Secretary Hobbs’s contention that the State Petitioners lack standing to seek review of the out-ofprecinct policy does not accurately state Arizona law.

Her argument ignores that the State of Arizona was

granted intervention in the Ninth Circuit—a decision

she did not challenge by cross-petition here. And unlike the single house in Virginia House of Delegates

v. Bethune-Hill, 139 S. Ct. 1945, 1956 (2019), the Arizona Attorney General has explicit statutory authority under Arizona law to “[r]epresent the state in any

action in a federal court.” A.R.S. §41-193(A)(3). This

Court has previously recognized as much: “Under Arizona law, the State Attorney General represents the

State in federal court.” Arizonans for Official English v. Arizona, 520 U.S. 43, 51 n.4 (1997). Secretary

Hobbs does not acknowledge that decision or A.R.S.

§41-193(A)(3), much less distinguish them or explain

why they do not apply here.

Instead, Secretary Hobbs relies almost entirely on

Santa Rita Mining Co. v. Department of Property

Valuation, 530 P.2d 360 (Ariz. 1975). Santa Rita,

however, involved a prior attorney general’s attempt

to appeal in state court in the name of a state agency

over the agency’s objection. Here, the Attorney General is not attempting to seek review in the Secretary’s name, but rather in his own name and, under

A.R.S. §41-193(A)(3), the State of Arizona’s name. 5

Santa Rita is thus inapposite.

The Court should reverse the judgment

without remanding for further proceedings.

To give Arizona’s legislature the certainty it deserves, this Court should direct entry of judgment for

Defendants. This case has already gone through a

full trial generating extensive factual findings—and

Plaintiffs failed to introduce evidence proving their

§2 claims. See Section II, supra.

The district court already performed the “intensely

local” inquiry that the DNC and Secretary Hobbs

acknowledge §2 requires. The results of that inquiry

after a full trial are not close: The district court

found (1) that the out-of-precinct policy did not affect

99.9% of non-minority voters vs. 99.8% of minority

voters, and (2) a total absence of quantitative proof

on the ballot-collection law’s impact. Because this

post-trial factual record supports only a single outcome if §2 is properly applied, a remand would serve

no purpose. Similarly, the bare en banc majority’s

discriminatory-intent holding is premised on obvious

legal error that can and should be corrected now.

Secretary Hobbs’s reliance on the election procedure manual

(“EPM”) is both irrelevant and incorrect. The out-of-precinct

policy arises from statute, not merely the EPM. State Br. 7.

Nor does Secretary Hobbs have authority to amend the EPM

unilaterally; instead, she needs the concurrence of both the Attorney General and the Governor. A.R.S. §16-452(B).

CONCLUSION

This Court should reverse the Ninth Circuit’s

judgment and direct entry of judgment for Defendants.

February 12, 2021

Respectfully submitted,