Feldman et al v. Arizona Secretary of State's Office et al

Filing 412

ORDER - Defendants' 384 oral motion for judgment on partial findings is DENIED as moot. The Court finds in favor of Defendants and against Plaintiffs on all claims. The Clerk of the Court shall enter judgment accordingly and terminate this case. See document for complete details. Signed by Judge Douglas L Rayes on 5/8/2018. (ATD)

1

WO

2
3
4
5
6

IN THE UNITED STATES DISTRICT COURT

7

FOR THE DISTRICT OF ARIZONA

8
9

Democratic National Committee, DSCC, and
Arizona Democratic Party,

10

Plaintiffs,

11
12
13
14

No. CV-16-01065-PHX-DLR
FINDINGS OF FACT
AND CONCLUSIONS OF LAW

v.
Michele Reagan and Mark Brnovich,
Defendants.

15
16

Plaintiffs challenge two aspects of Arizona’s election system: (1) Arizona’s policy

17

to not count provisional ballots cast in the wrong precinct, which derives from the

18

collective effect of A.R.S. §§ 16-122, -135, -584, and related rules in the Arizona

19

Election Procedures Manual; and (2) Arizona House Bill 2023 (“H.B. 2023”), codified at

20

A.R.S. § 16-1005(H)-(I), which makes it a felony for anyone other than the voter to

21

possess that voter’s early mail ballot, unless the possessor falls within a statutorily

22

enumerated exception. Plaintiffs allege that the challenged laws violate § 2 of the Voting

23

Rights Act of 1965 (“VRA”) by adversely and disparately impacting the electoral

24

opportunities of Hispanic, African American, and Native American Arizonans, who

25

Plaintiffs claim are among their core constituencies. Plaintiffs also contend that these

26

provisions violate the First and Fourteenth Amendments to the United States Constitution

27

by severely and unjustifiably burdening voting and associational rights. Lastly, Plaintiffs

28

claim that H.B. 2023 violates § 2 of the VRA and the Fifteenth Amendment to the United

1

States Constitution because it was enacted with the intent to suppress voting by Hispanic

2

and Native American voters. (Doc. 360 at 4-7.)1 Plaintiffs seek (1) a declaration that the

3

challenged election practices are unlawful and (2) a permanent injunction requiring

4

Defendants to partially count out-of-precinct (“OOP”) provisional ballots for races for

5

which the voter otherwise was eligible to cast a vote and enjoining Defendants from

6

implementing, enforcing, or giving any effect to H.B. 2023. (Doc. 233 at 41-42.)

7

The Court presided over a ten-day bench trial beginning October 3, 2017 and

8

ending October 18, 2017. Pursuant to Federal Rule of Civil Procedure 52, and for the

9

following reasons, the Court finds against Plaintiffs and in favor of Defendants on all

10

claims.2

11

I. PARTIES

12

Plaintiffs are the Democratic National Committee (“DNC”), the Democratic

13

Senatorial Campaign Committee (“DSCC”), and the Arizona Democratic Party (“ADP”).

14

The DNC is a national committee dedicated to electing local, state, and national

15

candidates of the Democratic Party to public office. The DSCC is a Democratic political

16

committee dedicated to encouraging the election of Democratic Senate candidates to

17

office and is comprised of sitting Democratic members of the United States Senate. The

18

ADP is a state committee dedicated to electing candidates of the Democratic Party to

19

public office throughout Arizona.

20

Defendants are Arizona Secretary of State Michele Reagan and Arizona Attorney

21

General Mark Brnovich. Secretary Reagan is Arizona’s chief elections officer. Attorney

22

General Brnovich is Arizona’s chief legal officer, charged with enforcing state criminal

23

statutes, including H.B. 2023 and other election-related offenses.

24

drafts, and Attorney General Brnovich (in conjunction with the Governor of Arizona)

25
26
27
28

1

Secretary Reagan

For purposes of this order, “Doc.” refers to documents on the Court’s electronic
docket, “Ex.” to trial exhibits, “Tr.” to the official trial transcript, and “Dep.” to
designated deposition transcripts. Record citations offer examples of supporting
evidence, but are not intended to be exhaustive of all evidence supporting a proposition.
2
Defendants’ oral motion, made during trial, for judgment on partial findings
pursuant to Federal Rule of Civil Procedure 52(c) is denied as moot. (Doc. 384.)

-2-

1

approves, the Election Procedures Manual. A.R.S. §§ 41-191 et seq, 16-1021, -452.

2

The Court also permitted the following parties to intervene as defendants: (1) the

3

Arizona Republican Party (“ARP”), a state committee dedicated to electing candidates of

4

the Republican Party to public office; (2) Debbie Lesko, who at the time of intervention

5

was an Arizona State Senator representing Arizona’s 21st legislative district and Precinct

6

Committeewoman for Arizona’s 21st legislative district, and who recently was elected to

7

represent Arizona’s 8th congressional district in the United States House of

8

Representatives; (3) Tony Rivero, a member of the Arizona House of Representatives

9

representing Arizona’s 21st legislative district; (3) Bill Gates, who at the time of

10

intervention served as a City of Phoenix Councilman and Precinct Committeeman for

11

Arizona’s 28th legislative district, and who now serves as a member of the Maricopa

12

County Board of Supervisors representing district 3; and (4) Suzanne Klapp, a City of

13

Scottsdale Councilwoman and Precinct Committeewoman for Arizona’s 23rd legislative

14

district. (Docs. 39, 44, 56, 126.)

15

II. OVERVIEW OF TRIAL TESTIMONY

16

A. Plaintiffs’ Expert Witnesses

17

1. Dr. Allan Lichtman

18

Dr. Allan Lichtman is a Distinguished Professor of History at American

19

University in Washington, D.C., where he has been employed for 42 years. Dr. Lichtman

20

formerly served as Chair of the History Department and Associate Dean of the College of

21

Arts and Sciences at American University.

22

Brandeis University in 1967 and his Ph.D. in History from Harvard University in 1973,

23

with a specialty in the mathematical analysis of historical data. Dr. Lichtman’s areas of

24

expertise include political history, electoral analysis, and historical and quantitative

25

methodology. (Ex. 91 at 3-4.)

He received his B.A. in History from

26

Dr. Lichtman has worked as a consultant or expert witness for plaintiffs and

27

defendants in more than 80 voting and civil rights cases, including League of United

28

Latin American Citizens (LULAC) v. Perry, 548 U.S. 399 (2006), in which Justice

-3-

1

Kennedy’s majority opinion authoritatively cited Dr. Lichtman’s statistical work. Dr.

2

Lichtman also has testified several times for plaintiffs and defendants on issues of

3

intentional discrimination and application of Section 2 in VRA cases. (Ex. 91 at 4.)

4

Dr. Lichtman opined, generally, that under the totality of the circumstances H.B.

5

2023 causes minority voters to have less opportunity to participate in the political process

6

than non-minority voters, and that the law was passed with the intent to suppress minority

7

voters.3

8

historical analysis, including scholarly books, articles, reports, newspapers, voter

9

registration and turnout data, and scientific surveys.

He supported his opinions with the standard sources used in political and

10

Dr. Lichtman’s underlying sources, research, and statistical information are useful.

11

The surveys and data he supplied reveal significant socioeconomic disparities between

12

non-minorities and minorities, including in areas of poverty, unemployment, education,

13

transportation, and health. (Ex. 91 at 3-4.) His report also contains evidence that

14

Arizona exhibits racially polarized voting and has a history of racial appeals in political

15

campaigns that continue to this day. (Ex. 91 at 30, 44-45.) Dr. Lichtman opined that the

16

strong ties between race and partisanship in Arizona make targeting minorities the most

17

effective and efficient way for Republicans to advance their political prospects. (Ex. 93

18

at 4-5.)

19

Although the Court finds Dr. Lichtman’s curation of material facts surrounding the

20

legislative history and his underlying research to be helpful and reliable, the Court did not

21

find Dr. Lichtman’s ultimate opinions useful.

22

interpreted it to the data he assembled. In this respect, his opinions presented more like

23

an attorney’s closing argument than an objective analysis of data, and the credibility of

24

his trial testimony was undermined by his seeming effort to advocate a position rather

25

than answer a question. Moreover, applying law to facts is this Court’s duty, and it is one

Dr. Lichtman applied the law as he

26
27
28

3

For ease, the Court uses the terms “minority” to refer to the racial minorities
alleged to be adversely impacted by the challenge laws, and “non-minority” to refer to
non-Hispanic white voters.

-4-

1

the Court can do without the assistance of an expert opining on how he interprets the law

2

and thinks it should be applied. The Court also has not considered Dr. Lichtman’s

3

opinions on the ultimate issue of legislative intent, both because this issue is not the

4

proper subject of expert testimony and because it invades the province of the Court.

5

2. Dr. David Berman

6

Dr. David Berman is a Professor Emeritus of Political Science and a Senior

7

Research Fellow at the Morrison Institute for Public Policy at Arizona State University.

8

As a political science professor, he has taught undergraduate survey courses in American

9

government and politics, state and local politics, and Arizona government and politics, as

10

well as more specialized courses, including undergraduate seminars on Arizona politics

11

during which students interacted with state and local office holders and political

12

participants. He has also taught advanced graduate courses focusing on research methods

13

in these areas. (Ex. 89 at 3.)

14

As a Senior Research fellow with the Morrison Institute, Dr. Berman specializes in

15

research and writing on governance and election issues in Arizona, including

16

redistricting, direct democracy, and campaign finance. He has been a professor at

17

Arizona State University since 1966, and his previous work experience was as a Research

18

Associate at the National League of Cities in Washington, D.C. from 1964 to 1966. (Ex.

19

89 at 3-4.)

20

Dr. Berman opined that Arizona has a long history of discrimination against the

21

voting rights of Native Americans, Hispanics, and African Americans, and that this

22

discrimination is part of a more general pattern of political, social, and economic

23

discrimination against minority groups in areas such as school segregation, educational

24

funding and programming, equal pay and the right to work, and immigration.

25

The Court finds Dr. Berman credible. His opinions were well-researched and

26

rendered using standard sources and methodologies in his field of expertise, and his

27

sources were well-identified. Dr. Berman has authored ten books and over 70 published

28

papers, book chapters, or refereed articles dealing with state and local government,

-5-

1

politics, and public policy, and his opinions were based substantially on these prior

2

works. In particular, Dr. Berman drew heavily upon his book Arizona Politics and

3

Government: The Quest for Autonomy, Democracy, and Development (University of

4

Nebraska Press, 1998) and his review of archival papers and collections. (Ex. 89 at 3-4.)

5

The Court affords great weight to Dr. Berman’s opinions.

6

3. Dr. Jonathan Rodden

7

Dr. Jonathan Rodden is a tenured Professor of Political Science at Stanford

8

University and the founder and director of the Stanford Spatial Social Science Lab, a

9

center for research and teaching with a focus on the analysis of geo-spatial data in the

10

social sciences. Students and faculty members affiliated with the lab are engaged in a

11

variety of research projects involving large, fine-grained, geo-spatial data sets, including

12

individual records of registered voters, Census data, survey responses, and election

13

results at the level of polling. Prior to joining the Stanford faculty, Dr. Rodden was the

14

Ford Professor of Political Science at the Massachusetts Institute of Technology. He

15

received his Ph.D. from Yale University and his B.A. from the University of Michigan,

16

Ann Arbor, both in political science. (Ex. 95 at 5-6.)

17

Dr. Rodden has expertise in the use of large data sets and geographic information

18

systems to analyze aspects of political representation. He has developed a national data

19

set of geo-coded precinct-level election results that has been used extensively in policy-

20

oriented research related to redistricting and representation.

21

extensively with Census data from the United States and other countries.

He also has worked

22

Dr. Rodden has published papers on political geography and representation in a

23

variety of academic journals and has been featured in popular publications like the Wall

24

Street Journal, the New York Times, and Boston Review. Dr. Rodden has testified as an

25

expert witness in three recent election law cases. (Ex. 95 at 6.)

26

Here, Dr. Rodden analyzed the rates and causes of OOP voting in Arizona during

27

the 2012, 2014, and 2016 general elections. The Court finds his use of a combination of

28

individual-level and aggregate data analyses, both of which have been accepted in

-6-

1

previous cases analyzing questions under the VRA, to be valid and generally trustworthy,

2

and affords them great weight. (Ex. 97 at 7-9.)

3

Dr. Rodden found that Hispanic, Native American, and African American voters

4

cast OOP ballots at statistically higher rates than their non-minority counterparts. (Ex. 95

5

at 3-4; Ex. 97 at 2-4.) Focusing on Maricopa County in the 2012 election, Dr. Rodden

6

found that the rate of OOP voting was “131 percent higher for Hispanics, 74 percent

7

higher for African Americans, and 39 percent higher for Native Americans than whites.”

8

(Ex. 95 at 3-4.)

9

Further, Dr. Rodden found that OOP voters are substantially more likely to be

10

young and to live in neighborhoods characterized by large numbers of renters and with

11

high rates of transience, and that the rate of OOP voting was 65 percent higher for

12

Democratic voters than for Republican voters in Maricopa County, and 56 percent higher

13

in Pima County.

14

associated with higher rates of out-of-precinct voting,” and that “African Americans and

15

Hispanics are substantially more affected by this than whites. In particular, the impact of

16

precinct consolidation, while statistically significant for all groups, is more than twice as

17

large for Hispanics and African Americans as for non-Hispanic whites.” (Ex. 95 at 3-4.)

18

When analyzing Arizona’s non-metropolitan counties, Dr. Rodden found that OOP

19

voting is “negligible in majority-white precincts, but increases dramatically in precincts

20

where Hispanics and Native Americans make up majorities.” (Ex. 96 at 58.)

Dr. Rodden found that “changes in polling place locations are

21

In addition to his analysis of OOP voting, Dr. Rodden employed standard and

22

accepted methods in his field to analyze the “mailability” of Arizona’s non-metropolitan

23

counties in order to estimate the populations that likely would be most affected by H.B.

24

2023’s ballot collection restrictions. Though somewhat imprecise, the Court finds his

25

method of analysis to be creative given the lack of direct data available on the subject,

26

generally reliable, and based on sufficient data given the circumstances. Dr. Rodden

27

found that “[o]utside of Maricopa and Pima counties” “around 86 percent of non-

28

Hispanic whites have home mail service,” but “only 80 percent of Hispanics do, and only

-7-

1

18 percent of Native Americans have such access.” (Ex. 97 at 4.)

2

Dr. Rodden’s error rate is unknown, however, due to the lack of direct data. Also,

3

his analysis did not include Arizona’s metropolitan counties and therefore does not reveal

4

whether, on a statewide basis, minorities have disparate access to home mail service as

5

compared to non-minorities. Further, mail access is an imprecise proxy for determining

6

the number and demographics of voters who use or rely on ballot collection services.

7

Simply because a voter lacks home mail access does not necessarily mean that she uses

8

or relies on a ballot collector to vote, let alone a ballot collector who does not fall into

9

one of H.B. 2023’s exceptions. Accordingly, although Dr. Rodden’s analysis provided

10

useful insight into home mail access in non-metropolitan counties, the Court is mindful of

11

its limitations and affords these opinions moderate weight.

12

B. Plaintiffs’ Lay Witnesses

13

Plaintiffs called the following lay witnesses to testify at trial: Carmen Arias,

14

Michael Blair, Delilah Correa, Charlene Fernandez, LeNora Fulton, Steve Gallardo, Kate

15

Gallego, Kathleen Giebelhausen, Marva Gilbreath, Leah Gillespie, Carolyn Glover,

16

Leonard Gorman, Shari Kelso, Scott Konopasek, Joeseph Larios, Daniel Magos, Lori

17

Noonan, Patrick O’Connor, Martin Quezada, Nellie Ruiz, Spencer Scharff, Sam Shaprio,

18

Ken Clark, and John Powers. These witnesses include individual voters, representatives

19

from state, county, and municipal governments, community advocates who have

20

collected ballots as part of get-out-the-vote (“GOTV”) efforts, community advocates

21

focusing of Native American issues, Democratic Party operatives, a California state

22

elections official, and a former United States Department of Justice (“DOJ”) official.

23
24

C. Defendants’ Expert Witnesses
1. Dr. Donald Critchlow

25

Dr. Donald Critchlow works at Arizona State University as the Director of the

26

Center for Political Thought and Leadership, an organization funded by a grant from the

27

Charles Koch Foundation. (Tr. 1533-37.) He opined on the relationship between racial

28

discrimination and voting in Arizona. Dr. Critchlow made credible observations that

-8-

1

discrimination in Arizona has not been linear and that Arizona has taken effective action

2

to combat discrimination and encourage participation in voting.

3

With that said, Dr. Critchlow has never published a book or article focused

4

specifically on Arizona history, nor has he taught courses in Arizona history or politics.

5

(Tr. 1531-32.) Further, in many respects he offered one-sided opinions of Arizona’s

6

history, ignored incidents of discrimination, and failed to address the key political shift

7

between the Democratic and Republican parties during the Civil Rights Movement. For

8

example, he either was unfamiliar with or totally discounted the Republican strategy of

9

confrontation of minority voters at the polls during “Operation Eagle Eye” in the 1960s.

10

(Ex. 89 at 16; Tr. 1549.)

11

Arizona has a history of discrimination, his report appears to attribute past racial

12

discrimination in Arizona only to the Democratic Party and claims that discrimination has

13

not existed since the 1960s (in the Republican era). (Ex. 521 at 4.) For these reasons, the

14

Court affords little weight to Dr. Critchlow’s opinions

15

Additionally, although Dr. Critchlow acknowledged that

2. Sean Trende

16

Sean Trende critiqued Dr. Lichtman’s analysis of Arizona’s voting patterns and

17

history of racial discrimination, but offered no new information or analysis. Though the

18

Court found some of his criticisms worth considering, overall they were insignificant.

19

For example, although Trende generally agreed with Dr. Lichtman that Arizona

20

experiences racially polarized voting, he made much of the irrelevant fact that Arizona

21

voting is not as racially polarized as voting in Alabama. (Tr. 1837.) Additionally,

22

Trende’s opinions on the weight to give certain evidence and on the proper interpretation

23

and application of the law and evidence—like those of Dr. Lichtman’s—were not helpful

24

and invade the province of the Court. Moreover, Trende does not have a Ph.D and has

25

never written a peer-reviewed article. He has spent most of his professional career

26

working as a lawyer or political commentator. He is not a historian and says nothing

27

about the historical methods Dr. Lichtman utilized. (Tr. 1861-62.) For these reasons, the

28

Court affords Trende’s opinions little weight.

-9-

1

3. Dr. M.V. Hood

2

Dr. M.V. Hood is a Professor of Political Science at the University of Georgia.

3

Dr. Hood responded to the reports of Drs. Lichtman, Rodden, and Berman. (Ex. 522 at 2-

4

3.) For a number of reasons, the Court affords little weight to Dr. Hood’s opinions.

5

Dr. Hood criticizes Dr. Berman’s use of older historical information. Yet Dr.

6

Critchlow, another expert retained by Defendants, agrees with Dr. Berman that older

7

historical information is relevant to understanding patterns. (Ex. 521 at 8-10; Ex. 522 at

8

11.) Moreover, Dr. Hood admitted at trial that he examines historical information going

9

back 50 to 200 years. (Tr. 2122-23.)

10

Dr. Hood opined that H.B. 2023 does not hinder Native American voting because

11

the rates of early voting on the Navajo Nation increased from 2012 to 2016. He based

12

that opinion on early votes cast in three counties. This opinion is not reliable. Dr.

13

Hood’s analysis did not include an assessment of racial disparities and turnout. He also

14

conceded that myriad factors could affect turnout. (Tr. 2111-14.)

15

Dr. Hood prepared a cross-state comparative analysis of ballot collection laws and

16

policies related to counting OOP ballots. Although his analysis offered some insight, it

17

overall was not useful because he did not address statutory differences and nuances, and

18

his analysis reflected an incomplete understanding of the laws he categorized. For

19

example, some of the states he labeled as prohibiting ballot collection do not have laws

20

comparable to H.B. 2023 because they prohibit only the delivery of the ballot, not the

21

collection and mailing of the ballot on someone else’s behalf. (Ex. 92 at 52-53.)

22

The Court also notes that Dr. Hood’s testimony either has been rejected or given

23

little weight in numerous other cases due to concerns over its reliability. See Ne. Ohio

24

Coal. for the Homeless v. Husted, No. 2:06-CV-896, 2016 WL 3166251, at *24 (S.D.

25

Ohio June 7, 2016), aff’d in part, rev’d in part, 837 F.3d 612 (6th Cir. 2016); Veasey v.

26

Perry, 71 F. Supp. 3d 627, 663 (S.D. Tex. 2014); Frank v. Walker, 17 F. Supp. 3d 837,

27

881-84 (E.D. Wis. 2014), rev’d on other grounds, 768 F.3d 744 (7th Cir. 2014); Fla. v.

28

United States, 885 F. Supp. 2d 299, 324 (D.D.C. 2012); Common Cause/Ga. v. Billups,

- 10 -

1

No. 4:05-cv-0201, 2007 WL 7600409, at *14 (N.D. Ga. Sept. 6, 2007). Additionally,

2

most of Dr. Hood’s work has been as an expert on behalf of states defending against

3

allegations that their laws violated the Constitution or the VRA. (Tr. 2123-25.)

4

4. Dr. Janet Thornton

5

Dr. Janet Thornton is a Managing Director at Berkeley Research Group. Dr.

6

Thornton did not conduct her own analysis, but instead offered her opinion that Dr.

7

Rodden’s statistical work is flawed. (Ex. 525 at 1.) For example, she challenged Dr.

8

Rodden’s approaches to measuring racial disparities in OOP voting. One approach uses

9

individual surname data and geographic coordinates to infer race. Among Dr. Thornton’s

10

critiques was the presence of measurement error, which is well-taken. Indeed, even Dr.

11

Rodden concedes measurement error exists, especially as it pertains to African American

12

probabilities. Dr. Thornton did not critique the Hispanic probabilities assessed by Dr.

13

Rodden, however, and Dr. Rodden credibly explained that the measurement error for

14

Hispanic probabilities leads only to the under-estimation of racial disparities.

15

The second approach that Dr. Rodden employed relied on data collected by the

16

Census Department on race and ethnicity at the lowest possible level of geographic

17

aggregation. Dr. Thornton’s challenge to the aggregate approach was neither about the

18

data nor the presence of racial disparities in OOP voting, but rather the statistical model

19

employed by Dr. Rodden. Dr. Rodden, however, credibly showed that results similar to

20

those reported by his analysis are obtained using the alternative model specification or

21

measurement strategies recommended by Dr. Thornton.

22

Dr. Thornton’s opinion that there should have been a systematic decline in the

23

number of ballots cast in Arizona’s 13 non-metropolitan counties during 2016 if the

24

limits on ballot collection impacted the ability of rural and minority persons to vote is

25

simplistic and not credible. The statistical evidence suggests that increased turnout in

26

rural counties for the 2016 election was driven by non-minority voters, not Native

27

American and Hispanic voters. (Ex. 98 at 21-26.) Moreover, many factors impact voter

28

turnout, including controversial candidates and partisan mobilization efforts, all of which

- 11 -

1

might drown out the potentially deleterious effects of H.B. 2023. Overall, the Court finds

2

that Dr. Thornton’s critiques do not significantly undermine Dr. Rodden’s opinions and

3

therefore affords them less weight.

4

D. Defendants’ Lay Witnesses

5

Defendants called the following lay witnesses to testify at trial: Brad Nelson, Eric

6

Spencer, Helen Purcell, James Drake, Michael Johnson, Michelle Ugenti-Rita, Amy

7

Chan (formerly Amy Bjelland), Tony Rivero, and Scott Freeman.

8

include current and former lawmakers, elections officials, and law enforcement officials.

9

These witnesses

E. Witnesses Testifying By Deposition

10

In addition to the live testimony, the following witnesses testified by deposition:

11

Sheila Healy, Randy Parraz, Samantha Pstross, Secretary Reagan, Spencer Scharff,

12

Donald Shooter, Eric Spencer, Robyn Stallworth-Pouquette, Alexis Tameron, Victor

13

Vasquez, and Dr. Muer Yang. The parties each raised admissibility objections to certain

14

of these deposition designations. The Court addresses these objections, along with other

15

outstanding evidentiary matters, in a separate order.

16

III. OVERVIEW OF CHALLENGED ELECTIONS PRACTICES

17

A. H.B. 2023

18

Voting in Arizona involves a flexible mixture of early in-person voting, early

19

voting by mail, and traditional, in-person voting at polling places on Election Day.

20

Arizona voters do not need an excuse to vote early and Arizona permits early voting both

21

in person and by mail during the 27 days before an election. A.R.S. § 16-541. For those

22

voters who prefer to vote early and in-person, all Arizona counties operate at least one in-

23

person early voting location. Some of these locations are open on Saturdays. (Doc. 361

24

¶ 59.)

25

Arizona has allowed early voting by mail for over 25 years, and it has since

26

become the most popular method of voting, accounting for approximately 80 percent of

27

all ballots cast in the 2016 election. In 2007, Arizona implemented permanent no-excuse

28

early voting by mail, known as the Permanent Early Voter List (“PEVL”). Arizonans

- 12 -

1

now may vote early by mail either by requesting an early ballot on an election-by-

2

election basis, or by joining the PEVL, in which case they will be sent an early ballot as a

3

matter of course no later than the first day of the 27-day early voting period. A.R.S. §§

4

16-542, -544. In 2002, Arizona also became the first state to make available an online

5

voter registration option, allowing voters to register online through Arizona’s Motor

6

Vehicle Division (“MVD”) website, www.servicearizona.com. When registering online

7

through the MVD, voters can enroll in the PEVL by clicking a box. (Doc. 361 ¶ 56.)

8

To be counted, an early ballot must be received by the county recorder by 7:00

9

p.m. on Election Day. A.R.S. § 16-548(A). Early ballots contain instructions that inform

10

voters of the 7:00 p.m. deadline. Voters may return their early ballots by mail postage-

11

free, but they must mail them early enough to ensure that they are received by this

12

deadline. Additionally, some Arizona counties provide special drop boxes for early

13

ballots, and voters in all counties may return their early ballots in person at any polling

14

place, vote center, or authorized election official’s office without waiting in line. (Doc.

15

361 ¶¶ 57, 61.)

16

Since 1997, it has been the law in Arizona that “[o]nly the elector may be in

17

possession of that elector’s unvoted early ballot.” A.R.S. § 16-542(D). In 2016, Arizona

18

amended A.R.S. § 16-1005 by enacting H.B. 2023, which limits who may collect a

19

voter’s voted or unvoted early ballot:

20
21
22
23
24
25
26
27
28

H. A person who knowingly collects voted or unvoted early
ballots from another person is guilty of a class 6 felony. An
election official, a United States postal service worker or any
other person who is allowed by law to transmit United States
mail is deemed not to have collected an early ballot if the
official, worker or other person is engaged in official duties.
I. Subsection H of this section does not apply to:
1. An election held by a special taxing district formed
pursuant to title 48 for the purpose of protecting or providing
services to agricultural lands or crops and that is authorized to
conduct elections pursuant to title 48.
2. A family member, household member or caregiver of the
voter. For the purposes of this paragraph:
(a) “Caregiver” means a person who provides medical or
- 13 -

health care assistance to the voter in a residence, nursing care
institution, hospice facility, assisted living center, assisted
living facility, assisted living home, residential care
institution, adult day health care facility or adult foster care
home.

1
2
3

(b) “Collects” means to gain possession or control of an early
ballot.

4
5
6

(c) “Family member” means a person who is related to the
voter by blood, marriage, adoption or legal guardianship.

7

(d) “Household member” means a person who resides at the
same residence as the voter.

8
9

A.R.S. § 16-1005(H)-(I). Voters therefore may entrust a caregiver, family member,

10

household member, mail carrier, or elections official to return their early ballots, but may

11

not entrust other, unauthorized third parties to do so.

12

B. Rejection of OOP Ballots

13

Since at least 1970, Arizona has required voters who choose to vote in person on

14

Election Day to cast their ballots in their assigned precinct and has enforced this system

15

by counting only those ballots cast in the correct precinct. (Doc. 361 ¶ 46.) Because

16

elections involve many different overlapping jurisdictions, the precinct-based system

17

ensures that each voter receives a ballot reflecting only the races for which that person is

18

entitled to vote. (Ex. 95 at 10.) If a voter arrives at a precinct but does not appear on the

19

precinct register, Arizona allows the voter to cast a provisional ballot. A.R.S. §§ 16-122,

20

-135, -584. After Election Day, county elections officials review all provisional ballots.

21

If a voter’s address is determined to be within the precinct, the provisional ballot is

22

counted. Arizona does not count any portion of a provisional ballot cast outside of a

23

voter’s correct precinct. A majority of states do not count OOP ballots, putting Arizona

24

well within the mainstream on this issue.4 Indeed, at no point has the DOJ objected to

25
26
27
28

4

See Ala. Code §§ 17-9-10, -10-2, -10-3; Ark. Code Ann. §§ 7-5-306(b), 308(d)(2); Conn. Gen. Stat. §§ 9-19j, -232n; Del. Code Ann. tit. 15, § 4948(b), h(7); Fla.
Stat. § 101.048(1),(2); Haw. Admin. Rules § 3-172-140; Ind. Code §§ 3-11.7-2-1, -11-82, and -11.7-5-3; Iowa Code §§ 49.9, 49.79(2)(c), 49.80, 49.81, 53.23; Tit. 31 Ky.
Admin. Regs. § 6:020(1),(14); Mich. Comp. Laws §§ 168.523a(1),(5),(7), 168.813(1);
Miss. Code, Ann. § 23-15- 573(1),(3)(b); Mo. Rev. Stat. § 115.430(2),(3),(6); Mont.
Code §§ 13-15-107(1),(3), 13-2-512, 13-13-114(1)(a),(2); Neb. Rev. Stat. §§ 32-915(1), -

- 14 -

1

this practice, and Plaintiffs object to it for the first time in this case.

2

In 2011, Arizona amended its elections code to allow counties to choose whether

3

to conduct elections under the traditional precinct model or to use a “vote center” system.

4

2011 Ariz. Legis. Serv. Ch. 331 (H.B. 2303) (April 29, 2011) (amending A.R.S. § 16-

5

411). Unlike the precinct-based system, the vote center model requires each vote center

6

to be equipped to print a specific ballot, depending on each voter’s particular district, that

7

includes all races for which that voter is eligible to vote. Thus, under a vote center

8

system, voters may cast their ballots at any vote center in the county in which they reside

9

and receive the appropriate ballot. A.R.S. § 16-411(B)(4). Graham, Greenlee, Cochise,

10

Navajo, Yavapai, and Yuma counties have adopted the vote center model.

11

counties are mostly rural and sparsely populated. Precinct-based voting requirements,

12

such as Arizona’s policy to not count OOP ballots, have no impact on voters in these

13

counties. By comparison, the most populous counties in Arizona, such as Maricopa,

14

Pima, and Pinal, currently adhere to the traditional precinct-based model.

15

IV. PRELIMINARY ISSUES

These

16

A. Standing

17

Article III of the United States Constitution limits federal courts to resolving

18

“Cases” and “Controversies,” one element of which is standing. To have standing to

19

litigate in federal court, a plaintiff “must have suffered or be imminently threatened with

20

a concrete and particularized ‘injury in fact’ that is fairly traceable to the challenged

21

action of the defendant and likely to be redressed by a favorable judicial decision.”

22

Lexmark Int’l, Inc. v. Static Control Components, Inc., -- U.S. --, 134 S. Ct. 1377, 1386

23

(2014) (citing Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61 (1992)). Only one

24

plaintiff needs to have standing when only injunctive relief is sought. Crawford v.

25
26
27
28

1002(5)(b),(e); Nev. Rev. Stat. § 293.3081(1), 293.3085(4); N.Y. U.C.C. Law §§ 8302(3)(e), § 9-209; Ohio Rev. Stat. §§ 3505.181(A)(1), 3505.183(B)(1), (4)(a); S.C. Code
Ann. §§ 7-13-820, 7-13-830; S.D. Sess. Laws § 12-18-39, 12-20-5.1; Tenn. Code Ann. §
2-7-112(a)(3)(A),(B); Tex. Elec. Code Ann. §§ 63.001(c),(g), 63.011(a),(b); Tex. Admin.
Code § 81.172.

- 15 -

1

Marion Cty. Election Bd., 472 F.3d 949, 951 (7th Cir. 2007), aff’d, 553 U.S. 181, 189 n.7

2

(2008).

3

Plaintiffs have organizational standing to challenge the election regulations at

4

issue. Ballot collection was a GOTV strategy used primarily by the Democratic Party to

5

increase electoral participation by otherwise low-efficacy voters. (Tr. 416-26, 632-33,

6

659, 902, 930; Healy Dep. 28:15-29:13.) H.B. 2023’s limitations will require Democratic

7

organizations, such as the ADP, to retool their GOTV strategies and divert more

8

resources to ensure that low-efficacy voters are returning their early mail ballots.

9

Additionally, credible expert testimony shows that minority voters, who tend to vote

10

disproportionately for Democratic candidates, vote OOP at higher rates than non-

11

minority voters. Thus, Arizona’s policy to not count OOP ballots places a greater

12

imperative on organizations like the ADP to educate their voters. These are sufficiently

13

concrete and particularized injuries that are fairly traceable to the challenged provisions.

14

See Crawford, 472 F.3d at 951 (“Thus the new law injures the Democratic Party by

15

compelling the party to devote resources to getting to the polls those of its supporters

16

who would otherwise be discouraged by the new law from bothering to vote.”); One Wis.

17

Inst., Inc. v. Nichol, 186 F. Supp. 3d 958, 967 (W.D. Wis. 2016) (finding expenditure of

18

resources for educating voters about how to comply with new state voter registration

19

requirements sufficient to establish standing).

20
21
22
23
24
25

Plaintiffs also have associational standing to challenge these provisions on behalf
of their members.
[A]n association has standing to bring suit on behalf of its
members when: (a) its members would otherwise have
standing to sue in their own right; (b) the interests it seeks to
protect are germane to the organization’s purpose; and (c)
neither the claim asserted nor the relief requested requires the
participation of individual members in the lawsuit.

26

Hunt v. Wash. State Apple Adver. Comm’n, 432 U.S. 333, 343 (1977). A number of self-

27

identified Democratic voters testified either that they have used ballot collection services

28

in the past, or that they have voted OOP. The voting rights of such individuals are

- 16 -

1

germane to Plaintiffs’ goal of electing Democratic candidates to local, state, and federal

2

offices. Further, neither the claims asserted nor the relief requested requires individual

3

members to participate in this lawsuit.

4

Finally, Plaintiffs’ asserted injuries can be redressed by a favorable decision of

5

this Court. “[W]hen a plaintiff challenges the constitutionality of a rule of law, it is the

6

state official designated to enforce that rule who is the proper defendant[.]” Am. Civil

7

Liberties Union v. The Fla. Bar, 999 F.2d 1486, 1490 (11th Cir. 1993). Here, county

8

officials are responsible for counting ballots and verifying proper voter registration, see

9

A.R.S. §§ 16-621(A), -584(E), but Secretary Reagan and Attorney General Brnovich also

10

play a role in determining how OOP ballots are counted. Arizona law requires Secretary

11

Reagan, after consulting with county officials, to “prescribe rules to achieve and maintain

12

the maximum degree of correctness, impartiality, uniformity and efficiency on the

13

procedures for early voting and voting, and of producing, distributing, collecting,

14

counting, tabulating and storing ballots.” A.R.S. § 16-452(A). These rules are prescribed

15

in the Election Procedures Manual and have the force of law. A.R.S. § 16-452(B)-(C).

16

“Any person who does not abide by the Secretary of State’s rules is subject to criminal

17

penalties,” Ariz. Libertarian Party, Inc. v. Bayless, 351 F.3d 1277, 1280 (9th Cir. 2003)

18

(citing A.R.S. § 16-452(C)), and Attorney General Brnovich is authorized to prosecute

19

such violations, A.R.S. § 16-1021.

20

physically counting ballots, they are not empowered to count or reject ballots at their

21

discretion. Rather, “[a]ll proceedings at the counting center shall be under the direction

22

of the board of supervisors or other officer in charge of elections and shall be conducted

23

in accordance with the approved instructions and procedures manual[.]” A.R.S. § 16-

24

621(A).

Although county officials are responsible for

25

Though the Court cannot require Secretary Reagan and Attorney General

26

Brnovich to physically count OOP ballots for races for which the voter was otherwise

27

eligible to cast a vote, it can require them to prescribe such a procedure in the Election

28

Procedures Manual, which county election officials then would be bound by law to

- 17 -

1

follow. Further, Attorney General Brnovich can ensure compliance with such a directive

2

because he is authorized to prosecute county officials who violate it.

3

Likewise, Attorney General Brnovich is empowered to enforce state election laws

4

like H.B. 2023.

5

Brnovich is authorized to enforce Arizona’s election laws “[i]n any election for state

6

office, members of the legislature, justices of the supreme court, judges of the court of

7

appeals or statewide initiative or referendum,” but in elections for “county, city or town

8

office, community college district governing board, judge or a county, city, or town

9

initiative or referendum,” that authority resides with “the appropriate county, city or town

10

attorney[.]” A.R.S. § 16-1021. But most elections will include statewide races and

11

therefore Attorney General Brnovich likely will share enforcement authority in most

12

circumstances. Moreover, although Attorney General Brnovich might lack authority to

13

direct the enforcement activities of county and municipal prosecutors, there is no reason

14

to believe that these local law enforcement officials will attempt to enforce H.B. 2023

15

should the Court declare it unconstitutional or unlawful under the VRA.

He is not the only official with such authority; Attorney General

16

Lastly, although there is no evidence that Secretary Reagan or other state or local

17

elections officials play a direct role in the enforcement of H.B. 2023, Secretary Reagan

18

has some indirect involvement in the law’s implementation by virtue of her responsibility

19

for drafting the Election Procedures Manual. If the Court were to enjoin H.B. 2023’s

20

implementation and enforcement, the Election Procedures Manual would need to reflect

21

as much.

22

B. Effect of Preliminary Appellate Proceedings

23

On September 23, 2016, the Court denied Plaintiffs’ motion to preliminarily

24

enjoin enforcement of H.B. 2023. (Doc. 204.) On October 4, 2016, the Court also

25

denied Plaintiffs’ motion to preliminary enjoin enforcement of H.B. 2023 pending

26

Plaintiffs’ appeal of the Court’s September 23 order. (Doc. 213.) Plaintiffs thereafter

27

moved the Ninth Circuit Court of Appeals for an injunction pending appeal, which was

28

denied by a three-judge motions panel. Later, on October 28, 2016, a divided three-judge

- 18 -

1

merits panel affirmed the Court’s order denying Plaintiffs’ preliminary injunction motion.

2

Chief Judge Thomas dissented.

3

On November 2, 2016, a majority the Ninth Circuit’s non-recused active judges

4

voted to rehear the case en banc. Two days later, a majority of the en banc panel voted to

5

preliminarily enjoin enforcement of H.B. 2023 pending the panel’s rehearing, essentially

6

for the reasons provided in Chief Judge Thomas’ dissent.5 This preliminary injunction

7

was short-lived, however, as the United States Supreme Court stayed the order on

8

November 5, 2016, pending the Ninth Circuit’s final disposition of the appeal.

9

In light of this history, the parties disagree over the effect that Chief Judge

10

Thomas’ dissent should have on the Court’s post-trial analysis. As explained during the

11

final pretrial conference, although the Court has considered Chief Judge Thomas’ dissent,

12

the Court is not bound by its factual analysis. To date, all appellate proceedings have

13

occurred at the preliminary injunction stage on a less developed factual record. Findings

14

and conclusions rendered at the preliminary injunction stage are just that—preliminary.

15

They do not necessarily preclude the Court from making different findings or conclusions

16

after thorough factual development and a full trial on the merits. Accordingly, although

17

the Court is mindful of Chief Judge Thomas’ critiques and their preliminary adoption by

18

a majority of the en banc panel, the Court is not bound to make identical findings and

19

conclusions as those made at a preliminary phase of the litigation.

20

And with that, the Court proceeds to the merits.

21

V. FIRST AND FOURTEENTH AMENDMENTS6

22

“[T]he Constitution of the United States protects the right of all qualified citizens

23

to vote, in state as well as in federal elections.” Reynolds v. Sims, 377 U.S. 533, 554

24
25
26
27
28

5

The en banc panel technically issued a stay of the Court’s order denying
Plaintiffs’ preliminary injunction motion, but the stay had the practical effect of an
injunction pending appeal.
6
Because Plaintiffs challenge state election laws, their claims technically arise
under the Fourteenth Amendment, which applies the First Amendment’s protections
against states and their political subdivisions. See City of Ladue v. Gilleo, 512 U.S. 43,
45 n.1 (1994).

- 19 -

1

(1964). Relatedly, the First and Fourteenth Amendments protect the right of the people

2

to associate for political purposes. Tashjian v. Republican Party of Conn., 479 U.S. 208,

3

214 (1986). “It does not follow, however, that the right to vote in any manner and the

4

right to associate for political purposes . . . are absolute.” Burdick v. Takushi, 504 U.S.

5

428, 433 (1992). Rather, the Constitution empowers states to regulate the “Times, Places

6

and Manner of holding Elections for Senators and Representatives,” U.S. Const. art. I, §

7

4, cl. 1, and states retain “control over the election process for state offices,” Tashjian,

8

479 U.S. at 217. “Common sense, as well as constitutional law, compels the conclusion

9

that government must play an active role in structuring elections.” Burdick, 504 U.S. at

10

433. “[A]s a practical matter, there must be a substantial regulation of elections if they

11

are to be fair and honest and if some sort of order, rather than chaos, is to accompany the

12

democratic processes.” Storer v. Brown, 415 U.S. 724, 730 (1974).

13

Like an individual’s voting and associational rights, however, a state’s power to

14

regulate elections is not absolute; it is “subject to the limitation that [it] may not be

15

exercised in a way that violates other . . . provisions of the Constitution.” Williams v.

16

Rhodes, 393 U.S. 23, 29 (1968); see Wash. State Grange v. Wash. State Republican

17

Party, 552 U.S. 442, 451 (2008). But because all election regulations “invariably impose

18

some burden upon individual voters,” Burdick, 504 U.S. at 433, “not every voting

19

regulation is subject to strict scrutiny,” Pub. Integrity Alliance, Inc. v. City of Tucson, 836

20

F.3d 1019, 1024 (9th Cir. 2016).

21
22
23
24
25
26
27
28

Instead, . . . a more flexible standard applies. A court
considering a challenge to a state election law must weigh
“the character and magnitude of the asserted injury to the
rights protected by the First and Fourteenth Amendments that
the plaintiff seeks to vindicate” against “the precise interests
put forward by the State as justifications for the burden
imposed by its rule,” taking into consideration “the extent to
which those interests make it necessary to burden the
plaintiff's rights.”
Burdick, 504 U.S. at 434 (quoting Anderson v. Celebrezze, 460 U.S. 780, 789 (1983)).
This framework commonly is referred to as the Anderson/Burdick test, after the two
Supreme Court decisions from which it derives.
- 20 -

1

Under this framework, the degree to which the Court scrutinizes “the propriety of

2

a state election law depends upon the extent to which a challenged regulation burdens

3

First and Fourteenth Amendment rights.” Id. A law that imposes severe burdens is

4

subject to strict scrutiny, meaning it must be narrowly tailored to serve a compelling state

5

interest. Id. “Regulations imposing . . . [l]esser burdens, however, trigger less exacting

6

review, and a State’s ‘important regulatory interests’ will usually be enough to justify

7

‘reasonable, nondiscriminatory restrictions.’” Timmons v. Twin Cities Area New Party,

8

520 U.S. 351, 358 (1997) (quoting Burdick, 504 U.S. at 434); see also Pub. Integrity

9

Alliance, 836 F.3d at 1024 (“Applying these precepts, ‘[w]e have repeatedly upheld as

10

‘not severe’ restrictions that are generally applicable, evenhanded, politically neutral, and

11

protect the reliability and integrity of the election process.’” (quoting Dudum v. Arntz,

12

640 F.3d 1098, 1106 (9th Cir. 2011)). Additionally, when applying Anderson/Burdick,

13

the Court considers the state’s election regime as a whole, including aspects that mitigate

14

the hardships that might be imposed by the challenged provisions. See Ohio Democratic

15

Party v. Husted, 834 F.3d 620, 627 (6th Cir. 2016); see also Crawford v. Marion Cty.

16

Election Bd., 553 U.S. 181, 199 (2008) (considering mitigating aspects of Indiana’s

17

election laws).

18
19

A. Application to H.B. 2023
1. Burden on Voting Rights

20

At most, H.B. 2023 minimally burdens Arizona voters as a whole. In fact, the vast

21

majority of Arizona voters are unaffected by the law. Although voting by early mail

22

ballot has steadily increased in Arizona, in any given election there remains a subset of

23

voters who choose to vote in person, either early at a designated early voting site or on

24

Election Day. In-person voters are not impacted by limitations on who may collect early

25

mail ballots. For example, 2,323,579 registered voters cast ballots during the 2012

26

general election. (Ex. 543 at 2.) Of these, 1,542,855 submitted early mail ballots, over

27

99 percent of which were counted. (Ex. 95 at 17.) Thus, roughly a third of all Arizonans

28

voted in person during the 2012 general election. Similarly, approximately 80 percent of

- 21 -

1

the 2,661,497 Arizonans who voted during the 2016 general election cast an early ballot,

2

meaning about 20 percent voted in person on Election Day. (Tr. 1925; Ex. 543.) H.B.

3

2023 has no impact on these voters.

4

Further, even under a generous interpretation of the evidence, the vast majority of

5

voters who choose to vote early by mail do not return their ballots with the assistance of a

6

third-party collector who does not fall within H.B. 2023’s exceptions. There are no

7

records of the numbers of voters who, in any given election, return their ballots with the

8

assistance of third parties. The ADP collected “a couple thousand” ballots in 2014.

9

(Tameron Dep. 52:12-17.) According to Secretary Reagan, community advocate Randy

10

Parraz testified before the Arizona Senate Elections Committee that he had once collected

11

4,000 ballots. (Regan Dep. 101:12-21.) During closing argument, the Court asked

12

Plaintiffs’ counsel for his best estimate of the number of voters affected by H.B. 2023

13

based on the evidence at trial, to which he responded: “Thousands . . . but I don’t have a

14

precise number of that.” (Tr. 2268.) An estimate of “thousands” offers little guidance for

15

determining where, on the scale of 1,000 to 999,999, the number falls, but the evidence

16

and Counsel’s response suggests that possibly fewer than 10,000 voters are impacted.

17

Purely as a hypothetical, if the Court were to draw the unjustified inference that

18

100,000 early mail ballots were collected and returned by third parties during the 2012

19

general election, that estimate would leave over 1.4 million early mail ballots that were

20

returned without such assistance. The point, of course, is that H.B. 2023’s limitations

21

have no effect on the vast majority of voters who vote by early mail ballot because, even

22

under generous assumptions, relatively few early voters give their ballots to individuals

23

who would be prohibited by H.B. 2023 from possessing them.

24

On its face, H.B. 2023 is generally applicable and does not increase the ordinary

25

burdens traditionally associated with voting. The law merely limits who may possess,

26

and therefore return, a voter’s early mail ballot. Early voters may return their own

27

ballots, either in person or by mail, or they may entrust a family member, household

28

member, or caregiver to do the same. Thus, the burden H.B. 2023 imposes is the burden

- 22 -

1

of traveling to a mail box, post office, early ballot drop box, any polling place or vote

2

center (without waiting in line), or an authorized election official’s office, either

3

personally or with the assistance of a statutorily authorized proxy, during a 27-day early

4

voting period.7

5

Even with H.B. 2023’s limitations, the burden on early voters to return their early

6

mail ballots is less severe than the burden on in-person voters, who must travel to a

7

designated polling place or vote center on Election Day, often necessitating taking time

8

off work and waiting in line. Indeed, the burden on early mail voters is less severe even

9

than the burden on early in-person voters, who must travel to a designated early voting

10

location during the 27-day early voting period. Plaintiffs do not contend that the more

11

onerous travel required of in-person voters is unconstitutionally burdensome, nor would

12

the law support such an argument.

13

For example, in Crawford the Supreme Court considered whether Indiana’s voter

14

identification law, which required in-person voters to present photo identification,

15

unconstitutionally burdened the right to vote. 553 U.S. at 185. A voter who had photo

16

identification but was unable to present it on Election Day, or a voter who was indigent

17

or had a religious objection to being photographed, could cast a provisional ballot, which

18

then would be counted if the voter traveled to the circuit court clerk within ten days after

19

the election and either presented photo identification or executed an affidavit. Id. at 185-

20

86.

21
22
23
24
25
26
27
28

7

Throughout this case, Plaintiffs have conflated the burden imposed by H.B. 2023
with the circumstances that might make that burden harder to surmount for certain voters.
That is, Plaintiffs conflate the burden with its severity. For example, during closing
argument the Court asked Plaintiffs’ counsel to summarize the precise burdens that H.B.
2023 imposes. (Tr. 2262.) Counsel responded that the burdens include lack of mail
access, inadequate transportation, disabilities, low education attainment, and residential
instability. (Tr. 2263.) But H.B. 2023 does not impose these conditions on any voter.
The sole burden H.B. 2023 imposes is the burden of traveling to a mail box, post office,
early ballot drop box, polling place or vote center, or authorized election official’s office,
either personally or with authorized assistance, during a 27-day early voting period. The
socioeconomic circumstances cited by Plaintiffs might explain why this process is more
difficult for some voters than others, but those circumstances are not themselves the
burden imposed by the challenged law.

- 23 -

1

In his controlling opinion upholding the constitutionality of the challenged law,

2

Justice Stevens explained “[t]he burdens that are relevant to the issue before us are those

3

imposed on persons who are eligible to vote but do not possess a current photo

4

identification that complies with the requirements of” the challenged law. Id. at 198.

5

The Court characterized these burdens as “the inconvenience of making a trip to the

6

[Indiana Bureau of Motor Vehicles], gathering the required documents, and posing for a

7

photograph,” to obtain the required identification, and concluded that this process “does

8

not qualify as a substantial burden on the right to vote, or even represent a significant

9

increase over the usual burdens of voting.”8 Id. The Court also reasoned that “[t]he

10

severity of that burden is . . . mitigated by the fact that, if eligible, voters without photo

11

identification may cast provisional ballots that will ultimately be counted,” although to do

12

so voters would need to make two trips: one to vote in the first instance and another to the

13

circuit court clerk’s office. Id. at 199.

14

At most, H.B. 2023 requires only that early mail voters make the first trip

15

described in Crawford—the trip to vote. Further, the trip H.B. 2023 requires voters to

16

make is less burdensome because an Arizona early mail voter has 27 days in which to

17

make it, can choose between traveling to the nearer and most convenient of either a

18

personal mailbox, post office, early ballot drop box, polling place or vote center, or

19

authorized election official’s office, and can have a family member, household member,

20

or caregiver make the trip on her behalf. Voting early by mail in Arizona is far easier

21

than traditional, in-person voting on Election Day, and if laws that do not “represent a

22

significant increase over the usual burdens of voting” do not severely burden the

23

franchise, id. at 198, it is illogical to conclude that H.B. 2023 imposes a severe burden on

24

Arizona voters.

25
26
27
28

For these reasons, Plaintiffs’ Fourteenth Amendment challenge is best understood
8

The Supreme Court did not characterize the burdens imposed by Indiana’s photo
identification law as the circumstances of particular voters that made it harder to obtain
the required identification. Rather, those conditions informed the analysis of the severity
of the burden on discrete subgroups.

- 24 -

1

as follows: H.B. 2023 has no impact on the vast majority of Arizona voters, but its

2

limitations on who may return a voter’s early mail ballot present special difficulties for a

3

small subset of socioeconomically disadvantaged voters. When evaluating the severity of

4

burdens imposed by a challenged law, “courts may consider not only a given law’s

5

impact on the electorate in general, but also its impact on subgroups, for whom the

6

burden, when considered in context, may be more severe.” Pub. Integrity Alliance, 836

7

F.3d at 1024 n.2 (citing Crawford, 553 U.S. at 199-203, 212-17 (Souter, J., dissenting)).

8

But to do so, the challengers must present sufficient evidence to enable the court to

9

quantify the magnitude of the burden imposed on the subgroup. Id.; see also Ne. Ohio

10

Coal. for the Homeless v. Husted, 837 F.3d 612, 631-32 (6th Cir. 2016) (explaining that,

11

even under this “more liberal approach to burden measuring,” the record must contain

12

“quantifiable evidence from which an arbiter could gauge the frequency with which this

13

narrow class of voters has been or will become disenfranchised as a result of” the

14

challenged law).

15

Thus, in Crawford the Supreme Court acknowledged that Indiana’s voter

16

identification law might place “a somewhat heavier burden . . . on a limited number of

17

persons,” such as “elderly persons born out of State, who may have difficulty obtaining a

18

birth certificate; persons who because of economic or other personal limitations may find

19

it difficult either to secure a copy of their birth certificate or to assemble the other

20

required documentation to obtain a state-issued identification; homeless persons; and

21

persons with a religious objection to being photographed.” 553 U.S. at 199. But the

22

Court declined to consider these burdens because “on the basis of the evidence in the

23

record it [was] not possible to quantify either the magnitude of the burden on this narrow

24

class of voters or the portion of the burden imposed on them that [was] fully justified.”

25

Id. at 200.

26

Like in Crawford, this Court has insufficient evidence from which to measure the

27

burdens on discrete subsets of voters. The Court cannot quantify with any degree of

28

certainty “the number of registered voters” who, in past elections, returned early mail

- 25 -

1

ballots with the assistance of ballot collectors who do not fall within H.B. 2023’s

2

exceptions. Id. at 200. The Court therefore cannot determine how frequently voters will

3

be impacted by H.B. 2023’s limitations. And of the nebulous “thousands” who, in past

4

elections, have entrusted their ballots to third parties, there is insufficient “concrete

5

evidence” for the Court to gauge the magnitude of that burden or the portion of it that is

6

justified. Id. at 201. Stated differently, it is not enough to know roughly how many

7

voters have used ballot collection services—which, in any event, the Court cannot

8

determine on this record. The Court also needs to know why voters used these services

9

so that it may determine whether those voters did so out of convenience or personal

10

preference, as opposed to meaningful hardship, and whether other aspects of Arizona’s

11

election system adequately mitigate those burdens.

12

The evidence available largely shows that voters who have used ballot collection

13

services in the past have done so out of convenience or personal preference, or because of

14

circumstances that Arizona law adequately accommodates in other ways. Joseph Larios,

15

a community advocate who has collected ballots in past elections, testified that in his

16

experience returning early mail ballots presents special challenges for communities that

17

lack easy access to outgoing mail services; the elderly, homebound, and disabled voters;

18

socioeconomically disadvantaged voters who lack reliable transportation; voters who

19

have trouble finding time to return mail because they work multiple jobs or lack childcare

20

services; and voters who are unfamiliar with the voting process and therefore do not vote

21

without assistance or tend to miss critical deadlines. (Tr. 416-26, 432-39.)

22

As to this latter category of voters who, due either to forgetfulness or unfamiliarity

23

with the voting process, choose not to vote or neglect to mail their ballots in time for

24

them to reach the county recorder by 7:00 p.m. on Election Day, H.B. 2023 does not

25

impose a severe burden. Remembering relevant election deadlines “does not qualify as a

26

substantial burden on the right to vote, or even represent a significant increase over the

27

usual burdens of voting.” Crawford, 553 U.S. at 198. Moreover, nothing in H.B. 2023

28

prohibits Plaintiffs or other organizations from educating voters and offering assistance in

- 26 -

1

understanding and completing a ballot.

2

As for the other types of voters Larios identified, Arizona accommodates many of

3

the circumstances that tend to make voting in general (and not just early mail voting)

4

more difficult for them. For example, all counties must provide special election boards

5

for voters who cannot travel to a polling location because of an illness or disability.

6

A.R.S. § 16-549. If an ill or disabled voter timely requests an accommodation, the

7

county recorder must arrange for a special election board to deliver a ballot to the voter in

8

person. Although relatively few voters are aware of this service (Tr. 864-65), nothing in

9

H.B. 2023 prevents Plaintiffs from educating voters about the special election board

10

option and assisting them in making those arrangements. Arizona also allows curbside

11

voting at polling places, where election officials will go out to a vehicle to assist voters as

12

necessary.9

13

For working voters, Arizona law requires employers to give an employee time off

14

to vote if the employee is scheduled to work a shift on Election Day that provides fewer

15

than three consecutive hours between either the opening of the polls and the beginning of

16

the shift, or the end of the shift and the closing of the polls. A.R.S. § 16-402. An

17

employer is prohibited from penalizing an employee for exercising this right. If voters

18

nonetheless feel uncomfortable requesting time off, they have a 27-day window to vote in

19

person at an on-site early voting location. Additionally, even under H.B. 2023 voters

20

with transportation difficulties or time limitations may entrust their early ballots to family

21

members, household members, caregivers, or elections officials.

22

The testimony of individual voters who have used ballot collection services in past

23

elections largely confirms that H.B. 2023 does not impose significant burdens. Five

24

voters testified at trial about their personal experiences with ballot collectors: Nellie Ruiz,

25
26
27
28

9

It is of no moment that entrusting a ballot to a volunteer is relatively more
convenient than arranging a special election board. In Crawford, voting without the
required identification certainly would have been easier than voting provisionally and
then travelling to the circuit court clerk’s office within ten days. Nonetheless, the
controlling opinion found this option to be an adequate mitigating alternative.

- 27 -

1

Carolyn Glover, Daniel Magos, Carmen Arias, and Marva Gilbreath. None of these

2

voters would be severely burdened by H.B. 2023’s limitations.

3

Ruiz, a 71-year-old early mail voter in Phoenix, testified that she typically asks her

4

neighbor to return her ballot because her rheumatoid arthritis and deteriorating eyesight

5

make it difficult for her to return it personally. Ruiz lives with her adult son and

6

daughter-in-law.

7

important documents, like bill payments and ballots. Instead, her son delivers her bill

8

payments whenever he delivers his own mail. Ruiz testified that she preferred to give her

9

ballot to her neighbor because she “didn’t want to impose on [her] children,” but could

10

not explain why her son could not return her ballot the same way he returns her bills, or

11

why asking him to deliver a ballot was any more of an imposition than asking him to

12

deliver her bills. Ruiz also was not aware that her son could drop off her ballot when he

13

goes to the polls to vote in person. Ruiz testified that she was not able to give her early

14

mail ballot to her neighbor during the 2016 general election because of H.B. 2023.

15

Nonetheless, Ruiz successfully returned her ballot by mailing it from her home mailbox.

16

(Tr. 93-96, 98-100, 102-103, 111.)

Although Ruiz has a personal mailbox, she prefers not to mail

17

H.B. 2023 does not burden voters like Ruiz. She admittedly was able to mail her

18

ballot in 2016 without relying on her neighbor and lives with her adult son who is capable

19

of returning her ballots, either by mail the same way he returns her bill payments, or at a

20

polling place when he votes in person.

21

Glover, a retired voter with mobility issues who resides in a senior citizens

22

apartment complex in Phoenix, testified that prior to the 2016 general election persons

23

affiliated with the Democratic Party would collet her early mail ballot. Glover initially

24

testified that her sister returned her ballot for her during the 2016 election, but on cross-

25

examination Glover claimed her ballot was returned by her “sister from church,” rather

26

than a family member. Glover testified that her apartment building has outgoing mail,

27

but the slots are too small for the ballot. Although a postal worker collects mail at the

28

building, Glover sometimes forgets to give the postal worker her outgoing mail. Glover

- 28 -

1

testified that others in the community have caregivers, but that she would not feel

2

comfortable giving her ballot to a caregiver. Glover also testified that she was unaware

3

she could request to vote via a special election board. (Tr. 222-25, 228-230, 232-33.)

4

H.B. 2023 does not severely burden voters like Glover, who admittedly can hand

5

her ballot to a postal worker, provided she remembers to do so. Further, if Glover’s

6

mobility issues make it difficult for her to travel to a post office, she can request to vote

7

via a special election board.

8

Democratic Party from assisting her with making those arrangements.

Nothing in H.B. 2023 prevents volunteers from the

9

Magos is a 72-year-old Phoenix resident who prefers to vote by mail. He has a

10

home mailbox but prefers not to use it to send important items because his mailbox has

11

been tampered with in the past. Magos once gave his ballot to a collector because a flood

12

impacted his home and he did not want to leave his wife alone. But in most elections, he

13

either takes his ballot to the post office or drops it off at a polling place. Magos is

14

capable of driving to a polling place and voting in person, and he has family members

15

who could return his ballot if he found himself in need of such a service, though he

16

testified that he “would hate to burden them with one more duty” because “they already

17

do enough for” him. Magos successfully voted in 2016, even though H.B. 2023 was in

18

effect. (Tr. 235, 238-40, 242, 247, 250.)

19

Arias is a registered voter in Phoenix who testified that she once gave her ballot to

20

a collector because her vehicle had broken down. Additionally, Arias voted by early

21

ballot in the 2016 presidential preference and general elections by driving to Democratic

22

Party headquarters and dropping her voted early ballots off there, presumably so

23

volunteers could later deliver those ballots to an appropriate destination. Although Arias

24

testified that the postal service in her neighborhood is unreliable, she did not explain why

25

she could drive her ballots to Democratic Party headquarters but not to a post office, early

26

ballot drop box, polling location, or elections office. (Tr. 1166-68, 1173.)

27

The only early mail voter who testified that she did not vote during the 2016

28

general election was Gilbreath, a 72-year-old Laveen resident. Gilbreath testified that she

- 29 -

1

has mobility issues due to her arthritis. During the 2014 election, Gilbreath gave her

2

early mail ballot to a friend because she waited too long to mail it. Gilbreath voted in the

3

2016 presidential preference election by mailing her early ballot herself. She received an

4

early mail ballot for the general election but did not return it because she waited too long

5

to mail it and was not sure where to go to deliver it in person. Thus, Gilbreath has access

6

to a mailbox; she simply must remember to timely mail her ballot.10 (Tr. 128, 130, 133,

7

135, 142.)

8

In addition to these voters, Plaintiffs designated for admission portions of the

9

deposition testimony of Victor Vasquez, who said that he suffered a heart attack during

10

the 2014 general election and asked a hospital nurse to return his early ballot for him, but

11

she refused. Accordingly, he checked himself out of the hospital on Election Day and

12

had a friend drive him to a polling place, where he cast a provisional ballot that

13

ultimately was not counted because Vasquez was not in his assigned precinct. (Vasquez

14

Dep. 15:18-18:13; 25:7-25.) The Court has concerns about the credibility of Vasquez’s

15

account. If Vasquez had already completed an early mail ballot, it is not clear why he

16

completed an entirely new, provisional ballot at the polling place rather than simply drop

17

off the early ballot he previously completed. Vasquez also stated that in a prior election

18

he gave his ballot to a friend to mail at the post office because he does not trust the

19

outgoing mail service where he lives, but he did not explain whether he easily can go to

20

the post office on his own.

21

In sum, though for voters like those who testified at trial H.B. 2023 might have

22

eliminated a preferred or convenient way of returning an early mail ballot, it does not

23

follow that what H.B. 2023 expects them to do instead is burdensome. The Constitution

24

does not demand “recognition and accommodation of such variable personal preferences,

25

even if the preferences are shown to be shared in higher numbers by members of certain

26
27
28

10

Plaintiffs do not challenge Arizona’s requirement that early mail ballots be
received by the county recorder by 7:00 p.m. on Election Day, which appears to cause
more problems for voters than H.B. 2023’s limitations on ballot collection.

- 30 -

1

identifiable segments of the voting public.”

2

Nor does it require states to prioritize voter convenience above all other regulatory

3

considerations. Id. at 629. H.B. 2023 has no impact on the vast majority of Arizona

4

voters, and the Court lacks sufficient evidence to assess whether the law imposes a more

5

severe burden for discrete subsets of voters. The evidence that was adduced at trial,

6

however, indicates that, for many, ballot collection is used out of convenience and not

7

because the alternatives are particularly difficult.

8

Ohio Democratic Party, 834 F.3d at 630.

2. Burden on Associational Rights

9

In Count V of their Second Amended Complaint, Plaintiffs alleged that H.B. 2023

10

unjustifiably infringes upon Plaintiffs’ associational rights, as distinct from voting rights.

11

(Doc. 233 ¶¶ 112-115.) The parties’ joint proposed pretrial order, however, does not

12

include this claim as a contested issue of fact and law. Instead, the proposed pretrial

13

order states that Plaintiffs challenge H.B. 2023 under the Fourteenth Amendment only

14

“because it imposes burdens on voters that outweigh the state’s interest in this policy.”

15

(Doc. 360 at 7.) Although Plaintiffs’ pretrial brief asserts that H.B. 2023 “infringes on

16

the right to associate,” it does not elaborate further on the issue. (Doc. 359 at 6.)

17

Moreover, Plaintiffs’ proposed findings of fact and conclusions of law contain no

18

proposed factual findings or legal conclusions regarding H.B. 2023’s impact on

19

associational rights. (Doc. 362 ¶ 131.) Defendants did not brief the associational rights

20

issue because, based on the parties’ joint description of contested issues in the joint

21

proposed pretrial order, they understood that Plaintiffs would not be seeking to prove that

22

claim at trial. (Doc. 356 at 11 n.6.) Plaintiffs did not seriously advance this issue at trial,

23

though when asked whether the claim still is at issue, Plaintiffs’ responded affirmatively

24

and explained that the claim is “part and parcel of our Anderson/Burdick claim.” (Tr.

25

1500.)

26

To the extent this claim has not been abandoned, Plaintiffs have offered no

27

evidence or argument that would lead the Court to deviate from the conclusion it reached

28

at the preliminary injunction stage, where Plaintiffs argued that H.B. 2023 burdens the

- 31 -

1

associational rights of groups that encourage and facilitate voting through ballot

2

collection. (Doc. 85 at 16-18.) The Anderson/Burdick framework applies to Plaintiff’s

3

First Amendment claim. Timmons, 520 U.S. at 358. As the party invoking the First

4

Amendment’s protection, however, Plaintiffs bear the additional, threshold burden of

5

proving that it applies. See Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 293

6

n.5 (1984).

7

Conduct, such as collecting a ballot, is not “speech” for purposes of the First

8

Amendment simply because “the person engaging in the conduct intends thereby to

9

express an idea.”

U.S. v. O’Brien, 391 U.S. 367, 376 (1968).

Rather, the First

10

Amendment extends “only to conduct that is inherently expressive.” Rumsfeld v. Forum

11

for Acad. & Institutional Rights, Inc., 547 U.S. 47, 66 (2006). The Court continues to

12

find persuasive the Fifth Circuit’s opinion in Voting for Am. v. Steen, 732 F.3d 382 (5th

13

Cir. 2013), which considered a challenge to various Texas laws that regulated the receipt

14

and delivery of completed voter registration applications. The Fifth Circuit rejected the

15

argument that collecting and delivering voter registration applications were inherently

16

expressive activities protected by the First Amendment. Id. at 392. In doing so, the court

17

agreed that “some voter registration activities involve speech—‘urging’ citizens to

18

register; ‘distributing’ voter registration forms; ‘helping’ voters fill out their forms; and

19

‘asking’ for information to verify registrations were processed successfully.” Id. at 389.

20

It determined, however, that “there is nothing inherently expressive about receiving a

21

person’s completed [voter registration] application and being charged with getting that

22

application to the proper place.” Id. at 392 (internal quotation and citation omitted).

23

Likewise, though many GOTV activities involve First Amendment protected activity,

24

there is nothing inherently expressive or communicative about collecting a voter’s

25

completed early ballot and delivering it to the proper place.

26

Moreover, assuming that H.B. 2023 implicates protected associational rights, it

27

does not impose severe burdens.

28

encouraging, urging, or reminding people to vote, informing and reminding them of

Nothing in H.B. 2023 prevents Plaintiffs from

- 32 -

1

relevant election deadlines, helping them fill out early ballots or request special election

2

boards, or arranging transportation to on-site early voting locations, post offices, county

3

recorder’s offices, or polling places.

4

volunteers remained “free to organize and run the registration drive, persuade others to

5

register to vote, distribute registration forms, and assist others in filling them out”);

6

League of Women Voters of Fla. v. Browning, 575 F. Supp. 2d 1298, 1322 (S.D. Fla.

7

2008) (“[The challenged law] does not place any restrictions on who is eligible to

8

participate in voter registration drives or what methods or means third-party voter

9

registration organizations may use to solicit new voters and distribute registration

See id. at 393 (noting that voter registration

10

applications.

11

process—the handling of voter registration applications by third-party voter registration

12

organizations after they have been collected from applications.”). H.B. 2023 merely

13

regulates who may possess, and therefore return, another’s early ballot. Accordingly,

14

H.B. 2023 no more than minimally burdens Plaintiffs’ associational rights.

15

Instead, [it] simply regulates an administrative aspect of the electoral

3. Justifications

16

Because H.B. 2023 no more than minimally burdens Plaintiffs’ First and

17

Fourteenth Amendment rights, Defendants must show only that it serves important

18

regulatory interests. Wash. State Grange, 552 U.S. at 452. Defendants advance two

19

justifications for H.B. 2023. First, they claim that H.B. 2023 is a prophylactic measure

20

intended to prevent absentee voter fraud by creating a chain of custody for early ballots

21

and minimizing the opportunities for ballot tampering, loss, and destruction. Second,

22

Defendants argue that H.B. 2023 improves and maintains public confidence in election

23

integrity.

24

Fraud prevention and preserving public confidence in election integrity are facially

25

important state regulatory interests.

26

(“Confidence in the integrity of our electoral process is essential to the functioning of our

27

participatory democracy.

28

process and breeds distrust of our government.”); Eu v. S.F. Cty. Democratic Cent.

Purcell v. Gonzalez, 549 U.S. 1, 4 (2006)

Voter fraud drives honest citizens out of the democratic

- 33 -

1

Comm., 489 U.S. 214, 231 (1989) (“A State indisputably has a compelling interest in

2

preserving the integrity of its election process.”); see also Crawford, 553 U.S. at 195

3

(“There is no question about the legitimacy or importance of the State’s interest in

4

counting only the votes of eligible voters. . . . While the most effective method of

5

preventing election fraud may well be debatable, the propriety of doing so is perfectly

6

clear.”).

7

unjustified because (1) there is no evidence of absentee voter fraud perpetrated by ballot

8

collectors or of widespread public perception that ballot collection leads to fraud and (2)

9

H.B. 2023 is not an appropriately tailored means of accomplishing Arizona’s objectives.

10

Plaintiffs do not argue otherwise.

Instead, they argue that H.B. 2023 is

On the first point, there has never been a case of voter fraud associated with ballot

11

collection charged in Arizona.

12

allegations of ballot collection voter fraud have been investigated in Arizona, none of the

13

incidents resulted in a criminal prosecution. (Tr. 834-37 1659, 1680-81, 2163-68, 2185-

14

87, 2202-05; Exs. 81, 372, 400.)

15

perpetrated through ballot collection was presented by or to the Arizona legislature

16

during the debates on H.B. 2023 or its predecessor bills. No Arizona county produced

17

evidence of confirmed ballot collection fraud in response to subpoenas issued in this case,

18

nor has the Attorney General’s Office produced such information. (Ex. 44, 65.)

(Tr. 1682, 1981, 2198.)

Although three specific

No specific, concrete example of voter fraud

19

The Republican National Lawyers Association (“RNLA”) performed a study

20

dedicated to uncovering cases of voter fraud between 2000 and 2011. (Tr. 1868.) The

21

study found no evidence of ballot collection or delivery fraud, nor did a follow-up study

22

through May 2015. (Ex. 91 at 19-20.) Although the RNLA reported instances of

23

absentee ballot fraud, none were tied to ballot collection and delivery. (Tr. 1368-69.)

24

Likewise, the Arizona Republic conducted a study of voter fraud in Maricopa County and

25

determined that, out of millions of ballots cast in Maricopa County from 2005 to 2013, a

26

total of 34 cases of fraud were prosecuted. Of these, 18 involved a felon voting without

27

her rights first being restored. Fourteen involved non-Arizona citizens voting. The study

28

uncovered no cases of fraud perpetrated though ballot collection. (Ex. 91 at 19.)

- 34 -

1

As for public perception of fraud, the legislative record contains no evidence of

2

widespread public concern that ballot collectors were engaging in voter fraud. (Ex. 91 at

3

19.) H.B. 2023’s sponsor, Representative Michelle Ugenti-Rita, was not aware of any

4

polling data indicating that Arizonans lacked confidence in the State’s election system at

5

the time she introduced the bill. (Tr. 1805.)

6

Although there is no direct evidence of ballot collection fraud or of widespread

7

public perception that ballot collection undermined election integrity, Arizona’s

8

legislature is not limited to reacting to problems as they occur, nor is it required to base

9

the laws it passes on evidence that would be admissible in court. See Voting for Am., 732

10

F.3d at 394 (explaining that states “need not show specific local evidence of fraud in

11

order to justify preventative measures”).

12

supporting Arizona’s concerns might be appropriate if H.B. 2023 severely burdened the

13

franchise. But because H.B. 2023’s burdens are at most minimal, the Court’s review is

14

less exacting. Timmons, 520 U.S. at 358.

A more exacting review of the evidence

15

For example, in Crawford the Supreme Court upheld Indiana’s voter identification

16

requirement as a measure designed to prevent in-person voter fraud even though “[t]he

17

record contain[ed] no evidence of any such fraud actually occurring in Indiana at any

18

time in its history.” 553 U.S. at 195. Similarly, in Munro v. Socialist Workers Party, the

19

Supreme Court upheld a Washington law requiring all minor party candidates for partisan

20

office to receive at least one percent of all votes cast during the primary election in order

21

to appear on the general election ballot. 479 U.S. 189 (1986). Washington argued that

22

the law prevented voter confusion from ballot overcrowding by ensuring candidates

23

appearing on the general election ballot had sufficient community support. Id. at 194. In

24

upholding the law, the Supreme Court explained: “We have never required a State to

25

make a particularized showing of the existence of voter confusion, ballot overcrowding,

26

or the presence of frivolous candidates prior to the imposition of reasonable restrictions

27

on ballot access.” Id. at 194-95. Rather, “[l]egislatures . . . should be permitted to

28

respond to potential deficiencies in the electoral process with foresight rather than

- 35 -

1

reactively[.]” Id. at 195; see also Lee v. Va. State Bd. of Elections, 188 F. Supp. 3d 577,

2

609 (E.D. Va. 2016) (“Outlawing criminal activity before it occurs is not only a wise

3

deterrent, but also sound public policy.”), aff’d, 843 F.3d 592 (4th Cir. 2016).

4

Furthermore, many courts have recognized that absentee voting presents a greater

5

opportunity for fraud. See Crawford, 553 U.S. at 225 (Souter, J. dissenting) (noting that

6

“absentee-ballot fraud . . . is a documented problem in Indiana”); Griffin v. Roupas, 385

7

F.3d 1128, 1131 (7th Cir. 2004) (“Voting fraud . . . is facilitated by absentee voting.”);

8

Qualkinbush v. Skubisz, 826 N.E.2d 1181, 1197 (Ill. App. Ct. 2004) (“It is evident that

9

the integrity of the vote is even more susceptible to influence and manipulation when

10

done by absentee ballot.”). Indeed, mail-in ballots by their very nature are less secure

11

than ballots cast in person at polling locations. Accordingly, the Court finds that the

12

regulatory interests Arizona seeks to advance are important.

13

The question then becomes one of means-end tailoring. Because H.B. 2023 does

14

not impose severe burdens, it need not be narrowly tailored to achieve the State’s goals.

15

Nevertheless, the Court still must take into consideration the extent to which Arizona’s

16

important regulatory interests make it necessary to impose those minimal burdens.

17

Burdick, 504 U.S. at 434

18

Plaintiffs contend that H.B. 2023 is not necessary because Arizona law already

19

includes measures designed to ensure the security of early mail ballots, and because H.B.

20

2023 is unlikely to be a useful tool to prevent or deter voter fraud or to preserve public

21

confidence in election integrity.

22

discarding someone else’s ballot all were illegal prior to the passage of H.B. 2023.

23

(Shooter Dep. 51:16-52:5.)

24

knowingly collects voted or unvoted ballots and does not turn those ballots in to an

25

elections official is guilty of a class 5 felony. A.R.S. § 16-1005. Further, Arizona has

26

long made all of the following class 5 felonies: “knowingly mark[ing] a voted or unvoted

27

ballot or ballot envelope with the intent to fix an election;” “receiv[ing] or agree[ing] to

28

receive any consideration in exchange for a voted or unvoted ballot;” possessing

For example, ballot tampering, vote buying, or

Arizona law has long provided that any person who

- 36 -

1

another’s voted or unvoted ballot with intent to sell; “knowingly solicit[ing] the

2

collection of voted or unvoted ballots by misrepresenting [one’s self] as an election

3

official or as an official ballot repository or . . . serv[ing] as a ballot drop off site, other

4

than those established and staffed by election officials;” and “knowingly collect[ing]

5

voted or unvoted ballots and . . . not turn[ing] those ballots in to an election official . . . or

6

any . . . entity permitted by law to transmit post.” A.R.S. §§ 16-1005(a)-(f). The early

7

voting process also includes a number of other safeguards, such as tamper evident

8

envelopes and a rigorous voter signature verification procedure. (Tr. 834-35, 1563-66,

9

1752, 1878, 2209.)

10

Plaintiffs also note that, to the extent Arizona wanted to create a chain of custody

11

for early ballots, the legislature rejected a less restrictive amendment to H.B. 2023

12

proposed by Representative Ken Clark and Senator Martin Quezada, which would have

13

allowed ballot collection if the collector issued a tracking receipt. (Shooter Dep. at

14

50:21-23; Ex. 91 at 12; Ex. 16 at 54.) As enacted, H.B. 2023 is less effective at creating

15

a chain of custody because it allows certain individuals to possess another’s voted early

16

ballot but does not require a record of that collection. (Reagan Dep. 83:25-85:20.) H.B.

17

2023 also is not enforced by county recorders. (Ex. 526 at 5 n.15; Ex. 75.) Instead,

18

county recorders will accept all ballots, even those returned by prohibited possessors

19

under H.B. 2023.

20

Plaintiffs raise fair concerns about whether, as a matter of public policy, H.B.

21

2023 is the best way to achieve Arizona’s stated goals. If H.B. 2023 severely burdened

22

the franchise, and Arizona consequently was required to narrowly tailor the law to

23

achieve compelling ends, Plaintiffs’ arguments would carry more weight. But because

24

H.B. 2023’s burdens are minimal, and the Court’s review consequently less exacting,

25

H.B. 2023’s means-end fit can be less precise.

26

Defendants contend that one of H.B. 2023’s purposes is to reduce the opportunity

27

for early mail ballot fraud by limiting who may possess a voter’s early ballot. They also

28

use the term “fraud” broadly to encompass not just vote tampering, which is amply

- 37 -

1

addressed by other provisions of Arizona law, but also early ballot loss or destruction.

2

By limiting who may possess another’s early ballot, H.B. 2023 reasonably reduces

3

opportunities for early ballots to be lost or destroyed.

4

Although Arizona’s legislature arguably could have addressed this concern

5

through a more narrowly tailored, but also more complex, system of training and

6

registering ballot collectors and requiring tracking receipts or other proof of delivery, the

7

Constitution does not require Arizona to erect such a bureaucracy if the alternative it has

8

chosen is not particularly burdensome. Arizona reasonably chose to limit possession of

9

early ballots to the voter herself, and to a handful of presumptively trustworthy proxies,

10

such as family and household members.

11

recommendation of the bipartisan Commission on Federal Election Reform, chaired by

12

former President Jimmy Carter and former Secretary of State James A. Baker III, which

13

in 2005 wrote:

14

Indeed, H.B. 2023 closely follows the

Fraud occurs in several ways. Absentee ballots remain the
largest source of potential voter fraud. . . . Absentee balloting
is vulnerable to abuse in several ways: . . . Citizens who vote
at home, at nursing homes, at the workplace, or in church are
more susceptible to pressure, overt and subtle, or to
intimidation. Vote buying schemes are far more difficult to
detect when citizens vote by mail. States therefore should
reduce the risks of fraud and abuse in absentee voting by
prohibiting “third-party” organizations, candidates, and
political party activists from handling absentee ballots.

15
16
17
18
19
20

Building Confidence in U.S. Elections § 5.2 (Sept. 2005) (“Carter-Baker Report”),

21

available at https://www.eac.gov/assets/1/6/Exhibit%20M.PDF.11 Though it might not

22
23
24
25
26
27
28

11

The Carter-Baker Report was not offered into evidence by either party. It was
part of the record in Crawford, however, and the Supreme Court cited it favorably. 553
U.S. at 193. It also was cited favorably by Judge Bybee in his dissent from the en banc
Ninth Circuit panel’s November 4, 2016 order temporarily enjoining enforcement of H.B.
2023 pending en banc review of this Court’s order denying a preliminary injunction. See
Feldman v. Ariz. Sec. of State’s Office, 843 F.3d 366, 414 (9th Cir. 2016) (Bybee, J.
dissenting). The Court may take judicial notice of the Carter-Baker Report’s
recommendations pursuant to Federal Rule of Evidence 201. The Carter-Baker Report is
a government document publicly available on the United States Election Assistance
Commission’s website. Though Plaintiffs might disagree with the Carter-Baker Report’s
recommendations, their continued validity, or their relevance to this case, there is no
question that this recommendation was made and is authentic.

- 38 -

1

be the most narrowly tailored provision, H.B. 2023 is one reasonable way to advance

2

what are otherwise important state regulatory interests. Accordingly, H.B. 2023 does not

3

violate the First and Fourteenth Amendments.

4

B. Application to OOP Ballot Policy

5

1. Burden on Voting Rights

6

Arizona consistently is at or near the top of the list of states that collect and reject

7

the largest number of provisional ballots each election. (Ex. 95 at 23-25.) In 2012 alone

8

“[m]ore than one in every five [Arizona in-person] voters . . . was asked to cast a

9

provisional ballot, and over 33,000 of these—more than 5 percent of all in-person ballots

10

cast—were rejected. No other state rejected a larger share of its in-person ballots in

11

2012.” (Ex. 95 at 24-25.) Interstate comparisons of provisional voting are complicated,

12

however, because states use provisional ballots in different ways, and some states do not

13

utilize provisional voting in any form.

14

proportion of provisional votes are rejected for reasons not specified or because the voter

15

voted in an incorrect jurisdiction, as compared to Arizona. Moreover, the overall number

16

of provisional ballots in Arizona, both as a percentage of the registered voters and as a

17

percentage of the number of ballots cast, has consistently declined.

For example, nationwide a much higher

18

One of the most frequent reasons that provisional ballots are rejected in Arizona is

19

because they are cast OOP. (Ex. 95 at 22-29.) Arizona’s rejection of OOP ballots,

20

however, has no impact on the vast majority of Arizona voters. Early mail voting is the

21

most popular method of voting in Arizona, accounting for approximately 80 percent of all

22

ballots cast in the 2016 election. Voters who cast early mail ballots are unaffected by

23

Arizona’s policy to not count OOP ballots. Likewise, this policy has no impact on voters

24

in Graham, Greenlee, Cochise, Navajo, Yavapai, and Yuma counties, which have

25

adopted the vote center model.

26

Moreover, the vast majority of in-person voters successfully vote in their assigned

27

precincts, and OOP voting has consistently declined as a percentage of the total ballots

28

cast in Arizona. In the 2008 general election, Arizona voters cast 14,885 OOP ballots out

- 39 -

1

of the 2,320,851 ballots cast statewide, meaning OOP ballots constituted 0.64 percent of

2

all votes cast in that election. In the 2012 general election, Arizona voters cast 10,979

3

OOP ballots out of the 2,323,579 ballots cast statewide, accounting for 0.47 percent of all

4

votes cast. In that same election, 1,542,855 Arizona voters submitted early ballots, and

5

more than 99 percent were counted. In the 2016 general election, Arizona voters cast

6

3,970 OOP ballots out of the 2,661,497 ballots cast statewide, representing 0.15 percent

7

of all votes cast. Since 2008, OOP voting during general presidential elections has

8

declined 73 percent statewide, dropping from 14,885 in 2008 to 3,970 in the 2016

9

election. (Tr. 1927-32; Exs. 578, 581.)

10

OOP voting has declined in midterm elections, as well. In the 2010 general

11

election, Arizona voters cast 4,919 OOP ballots out of the 1,750,840 ballots cast

12

statewide, constituting 0.28 percent of all votes cast. By comparison, in the 2014 general

13

election, Arizona voters cast 3,582 OOP ballots out of the 1,537,671 ballots cast

14

statewide, constituting 0.23 percent of all votes cast. During this same period, the

15

number of registered voters in Arizona increased as follows: 2,987,451 in 2008;

16

3,146,418 in 2010; 3,124,712 in 2012; 3,235,963 in 2014; and 3,588,466 in 2016. (Exs.

17

577, 578.)

18

These trends also hold true at the county level. For example, Maricopa County

19

(Arizona’s most populous) has experienced a consistent decline in the number of OOP

20

ballots, both in terms of raw numbers and as a percentage of the total ballots cast. In the

21

2008 general election, Maricopa County voters cast 9,159 OOP ballots out of the

22

1,380,571 ballots cast countywide, accounting for 0.66 percent of the all votes cast. In

23

the 2012 general election, Maricopa County voters cast 7,529 OOP ballots out of the

24

1,390,836 ballots cast countywide, representing 0.54 percent of all votes. In the 2016

25

general election, Maricopa County voters cast 2,197 OOP ballots out of the 1,608,875

26

ballots cast countywide, representing 0.14 percent of all votes cast. Likewise, in the 2010

27

general midterm election, Maricopa County voters cast 3,527 OOP ballots out of the

28

1,004,125 ballots cast countywide, accounting for 0.35 percent of all votes. In the 2014

- 40 -

1

general election, Maricopa County voters cast 2,781 OOP ballots out of the 877,187

2

ballots cast countywide, constituting 0.32 percent of all votes. Between 2008 and 2016,

3

Maricopa County had a staggering decrease of 76 percent in the raw number of OOP

4

ballots. During this same period, the number of registered voters in Maricopa County

5

increased as follows: 1,730,886 in 2008; 1,851,956 in 2010; 1,817,832 in 2012;

6

1,935,729 in 2014; and 2,161,716 in 2016. (Exs. 579, 582.)

7

Pima County (Arizona’s second most populous) also has experienced a consistent

8

decline in OOP voting. In the 2008 general election, Pima County voters cast 3,227 OOP

9

ballots out of the 397,503 ballots cast countywide, accounting for 0.81 percent of all

10

votes. In the 2012 general election, Pima County voters cast 2,212 OOP ballots out of

11

the 385,725 ballots cast countywide, accounting for 0.57 percent of all votes. In the 2016

12

general election, Pima County voters cast 1,150 OOP ballots out of the 427,102 ballots

13

cast countywide, representing 0.27 percent of all votes. As for Pima County midterm

14

elections, in the 2010 general election Pima County voters cast 641 OOP ballots out of

15

the 318,995 ballots cast countywide, or 0.20 percent of all votes. By comparison, in the

16

2014 general election, Pima County voters cast just 371 OOP ballots out of the 274,449

17

ballots cast countywide, constituting 0.14 percent of the total ballots. The raw number of

18

OOP ballots thus dropped by 64 percent in Pima County between 2008 and 2016. During

19

this same period, the number of registered voters in Pima County increased as follows:

20

498,777 in 2008; 486,697 in 2010; 494,630 in 2012; 497,542 in 2014; and 544,270 in

21

2016. (Exs. 580, 583.)

22

In light of these figures, and much like their H.B. 2023 claim, Plaintiffs’ challenge

23

to Arizona’s treatment of OOP ballots is best described as follows: Arizona’s rejection of

24

OOP ballots has no impact on the vast majority of Arizona voters, though a small subset

25

of voters is affected more often because of their special circumstances. But Plaintiffs’

26

contention that Arizona’s rejection of OOP ballots severely burdens this small subset of

27

voters is unavailing for two independent reasons.

28

First, Plaintiffs do not directly challenge the electoral practices actually

- 41 -

1

responsible for higher rates of OOP voting. For example, high rates of residential

2

mobility are associated with higher rates of OOP voting. Almost 70 percent of Arizonans

3

have changed their residential address in the decade between 2000 and 2010, the second

4

highest rate of any state. The vast majority of Arizonans who moved in the last year

5

moved to another address within their current city of residence and, compared with other

6

states, Arizona has the second highest rate of within-city moves. Most of these within-

7

city moves took place in Maricopa and Pima Counties. (Ex. 95 at 11-12.) Relatedly,

8

rates of OOP voting are higher in neighborhoods where renters make up a larger share of

9

householders. (Ex. 96 at 41.) One significant reason residential mobility tends to result

10

in higher rates of OOP voting is because voters who move sometimes neglect to timely

11

update their voter registration. (See, e.g., Tr. 602-06.) Relatedly, voters registered for

12

PEVL who move and do not update their address information will not have their early

13

ballot forwarded to their new address. Arizona in-person voters are more likely to vote

14

OOP if they have signed up for the PEVL and have moved. (See, e.g., Tr. 124, 987-89.)

15

Additionally, changes in polling locations from election to election, inconsistent

16

election regimes used by and within counties, and placement of polling locations all tend

17

to increase OOP voting rates. (Ex. 95 at 12-15, 26-27, 44-52, 54-58.) In Maricopa

18

County, between 2006 and 2008 at least 43 percent of polling locations changed from one

19

year to the next. Likewise, approximately 40 percent of Maricopa County’s active

20

registered voters’ polling locations changed between 2010 and 2012.

21

Maricopa County polling locations and election regimes continued to occur in 2016,

22

when Maricopa County experimented with 60 vote centers for the presidential preference

23

election, then reverted to a precinct-based system with 122 polling locations for the May

24

special election, and then implemented over 700 assigned polling places in the August

25

primary and November general elections. The OOP voting rate was 40 percent higher for

26

voters who had experienced such polling place changes. (Ex. 95 at 14-15, 56-57.)

27

Further, some individual voters testified that they arrived at an incorrect polling place but

28

were not redirected by poll workers to the correct location, nor were the implications of

- 42 -

Changes in

1

casting a provisional ballot explained. These voters stated that they would have gone to

2

the correct polling location had they been so advised. (Tr. 120, 265-66, 352-54, 493,

3

935-36.)

4

Plaintiffs do not challenge as unconstitutional the manner in which Arizona and its

5

counties allocate or relocate polling places, inform voters of their assigned precincts, or

6

train poll workers. They do not challenge Arizona’s requirement that voters update their

7

voter registrations after moving to a new address. Nor do Plaintiffs challenge Arizona’s

8

use of the precinct-based system, though the logical implication of their argument is that

9

Arizona may utilize a precinct-based system but cannot enforce it as to races for which an

10

OOP voter otherwise would be eligible to vote (usually so-called “top of the ticket”

11

races). (Tr. 1495-96.) Instead, Plaintiffs challenge what Arizona does with OOP ballots

12

after they have been cast. But there is no evidence that it will be easier for voters to

13

identify their correct precincts if Arizona eliminated its prohibition on counting OOP

14

ballots. Though the consequence of voting OOP might make it more imperative for

15

voters to correctly identify their precincts, it does not increase the burdens associated

16

with doing so.

17

Second, the burdens imposed by precinct-based voting—a system which, again,

18

Plaintiffs do not directly challenge—are not severe.

19

requires voters to locate and travel to their assigned precincts, which are ordinary burdens

20

traditionally associated with voting.12

21

04CV7709, 2004 WL 2360485, at *14 (Colo. Dist. Ct. Oct. 18, 2004) (“[I]t does not

22

seem to be much of an intrusion into the right to vote to expect citizens, whose judgment

23

we trust to elect our government leaders, to be able to figure out their polling place.”); see

24
25
26
27
28

Precinct-based voting merely

See Colo. Common Cause v. Davidson, No.

12

Plaintiffs again conflate the burdens imposed by the (indirectly) challenged
practice with the socioeconomic circumstances that can make those burdens more
difficult for certain subsets of voters to surmount. Arizona’s precinct-based system does
not impose residential instability, transportation difficulties, or informational deficits on
any voter. These circumstances exist independent of the precinct-based system. Instead,
the precinct-based system imposes on voters the burden of locating and travelling to an
assigned precinct, which might be more difficult for some voters to do because of their
circumstances.

- 43 -

1

also Serv. Emps. Int’l Union Local 1 v. Husted, 698 F.3d 341, 344 (6th Cir. 2012)

2

(explaining that voters cannot be absolved “of all responsibility for voting in the correct

3

precinct or correct polling place by assessing voter burden solely on the basis of the

4

outcome—i.e., the state’s ballot validity determination”).

5

Moreover, Arizona does not make it needlessly difficult for voters to find their

6

assigned precincts.

7

(“SPAE”) found that none of the survey respondents for Arizona reported that it was

8

“very difficult” to find their polling places. By comparison, several other states had

9

respondents who reported that it was very difficult to find their polling places. The 2016

10

SPAE also reported that approximately 94 percent of the Arizona respondents thought it

11

was very easy or somewhat easy to find their polling places. (Tr. 1350-51.)

Indeed, a 2016 Survey of Performance of American Elections

12

In Arizona counties with precinct-based systems, voters generally are assigned to

13

precincts near where they live, and county officials consider access to public

14

transportation when assigning polling places. (Tr. 1570-73.) Arizona voters also can

15

learn of their assigned precincts in a variety of ways. (See Ex. 526 at 11-18.)

16

If precincts or polling places have been altered since the previous election,

17

registered voters are sent a mailing informing them of this fact and of where their new

18

polling places are located. (Tr. 1575-76.) State law requires that election officials send

19

each household with a registered voter who is not on the PEVL a sample ballot at least

20

eleven days prior to election day, A.R.S. § 16-510(C), which contains instructions and

21

identifies their polling location. (Doc. 361 ¶ 52.) The Secretary of State’s Office

22

operates several websites that make voter-specific polling place information available and

23

allow the Secretary’s staff to respond directly to voter inquiries. The Secretary of State’s

24

Office also mails a publicity pamphlet to voters, which includes information on how to

25

locate their correct precincts. This information is provided in English and Spanish. The

26

Secretary also uses social media, town halls, and live events (such as county and state

27

fairs) to register voters and answer questions.

28

In addition, several Arizona counties, including Maricopa and Pima Counties,

- 44 -

1

operate online polling place locators that are available in English and Spanish. Voters

2

also can learn their assigned polling locations by calling the office of the county recorder

3

for the county in which they reside. Counties spread awareness about polling place

4

locations and the consequences of OOP voting through news and social media. This

5

information is communicated in both English and Spanish. Some counties—including

6

the state’s most populous, Maricopa and Pima—post signs at polling places informing

7

voters that OOP ballots will not be counted. (Tr. 1586-88; Ex. 368.) Poll workers also

8

are trained to direct voters who appear at an incorrect polling location to their correct

9

polling location and to notify such voters that their votes will not be counted if they vote

10

with a provisional ballot at the wrong location.

11

The Arizona Citizens Clean Elections Commission (“CCEC”) operates a website

12

in English and Spanish that provides a tool for voters to determine their polling place.

13

The CCEC also engages in advertising to help educate voters on where to vote. Partisan

14

groups, such as the ADP and political campaigns, also help educate voters on how to find

15

their assigned polling places. (Tr. 1575-76.)

16

In sum, Arizona’s rejection of OOP ballots has no impact on the vast majority of

17

voters. Although a small and ever-dwindling subset of voters still vote OOP, how

18

Arizona treats OOP ballots after they have been cast does not make it difficult for these

19

voters to find and travel to their correct precincts. To the extent Plaintiffs’ claim may

20

properly be considered as an indirect challenge to Arizona’s strictly enforced precinct-

21

based system, the burdens imposed on voters to find and travel to their assigned precincts

22

are minimal and do not represent significant increases in the ordinary burdens

23

traditionally associated with voting. Moreover, for those who find it too difficult to

24

locate their assigned precinct, Arizona offers generous early mail voting alternatives.13

25

2. Justifications

26
27
28

13

If a voter is capable of travelling to an incorrect precinct, she certainly is
capable of mailing an early ballot. Moreover, early mail voters may drop their ballots off
at any polling place, even one to which they are not assigned.

- 45 -

1

Weighing against the minimal burdens imposed by precinct-based voting are

2

numerous important state regulatory interests. Precinct-based voting serves an important

3

planning function for Arizona counties by helping them estimate the number of voters

4

who may be expected at any particular precinct, which allows for better allocation of

5

resources and personnel. In turn, orderly administration of elections helps to increase

6

voter confidence in the election system and reduces wait times. (Tr. 1608-10, 1896-913.)

7

Because elections involve many different overlapping jurisdictions, the precinct-based

8

system also ensures that each voter receives a ballot reflecting only the races for which

9

that person is entitled to vote. Precincts must be created, and ballots printed, so that the

10

residential address of every voter is connected to the correct assortment of local elected

11

officials. The system thus promotes voting for local candidates and issues and helps

12

make ballots less confusing by not providing voters with ballots that include races for

13

which they are not eligible to vote. (Ex. 95 at 10; Doc. 361 ¶ 47.)

14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

Indeed, other courts have recognized these numerous and significant advantages:
[Precinct-based voting] caps the number of voters attempting
to vote in the same place on election day; it allows each
precinct ballot to list all of the votes a citizen may cast for all
pertinent federal, state, and local elections, referenda,
initiatives, and levies; it allows each precinct ballot to list
only those votes a citizen may cast, making ballots less
confusing; it makes it easier for election officials to monitor
votes and prevent election fraud; and it generally puts polling
places in closer proximity to voter residences.
Sandusky Cty. Democratic Party v. Blackwell, 387 F.3d 565, 569 (6th Cir. 2004).
Plaintiffs do not quarrel with the importance or legitimacy of these interests or
contest that precinct-based voting brings significant advantages. Instead, they argue that
Arizona need not reject OOP ballots in their entirety to accomplish these goals. Plaintiffs
contend that Arizona can just as easily accomplish these goals and reap these benefits by
partially counting OOP ballots, accepting votes in races for which the voter is eligible to
vote and rejecting votes in races for which the voter is not.
Counting OOP ballots is administratively feasible. Twenty states partially count
OOP ballots. (Ex. 94 at 32-33.) These include the neighboring states of California, Utah,
- 46 -

1

and New Mexico. Cal. Elec. Code §§ 14310(a)(3), 14310 (c)(3), 15350; Utah Code Ann.

2

§ 20A-4-107(1)(b)(iii), 2(a)(ii), 2(c); N.M. Stat. Ann § 1-12-25.4(F); N.M. Admin. Code

3

1.10.22.9(N).

4

processes for counting only the offices for which the OOP voter is eligible to vote. Some

5

states, such as New Mexico, use a hand tally procedure, whereby a team of elections

6

workers reviews each OOP ballot, determines the precinct in which the voter was

7

qualified to vote, and marks on a tally sheet for that precinct the votes cast for each

8

eligible office.

9

California, use a duplication method, whereby a team of elections workers reviews each

10

OOP ballot, determines the precinct in which the voter was qualified to vote, obtains a

11

new paper ballot for the correct precinct, and duplicates the votes cast on the OOP ballot

12

onto the ballot for the correct precinct. Only the offices that appear on both the OOP

13

ballot and the ballot for the correct precinct are copied. The duplicated ballot then is

14

scanned through the optical scan voting machine and electronically tallied. (Tr. 777-81.)

Elections administrators in these and other states have established

See N.M. Admin Code 1.10.22.9(H)-(N).

Other states, such as

15

Arizona has a similar duplication procedure that it uses to process certain types of

16

ballots that cannot be read by an optical scan voting machine, such as ballots that are

17

damaged, marked with the wrong color pen, or submitted to the county recorder by a

18

military or overseas voter via facsimile. (Tr. 1564-66; Ex. 455 at 177-78.) Arizona also

19

uses the duplication procedure to process some provisional ballots cast by voters who are

20

eligible to vote in federal elections, but whom Arizona does not permit to vote in state

21

elections. (Ex. 455 at 187.) This duplication procedure takes about twenty minutes per

22

ballot. (Tr. 1604-606.)

23

If strict scrutiny applied and Arizona were required to narrowly tailor its precinct

24

enforcement to achieve compelling state interests, Plaintiffs’ critiques might carry more

25

weight. But in light of the minimal burdens associated with the precinct-based system,

26

Arizona’s policy need not be the narrowest means of enforcement.

27

Moreover, Plaintiffs are incorrect that Arizona can accomplish all of its goals

28

without its strict enforcement regime. If voters in precinct-based counties can have their

- 47 -

1

ballots counted for statewide and countywide races even if they vote in the wrong

2

precinct, they will have far less incentive to vote in their assigned precincts and might

3

decide to vote elsewhere. Other voters might incorrectly believe that that they can vote at

4

any location and receive the correct ballot. Voters might also be nefariously directed to

5

vote elsewhere. North Carolina, for example, has experienced a problem with “political

6

organizations intentionally transporting voters to the wrong precinct.” See N.C. State

7

Conf. of the NAACP v. McCrory, 182 F. Supp. 3d 320, 461 (M.D.N.C. 2016) rev’d on

8

other grounds, 831 F.3d 204 (4th Cir. 2016). This, in turn, would undermine both the

9

ability of Arizona counties to accurately estimate the number of voters who may be

10

expected at any particular precinct and allocate appropriate resources and personnel, and

11

Arizona’s goal of promoting voting for local candidates. Consequently, if OOP ballots

12

are partially counted in Arizona, candidates for local office will have to expend resources

13

to educate voters on why it nevertheless is important to vote within their assigned

14

precincts. Moreover, requiring counties to review all OOP ballots for any given election

15

and determine the specific contests in which each voter was eligible to vote would

16

impose a significant financial and administrative burden on Maricopa and Pima Counties

17

because of their high populations.

18

Plaintiffs’ requested relief essentially would transform Arizona’s precinct-based

19

counties, including its two most populous, into quasi-vote-center counties. But the vote-

20

center model is not appropriate for every jurisdiction.

21

polling places, it can be difficult for counties to predict the number of voters at each vote

22

center.

23

corresponding decreases in turnout, due to the potential for uneven distribution of voters.

24

(Tr. 1607-611, 1896-913.) Plaintiffs’ requested relief therefore would deprive precinct-

25

based counties of the full range of benefits that correspond with the precinct-based

26

system.

Compared to precinct-based

Consequently, vote centers can cause voter wait times to increase, with

27

Precinct-based voting is a quintessential time, place, and manner election

28

regulation. Arizona’s policy to not count OOP ballots is one mechanism by which it

- 48 -

1

enforces and administers this precinct-based system to ensure that it reaps the full extent

2

of its benefits. This policy is sufficiently justified in light of the minimal burdens it

3

imposes.

4

Fourteenth Amendment.

5

VI. SECTION 2 OF THE VOTING RIGHTS ACT (RESULTS TEST)

Accordingly, Arizona’s rejection of OOP ballots does not violate the

6

“Inspired to action by the civil rights movement, Congress responded in 1965 with

7

the Voting Rights Act.” Shelby Cty. v. Holder, 570 U.S. 529, 536 (2013). In its original

8

form, § 2 of the VRA prohibited all states from enacting any “standard, practice, or

9

procedure . . . imposed or applied . . . to deny or abridge the right of any citizen of the

10

United States to vote on account of race or color.” Id. (quoting § 2, 79 Stat. 437).

11

“At the time of passage of the Voting Rights Act of 1965, § 2, unlike other

12

provisions of the Act, did not provoke significant debate in Congress because it was

13

viewed largely as a restatement of the Fifteenth Amendment.” Chisom v. Roemer, 501

14

U.S. 380, 392 (1991). The Supreme Court took a similar view, holding in a 1980

15

plurality opinion that “the language of § 2 no more than elaborates upon that of the

16

Fifteenth Amendment,” and therefore § 2 is violated only if a state enacted the challenged

17

law with the intent to discriminate on account of race or color. City of Mobile v. Bolden,

18

446 U.S. 55, 60-62 (1980) (plurality opinion).

19

In 1982, in response to the Supreme Court’s opinion in Bolden, “Congress

20

substantially revised § 2 to make clear that a violation could be proved by showing

21

discriminatory effect alone[.]” Thornburg v. Gingles, 478 U.S. 30, 35 (1986). In its

22

current form, § 2 provides:

23
24
25
26
27
28

(a) No voting qualification or prerequisite to voting or
standard, practice, or procedure shall be imposed or applied
by any State or political subdivision in a manner which
results in a denial or abridgement of the right of any citizen of
the United States to vote on account of race or color, or in
contravention of the guarantees set forth in section
10303(f)(2) of this title, as provided in subsection (b).
(b) A violation of subsection (a) is established if, based on the
totality of circumstances, it is shown that the political
processes leading to nomination or election in the State or
political subdivision are not equally open to participation by
- 49 -

members of a class of citizens protected by subsection (a) in
that its members have less opportunity than other members of
the electorate to participate in the political process and to
elect representatives of their choice.

1
2
3
4

52 U.S.C. § 10301. To succeed on a § 2 claim, a plaintiff now may show either that the

5

challenge law was enacted with the intent to discriminate on account of race or color, or

6

that “under the totality of the circumstances, a challenged election law or procedure ha[s]

7

the effect of denying a protected minority equal chance to participate in the electoral

8

process.” Gingles, 478 U.S. at 44 n.8. “The essence of a § 2 claim” brought under the

9

so-called “effects” or “results test” “is that a certain electoral law, practice, or structure

10

interacts with social and historical conditions to cause an inequality in the opportunities

11

enjoyed by [minority and non-minority] voters to elect their preferred representatives.”

12

Id. at 47.

13

When determining whether, under the totality of the circumstances, a challenged

14

voting practice interacts with social and historical conditions to cause inequality in the

15

electoral opportunities of minority and non-minority voters, courts may consider, as

16

relevant, the following factors derived from the Senate Report accompanying the 1982

17

amendments to the VRA (“Senate Factors”):

18
19
20
21
22
23
24
25
26
27
28

1. the extent of any history of official discrimination in the
state or political subdivision that touched the right of the
members of the minority group to register, to vote, or
otherwise to participate in the democratic process;
2. the extent to which voting in the elections of the state or
political subdivision is racially polarized;
3. the extent to which the state or political subdivision has
used unusually large election districts, majority vote
requirements, anti-single shot provisions, or other voting
practices or procedures that may enhance the opportunity for
discrimination against the minority group;
4. if there is a candidate slating process, whether the members
of the minority group have been denied access to that process;
5. the extent to which members of the minority group in the
state or political subdivision bear the effects of discrimination
in such areas as education, employment and health, which
hinder their ability to participate effectively in the political
process;
- 50 -

6. whether political campaigns have been characterized by
overt or subtle racial appeals; [and]

1
2

7. the extent to which members of the minority group have
been elected to public office in the jurisdiction.

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

Id. at 36-37 (quoting S. Rep. No. 97-417, at 28-29 (1982)). Courts also may consider
“whether there is a significant lack of responsiveness on the part of elected officials to the
particularized needs of the members of the minority group,” and “whether the policy
underlying the state or political subdivision’s use of such voting qualification,
prerequisite to voting, or standard, practice or procedure is tenuous.” Id.; see also
Houston Lawyers’ Ass’n v. Attorney Gen. of Tex., 501 U.S. 419, 426-27 (1991)
(explaining that courts may consider a state’s “justification for its electoral system”).
“[T]here is no requirement that any particular number of factors be proved, or that a
majority of them point one way or the other.” Gingles, 478 U.S. at 45.
Until relatively recently, “Section 2’s use . . . has primarily been in the context of
vote-dilution cases,” which “‘involve challenges to methods of electing representatives—
like redistricting or at-large districts—as having the effect of diminishing minorities
voting strength.’” League of Women Voters of N.C. v. N.C., 769 F.3d 224, 239 (4th Cir.
2014) (quoting Ohio State Conference of N.A.A.C.P. v. Husted, 768 F.3d 524, 554 (6th
Cir. 2014), vacated on other grounds by 2014 WL 10384647 (6th Cir. 2014)). Gingles
itself was a vote dilution case. “While vote-dilution jurisprudence is well-developed,
numerous courts and commentators have noted that applying Section 2’s ‘results test’ to
vote-denial claims is challenging, and a clear standard for its application has not been
conclusively established.”14 Ohio Democratic Party, 834 F.3d at 636; see also Veasey v.
Abbott, 830 F.3d 216, 243-44 (5th Cir. 2016) (“Although courts have often applied the
Gingles factors to analyze claims of vote dilution . . . there is little authority on the proper
test to determine whether the right to vote has been denied or abridged on account of
race.”); League of Women Voters, 769 F.3d at 239 (“[T]here is a paucity of appellate case
14

A “vote denial” claim generally refers “to any claim that is not a vote dilution
claim.” Ohio State Conference, 768 F.3d at 554

- 51 -

1

law evaluating the merits of Section 2 claims in the vote-denial context.”); Ohio State

2

Conference, 768 F.3d at 554 (“A clear test for Section 2 vote denial claims . . . has yet to

3

emerge.”); Simmons v. Galvin, 575 F.3d 24, 41-42 n.24 (1st Cir. 2009) (“‘While Gingles

4

and its progeny have generated a well-established standard for vote dilution, a

5

satisfactory test for vote denial cases under Section 2 has yet to emerge . . . [and] the

6

Supreme Court’s seminal opinion in Gingles . . . is of little use in vote denial cases.’”

7

(quoting Daniel P. Tokaji, The New Vote Denial: Where Election Reform Meets the

8

Voting Rights Act, 57 S.C. L. Rev. 689, 709 (2006))); Janai S. Nelson, The Causal

9

Context of Disparate Vote Denial, 54 B.C. L. Rev. 579, 595 (2013) (“[T]he legal

10

contours of vote denial claims remain woefully underdeveloped as compared to vote

11

dilution claims.”). Indeed, some of the Senate Factors cited by the Gingles Court as

12

relevant to the totality of the circumstances inquiry do not seem particularly germane to

13

vote denial claims. Cf. Gingles, 478 U.S. at 45 (“While the enumerated factors will often

14

be pertinent to certain types of § 2 violations, particularly to vote dilution claims, other

15

factors may also be relevant and considered.”); see Frank v. Walker, 768 F.3d 744, 752-

16

55 (7th Cir. 2014) (questioning usefulness of Senate Factors in vote denial claims); Ohio

17

Democratic Party, 834 F.3d at 638 (explaining that totality of circumstances inquiry in

18

vote denial cases is “potentially informed by the ‘Senate Factors’ discussed in Gingles”

19

(emphasis added)).

20

Several circuit courts that recently have analyzed vote denial claims have adopted

21

the following two-part framework based on the text of § 2 and the Supreme Court’s

22

guidance in Gingles:

23
24
25
26
27
28

First, the challenged standard, practice, or procedure must
impose a discriminatory burden on members of a protected
class, meaning that members of the protected class have less
opportunity than other members of the electorate to
participate in the political process and to elect representatives
of their choice.
Second, that burden must in part be caused by or linked to
social and historical conditions that have or currently produce
discrimination against members of the protected class.

- 52 -

1

League of Women Voters, 769 F.3d at 240 (internal quotations and citations omitted); see

2

also Veasey, 830 F.3d at 244; Ohio Democratic Party, 834 F.3d at 636; Frank, 768 F.3d

3

at 754-55 (adopting two-park framework for sake of argument but expressing skepticism

4

of the second step “because it does not distinguish between discrimination by [the state]

5

from other persons’ discrimination”). The Ninth Circuit likewise endorsed this two-part

6

framework in the context of this case. Feldman v. Ariz. Sec. of State’s Office, 840 F.3d

7

1057, 1070 (9th Cir. 2016); id. at 1091 (Thomas, C.J. dissenting); Feldman, 843 F.3d at

8

367.

9

“The first part of this two-prong framework inquires about the nature of the

10

burden imposed and whether it creates a disparate effect[.]” Veasey, 830 F.3d at 244.

11

Drawing on the Supreme Court’s guidance in Gingles, “[t]he second part . . . provides the

12

requisite causal link between the burden on voting rights and the fact that this burden

13

affects minorities disparately because it interacts with social and historical conditions that

14

have produced discrimination against minorities currently, in the past, or both.” Id. That

15

is, “the second step asks not just whether social and historical conditions ‘result in’ a

16

disparate impact, but whether the challenged voting standard or practice causes the

17

discriminatory impact as it interacts with social and historical conditions.”

18

Democratic Party, 834 F.3d at 638 (emphasis in original).

19
20
21
22
23

Ohio

Although proving a violation of § 2 does not require a
showing of discriminatory intent, only discriminatory results,
proof of a causal connection between the challenged voting
practice and a prohibited result is crucial. Said otherwise, a §
2 challenge based purely on a showing of some relevant
statistical disparity between minorities and whites, without
any evidence that the challenged voting qualification causes
that disparity, will be rejected.

24

Gonzales v. Arizona, 677 F.3d 383, 405 (9th Cir. 2012) (internal quotations and citations

25

omitted); see also Smith v. Salt River Project Agric. Improvement & Power Dist., 109

26

F.3d 586, 595 (9th Cir. 1997).

27

A close reading of the decisions of courts that recently have grappled with vote

28

denial claims reveals two important nuances of the results test. The first bears on the

- 53 -

1

meaning of “disparity” as it has been used in vote denial cases. “No state has exactly

2

equal registration rates, exactly equal turnout rates, and so on, at every stage of its voting

3

system.” Frank, 768 F.3d at 754. Perfect racial parity is unlikely to exist in any aspect of

4

a state’s election system, which is to say it is unlikely that minorities and non-minorities

5

will be impacted by laws in perfect proportion to their representation in the overall voting

6

population. Unless the VRA is to be interpreted to sweep away all elections regulations,

7

some degree of disproportionality must be tolerable.

8

Therefore, not every disparity between minority and non-minority voters is

9

cognizable under the VRA. Rather, to be cognizable the disparity must be meaningful

10

enough to work “an inequality in the opportunities enjoyed by [minority as compared to

11

non-minority] voters to elect their preferred representatives.” Gingles, 478 U.S. at 47;

12

see Gonzales, 677 F.3d at 405 (suggesting that disparity must be “relevant”).

13

example, the Seventh Circuit rejected the notion that:

14
15
16
17
18
19
20
21
22

For

if whites are 2% more likely to register than blacks, then the
registration system top to bottom violates § 2; and if white
turnout on election day is 2% higher, then the requirement of
in-person voting violates § 2. Motor-voter registration, which
makes it simple for people to register by checking a box when
they get drivers’ licenses, would be invalid, because black
and Latino citizens are less likely to own cars and therefore
less likely to get drivers’ licenses. . . . Yet it would be
implausible to read § 2 as sweeping away almost all
registration and voting rules. It is better to understand § 2(b)
as an equal-treatment requirement (which is how it reads)
than as an equal-outcome command[.]
Frank, 768 F.3d at 754.
The second nuance bears on the definition of “impact” or “effect.”

To be

23

cognizable, the challenged voting practice must “impose a discriminatory burden,”

24

League of Women Voters, 769 F.3d at 240, and not merely result in a “disproportionate

25

impact,” Salt River Project, 109 F.3d at 595. Section 2 “does not sweep away all election

26

rules that result in a disparity in the convenience of voting.” Lee v. Va. State Bd. of

27

Elections, 843 F.3d 592, 601 (4th Cir. 2016). A contrary interpretation would require the

28

Court to accept:

- 54 -

1
2
3
4
5
6
7

an unjustified leap from the disparate inconveniences that
voters face when voting to the denial or abridgment of the
right to vote. Every decision that a State makes in regulating
its elections will, inevitably, result in somewhat more
inconvenience for some voters than for others. For example,
every polling place will, by necessity, be located closer to
some voters than to others. To interpret § 2 as prohibiting
any regulation that imposes a disparate inconvenience 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. Similarly, motor-voter registration
would be found to be invalid [if] members of [a] protected
class were less likely to possess a driver’s license. Yet, courts
have also correctly rejected that hypothetical.

8
9

Id.; see also N.E. Ohio Coal. for the Homeless, 837 F.3d at 628 (“A law cannot

10

disparately impact minority voters if its impact is insignificant to begin with.”); Ohio

11

Democratic Party, 834 F.3d at 623 (“[W]hile the challenged regulation may slightly

12

diminish the convenience of registration and voting, it applies even-handedly to all

13

voters, and, despite the change, Ohio continues to provide generous, reasonable, and

14

accessible voting options to all Ohioans. The issue is not whether some voter somewhere

15

would benefit from six additional days of early voting or from the opportunity to register

16

and vote at the same time. Rather, the issue is whether the challenged law results in a

17

cognizable injury under the Constitution or the Voting Rights Act.”); Frank, 843 F.3d at

18

753 (“[U]nless the State of Wisconsin made it ‘needlessly hard’ to obtain the requisite

19

photo identification for voting, this requirement did not result in a ‘denial’ of anything by

20

Wisconsin, as § 2(a) requires.”); Jacksonville Coal. for Voter Prot. v. Hood, 351 F. Supp.

21

2d 1326, 1335 (M.D. Fla. 2004) (“While it may be true that having to drive to an early

22

voting site and having to wait in line may cause people to be inconvenienced,

23

inconvenience does not result in a denial of ‘meaningful access to the political process.’”

24

(quoting Osburn v. Cox, 369 F.2d 1283, 1289 (11th Cir. 2004)); Glover v. S.C.

25

Democratic Party, No. C/A 4-04-CV-2171-25, 2004 WL 3262756, at *6 (D.S.C. Sept. 3,

26

2004) (“[T]he Court does not find that difficulty voting equates with a ‘denial or

27

abridgment’ of the right to vote.”).

28

With these principles in mind, the Court first will apply step one of the two-part

- 55 -

1

vote denial framework to each of the challenged voting practices to determine whether

2

either disparately burdens minority voters. The Court then will discuss step two and the

3

Senate Factors.

4
5

A. Step One (Disparate Impact)
1. H.B. 2023

6

H.B. 2023 is facially neutral. It applies to all Arizonans regardless of race or

7

color. Plaintiffs nonetheless allege that H.B. 2023 disparately burdens Hispanic, Native

8

American, and African American voters as compared to non-minority voters because

9

these groups disproportionately rely on others to collect and return their early ballots.

10

But there are no records of the numbers of people who, in past elections, have relied on

11

now-prohibited third parties to collect and return their early mail ballots, and of this

12

unknown number Plaintiffs have provided no quantitative or statistical evidence

13

comparing the proportion that is minority versus non-minority.

14

This evidentiary hole presents a practical problem. Disparate impact analysis is a

15

comparative exercise. To determine whether a practice disparately impacts minorities,

16

the Court generally must know approximately: (1) how many people will be affected by

17

the practice, and (2) their racial composition. Without this information, it becomes

18

difficult to compare the law’s impact on different demographic populations and to

19

determine whether the disparities, if any, are meaningful. That is, it might be true that

20

minorities qualitatively have used ballot collection services more often than non-

21

minorities, but the discrepancy might be slight enough that it does not meaningfully deny

22

minorities an equal opportunity to participate in the political process and elect their

23

preferred representatives. See Frank, 768 F.3d at 754.

24

Indeed, the Court is aware of no vote denial case in which a § 2 violation has been

25

found without quantitative evidence measuring the alleged disparate impact of a

26

challenged law on minority voters. Rather, the standards developed for analyzing § 2

27

vote denial cases suggest that proof of a relevant statistical disparity might be necessary

28

at step one, even though it is not alone sufficient to prove a § 2 violation because of the

- 56 -

1

causation requirement at step two.15 See Gonzales, 677 F.3d at 405; Veasey, 830 F.3d at

2

244 (noting that “courts regularly utilize statistical analyses to discern whether a law has

3

a discriminatory impact”).

4

necessity of statistical evidence in disparate impact cases.” Budnick v. Town of Carefree,

5

518 F.3d 1109, 1118 (9th Cir. 2008) (Fair Housing Act); Pottenger v. Potlatch Corp., 329

6

F.3d 740, 749 (9th Cir. 2003) (Age Discrimination in Employment Act); Cooper v. S.

7

Co., 390 F.3d 695, 716 (11th Cir. 2004), overruled on other grounds by Ash v. Tyson

8

Foods, Inc., 546 U.S. 454 (2006) (Title VII); Rollins v. Alabama Cmty. Coll. Sys., No.

9

2:09-CV-636-WHA, 2010 WL 4269133, at *9 (M.D. Ala. Oct. 25, 2010) (Equal Pay

10

Act); Davis v. City of Panama City, Fla., 510 F. Supp. 2d 671, 689 (N.D. Fla. 2007)

11

(Title VII and 42 U.S.C. § 1983).

Further, in other contexts courts have “recognized the

12

The Court is not suggesting that quantitative evidence of a challenged voting

13

practice’s actual effect is needed in vote denial cases. As Chief Judge Thomas noted in

14

his dissent during the preliminary appellate phase of this case, “quantitative evidence of

15

the effect of a rule on voting behavior is only available after an election has occurred, at

16

which point the remedial purpose of the Voting Rights Act is no longer served.”

17

Feldman, 840 F.3d at 1092 n.5 (Thomas, C.J. dissenting). But quantitative evidence of

18

the number of voters who used ballot collection before H.B. 2023’s enactment, together

19

with similar evidence of those voters’ demographics, would permit the Court to

20

reasonably infer how many voters would be affected by H.B. 2023’s limitations in future

21

elections, and whether those voters disproportionately would be minorities.16 As one

22
23
24
25
26
27
28

15

In vote dilution cases the Senate Factors “are sometimes used as a non-statistical
proxy . . . to link disparate impacts to current or historical conditions of discrimination.”
Ohio Democratic Party, 834 F.3d at 637 n.11. But to use the Gingles factors to prove the
existence of a disparity essentially would collapse the step one and step two inquiries.
That is, a plaintiff could simply assume that the challenged law causes a meaningful
disparity between minorities and non-minorities because of social and historical
discrimination in the state. This perhaps is another illustration of how the Gingles votedilution framework is an imperfect fit for vote denial claims.
16
Notably, the trial in this matter occurred after H.B. 2023 had been in effect for
two major elections—the 2016 presidential preference election and the 2016 general
election—yet Plaintiffs still were unable to produce data on the law’s impact.

- 57 -

1
2
3
4
5
6
7

commentator has argued:
It can be difficult to document the racial composition of those
who use a voting opportunity . . . , given that election and
other public records often do not include racial or ethnic data.
There is no getting around this problem. But given that § 2
forbids the denial or abridgement of the vote on account of
race, it is reasonable that plaintiffs be required to make a
threshold showing they are disproportionately burdened by
the challenged practice, in the sense that it eliminates an
opportunity they are more likely to use or imposes a
requirement they are less likely to satisfy.

8

Daniel P. Tokaji, Applying Section 2 to the New Vote Denial, 50 Harv. C.R.-C.L. L. Rev.

9

439, 476 (2015).

10

The Court is mindful, however, that no court has explicitly required quantitative

11

evidence to prove a vote denial claim, and a majority of the en banc Ninth Circuit panel

12

reviewing the preliminary phase of this case appears to have rejected such a rule. The

13

Court therefore does not find against Plaintiffs on this basis. Rather, the Court finds that

14

Plaintiffs’ circumstantial, qualitative evidence is insufficient to establish a cognizable

15

disparity under § 2.

16

To overcome the lack of quantification, Plaintiffs attempt to prove the existence of

17

a meaningful disparity through two general categories of circumstantial evidence. First,

18

lawmakers, elections officials, and community advocates testified that ballot collection

19

tends to be used more by communities that lack easy access to secure, outgoing mail

20

services; the elderly, homebound, and disabled; the poor; those who lack reliable

21

transportation; those who work multiple jobs or lack childcare; and less educated voters

22

who are unfamiliar with or more intimidated by the voting process. In turn, data shows

23

that these socioeconomic circumstances are disproportionately reflected in minority

24

communities. (See Ex. 97 at 57; Tr. 59-60, 416-26, 432-39, 629-35, 895-900.) It stands

25

to reason, then, that prior to H.B. 2023’s enactment minorities qualitatively were more

26

likely than non-minorities to give their early ballots to third parties.

27

For example, relative to non-minorities, Hispanics and African Americans are

28

nearly two times more likely to live in poverty, and the poverty rate for Native Americans

- 58 -

1

is over three times higher. (Ex. 93 at 15.) Wages and unemployment rates for Hispanics,

2

African Americans, and Native Americans consistently have exceeded non-minority

3

unemployment rates for the period of 2010 to 2015. (Ex. 91 at 40; Ex. 93 at 15.)

4

According to the 2015 American Community Survey 1-year estimates, unemployment

5

rates were 10.5 percent for African Americans, 7.7 percent for Hispanics, 16.8 percent for

6

Native Americans, and only 5.6 percent for non-minorities. In Arizona, 68.9 percent of

7

non-minorities own a home, whereas only 32.3 percent of African Americans, 49 percent

8

of Hispanics, and 56.1 percent of Native Americans do so. (Ex. 93 at 15, 17; Ex. 98 at

9

33.)

10

Non-minorities remain more likely than Hispanics, Native Americans, and African

11

Americans to graduate from high school, and are nearly three times more likely to have a

12

bachelor’s degree than Hispanics and Native Americans.

13

survey, over 22.4 percent of Hispanics and 11.2 percent of Native Americans rated

14

themselves as speaking English less than “very well,” as compared to only 1.2 percent of

15

non-minorities. (Ex. 93 at 16.) Due to their lower levels of literacy and education,

16

minority voters are more likely to be unaware of certain technical rules, such as the

17

requirement that early ballots be received by the county recorder, rather than merely

18

postmarked, by 7:00 p.m. on Election Day. (Ex. 91 at 38.)

19

Additionally, in a recent

As of 2015, Hispanics, Native Americans, and African Americans fared worse

20

than non-minorities on a number of key health indicators. (Ex. 93 at 18.)

21

Americans in particular have much higher rates of disability than non-minorities, and

22

Arizona counties with large Native American populations have much higher rates of

23

residents with ambulatory disabilities. For example, “17 percent of Native Americans are

24

disabled in Apache County, 22 percent in Navajo County, and 30 percent in Coconino

25

County.” Further, “11 percent [of individuals] have ambulatory difficulties in Apache

26

County, 13 percent in Navajo County, and 12 percent in Coconino County, all of which

27

contain significant Native American populations and reservations.” (Ex. 97 at 60.)

28

Native

Hispanics, Native Americans, and African Americans also are significantly less

- 59 -

1

likely than non-minorities to own a vehicle, more likely to rely upon public

2

transportation, more likely to have inflexible work schedules, and more likely to rely on

3

income from hourly wage jobs. (Ex. 93 at 12-18; Ex. 97 at 51-52; Tameron Dep. 155:5-

4

20; Pstross Dep. 34:11-22.)

5

nonexistent in some minority communities. (Ex. 97 at 57; Ex. 98 at 18; Tr. 506.)

Ready access to reliable and secure mail service is

6

These disparities exist in both urban and rural areas. For example, Representative

7

Charlene Fernandez described a lack of home mail service in rural San Luis, a city that is

8

98 percent Hispanic. Almost 13,000 residents rely on a post office located across a major

9

highway. With no mass transit, a median income of $22,000, and many people not

10

owning cars, receiving and sending mail in San Luis can be more difficult than in other

11

communities. (Tr. 40-46.) A surprising number of voters in the Hispanic community

12

also distrust returning their voted ballot via mail, particularly in low-income communities

13

where mail theft is common. Although a lack of outgoing mail presents a problem for

14

rural minority voters, unsecure mailboxes are an impediment for urban minorities who

15

distrust the mail service and prefer instead to give their ballots to a volunteer. (Tr. 98,

16

238-39, 896-97, 1170; Healy Dep. 97:18-24; Scharff Dep. 92:5-17.)

17

These problems are particularly acute in Arizona’s Native American communities,

18

in which vehicle ownership is significantly lower than non-minority Arizonans. (Ex. 91

19

at 42.)

20

reservations lack access to a vehicle. (Ex. 98 at 16.) Moreover, according to Dr. Rodden,

21

“the extent to which rural Native Americans lack mail service is quite striking.” “[T]he

22

majority of Native Americans in non-metropolitan Arizona do not have residential mail

23

service.” “Only 18 percent of Native American registered voters have home mail

24

delivery. . . . The rate at which registered voters have home mail service is over 350

25

percent higher for non-Hispanic whites than for Native Americans.” As such, most

26

Native American registered voters must travel to a town to retrieve their mail, “[y]et rates

27

of vehicle access are quite low.” (Ex. 97 at 57.) On the Navajo Reservation, most people

28

live in remote communities, many communities have little to no vehicle access, and there

Between one quarter and one half of all households on Native American

- 60 -

1

is no home incoming or outgoing mail, only post office boxes, sometimes shared by

2

multiple families. (Tr. 172-75, 297-98.)

3

There is no home delivery in the Tohono O’odham Nation, where there are 1,900

4

post office boxes and some cluster mail boxes. The postmaster for the Tohono O’odham

5

Nation anecdotally related to Representative Fernandez that she observes residents come

6

to the post office every two or three weeks to get their mail. Due to the lack of

7

transportation, the condition of the roads, and health issues, some go to post office only

8

once per month. (Tr. 52-58, 315-17.)

9

Thus, “for many Native Americans living in rural locations, especially on

10

reservations, voting is an activity that requires the active assistance of friends and

11

neighbors.” (Ex. 97 at 60.) LeNora Fulton—a member of the Navajo Nation, former

12

representative of the Fort Defiance Chapter of the Navajo National Council, member of

13

the Navajo Election Board of Supervisors, and the Apache County Recorder from 2004

14

through 2016, where her responsibilities included overseeing voter registration, early

15

voting and voter outreach—explained that people in the Navajo Nation trust non-family

16

members to deliver their early ballots because “[i]t’s part of the culture. . . . [T]here is a

17

clan system. They may not be related by blood, but they are related by clan. Everyone

18

on the Navajo Nation is related on way or another through the clan system.” Ballot

19

collection and delivery by those with the means to travel “was the standard practice with

20

the Apache County . . . but also with the Nation[.]” “We have many people that would

21

come into our office in St. Johns that help individuals that not are not able to get a ballot,

22

you know, to the office.

23

practice . . . It was a norm for us.” According to Fulton, limiting who may collection and

24

deliver early ballots “would be a huge devastation . . . . The laws are supposed to be

25

helpful to people, but in this instance, it’s harmful.” (Tr. 283-85, 300, 322-324.)

They would bring it in.

And so it was just a standard

26

The second category of circumstantial evidence concerns those who tend to offer

27

ballot collection services. Within the last decade, ballot collection has become a larger

28

part of the Democratic Party’s GOTV strategy. The Democratic Party and community

- 61 -

1

advocacy organizations have focused their ballot collection efforts on low-efficacy

2

voters, who trend disproportionately minority. In turn, minorities in Arizona tend to vote

3

for Democratic candidates. (Ex. 93 at 4-6, 11; Tr. 92, 283, 309, 416-26, 632-33, 659,

4

902-03, 1143, 1191-96, 1200, 1407, 1770-71, 1843-44; Healy Dep. 28:15-29:13.)

5

Individuals who have collected ballots in past elections observed that minority voters,

6

especially Hispanics, were more interested in utilizing their services. Indeed, Helen

7

Purcell, who served as the Maricopa County Recorder for 28 years from 1988 to 2016,

8

observed that ballot collection was disproportionately used by Hispanic voters. (Tr. 417-

9

19, 635, 642, 866, 895-900, 931-32, 1039-40, 1071, 1170.)

10

In contrast, the Republican Party has not significantly engaged in ballot collection

11

as a GOTV strategy. The base of the Republican Party in Arizona trends non-minority.

12

On average, non-minorities in Arizona vote 59 percent for Republican candidates, as

13

compared with 35 percent of Hispanic voters. Individuals who have collected ballots in

14

past elections have observed that voters in predominately non-minority areas were not as

15

interested in ballot collection services. (Tr. 430-31, 898, 1170, 1192, 1408; Ex. 91 at 31.)

16

Based on this evidence, the Court finds that prior to H.B. 2023’s enactment

17

minorities qualitatively were more likely than non-minorities to return their early ballots

18

with the assistance of third parties. The Court, however, cannot speak in more specific or

19

precise terms than “more” or “less.”

20

disparities between minorities and non-minorities in Arizona, these disparities are an

21

imprecise proxy for disparities in ballot collection use. Plaintiffs do not argue that all or

22

even most socioeconomically disadvantaged voters use ballot collection services, nor

23

does the evidence support such a finding. Rather, the anecdotal estimates from individual

24

ballot collectors indicate that a relatively small number of voters have used ballot

25

collection services in past elections.

26

socioeconomically disadvantaged voters, most do not use ballot collection services to

27

vote. Considering the vast majority of Arizonans, minority and non-minority alike, vote

28

without the assistance of third-parties who would not fall within H.B. 2023’s exceptions,

Although there are significant socioeconomic

It reasonably follows, then, that even among

- 62 -

1

it is unlikely that H.B. 2023’s limitations on who may collect an early ballot cause a

2

meaningful inequality in the electoral opportunities of minorities as compared to non-

3

minorities.

4

Moreover, H.B. 2023 does not impose burdens beyond those traditionally

5

associated with voting. Although, for some voters, ballot collection is a preferred and

6

more convenient method of voting, H.B. 2023 does not deny minority voters meaningful

7

access to the political process simply because the law makes it slightly more difficult or

8

inconvenient for a small, yet unquantified subset of voters to return their early ballots. In

9

fact, no individual voter testified that H.B. 2023’s limitations on who may collect an

10

early ballot would make it significantly more difficult to vote. The Court therefore finds

11

that Plaintiffs have not carried their burden at step one of the vote denial framework.

12

2. OOP Voting

13

Unlike their H.B. 2023 challenge, Plaintiffs provided quantitative and statistical

14

evidence of disparities in OOP voting through the expert testimony of Dr. Rodden.

15

Because Arizona does not track the racial demographics of its voters, Dr. Rodden used an

16

open-source software algorithm that he found online to predict each voter’s race. (Tr.

17

378-414, 520-24.) Specifically, Dr. Rodden placed the names of individuals who cast

18

ballots in particular elections into Census blocks and tracts, for which racial data is

19

available from the Census, and then combined that information with surname data to

20

estimate the race of each voter. This approach has become common in academic studies,

21

as well as VRA litigation. Dr. Rodden added further precision to his estimates by

22

conditioning them on not just surnames and neighborhood race statistics, but also on

23

additional information found in the voter file, such as the individual’s age and party.17

24
25
26
27
28

17

Dr. Thornton criticized Dr. Rodden’s analysis of racial disparities in OOP
voting among the smallest minority groups in Arizona’s smaller counties, but when Dr.
Rodden conducted an analysis addressing Dr. Thornton’s criticisms he reached the same
results. Dr. Thornton also was critical of Dr. Rodden’s analysis because the application
of the algorithm to Arizona voters includes unidentifiable measurement error. But
because there is no concrete racial data for individual voters, Dr. Rodden has no means to
compare his estimates. Indeed, if such concrete data existed, there would have been no
need for Dr. Rodden’s estimates. Moreover, as Dr. Rodden explains in his Second Expert

- 63 -

1

(Ex. 97 at 10-15.)

2

Dr. Rodden’s analysis is credible and shows that minorities are over-represented

3

among the small number of voters casting OOP ballots. For example, in Maricopa

4

County—which accounts for 61 percent of Arizona’s population—non-minority voters

5

accounted for only 56 percent of OOP ballots cast during the 2012 general election,

6

despite casting 70 percent of all in-person voters. In contrast, African American and

7

Hispanic voters made up 10 percent and 15 percent of in-person voters, but accounted for

8

13 percent and 26 percent of OOP ballots, respectively.

9

accounted for 1.1 percent of in-person voters, and 1.3 percent of OOP ballots. Dr.

10

Rodden observed similar results for Pima County. (Ex. 91 at 52; Ex. 95 at 31-33, 37, 43.)

11

Similarly, minority voters cast a disproportionate share of OOP ballots during the

12

2016 general election. In Maricopa County, estimated rates of OOP voting were twice as

13

high for Hispanics, 86 percent higher for African Americans, and 73 percent higher for

14

Native Americans than for their non-minority counterparts. In Pima County, rates of

15

OOP voting were 150 percent higher for Hispanics, 80 percent higher for African

16

Americans, and 74 percent higher for Native Americans than for non-minorities.

17

Moreover, in Pima County the overall rate of OOP voting was higher, and the racial

18

disparities larger, in 2016 than in 2014. Among all counties that reported OOP ballots in

19

the 2016 general election, a little over 1 in every 100 Hispanic voters, 1 in every 100

20

African-American voters, and 1 in every 100 Native American voters cast an OOP ballot.

21

For non-minority voters, the figure was around 1 in every 200 voters. Racial disparities

22

in OOP voting were found in all counties except La Paz County, which has a small

23

minority population. (Ex. 97 at 3, 19-21, 28-34.)

Native American voters

24

Although Dr. Rodden’s race estimation is credible, his analysis paints an

25

incomplete picture of the practical impact of OOP voting because the majority of

26

Arizonans successfully vote by mail and therefore are unaffected by precinct

27
28

Report, his methods lead to more conservative estimates of disparities, a fact not
challenged by Dr. Thornton. (Ex. 98 at 3, 8-11.)

- 64 -

1

requirements. For example, in the 2012 general election Arizona voters cast 10,979 OOP

2

ballots out of the 2,323,579 ballots cast statewide, accounting for 0.47 percent of all votes

3

cast. In Maricopa County, 1,390,836 total ballots were cast, of which 7,529 (or 0.54

4

percent) were rejected for being cast out of the voter’s assigned precinct. OOP ballots

5

cast by non-minority voters therefore accounted for only 0.3 percent of all votes cast in

6

Maricopa County during the 2012 election, whereas OOP ballots cast by Hispanic and

7

African American voters accounted only for approximately 0.14 percent and 0.07

8

percent, respectively. These figures dropped substantially in the 2016 general election,

9

during which only 3,970 Arizonans voted OOP out of the 2,661,497 ballots cast

10

statewide, representing only 0.15 percent of all votes cast.

11

1,608,875 total ballots were cast, of which only 2,197 (or 0.14 percent) were cast OOP.

12

(Tr. 1927-32; Exs. 578-79, 581.)

In Maricopa County,

13

Considering OOP ballots represent such a small and ever-decreasing fraction of

14

the overall votes cast in any given election, OOP ballot rejection has no meaningfully

15

disparate impact on the opportunities of minority voters to elect their preferred

16

representatives. To be clear, the Court is not suggesting that the votes of individuals who

17

show up at the wrong precinct are unimportant. But, as a practical matter, the disparity

18

between the proportion of minorities who vote at the wrong precinct and the proportion

19

of non-minorities who vote at the wrong precinct does not result in minorities having

20

unequal access to the political process. See Fed. Judicial Ctr., Reference Manual on

21

Scientific Evidence 252 (3d ed. 2011) (discussing difference between statistical

22

significance and practical significance). No state has exactly equal rates at every stage of

23

its voting system, and in the end the vast majority of all votes in Arizona—cast by

24

minority and non-minority voters alike—are counted. See Frank, 768 F.3d at 754.

25

Moreover, Arizona’s policy to not count OOP ballots is not the cause of the

26

disparities in OOP voting.

27

concentrated in relatively dense precincts that are disproportionately populated with

28

renters and those who move frequently. These groups, in turn, are disproportionately

Dr. Rodden’s analysis confirms that OOP voting is

- 65 -

1

composed of minorities. (Ex. 97 at 16-18.) Because minority voters in Arizona have

2

disproportionately higher rates of residential mobility, they are more likely to need to

3

renew their voter registration and reeducate themselves about their new voting locations.

4

(Ex. 91 at 39; Ex. 93 at 17; Ex. 95 at 4, 7-12; Ex. 98 at 33.)

5

Polling place locations present additional challenges for Native American voters.

6

For example, Navajo voters in Northern Apache County lack standard addresses, and

7

their precinct assignments for state and county elections are based upon guesswork,

8

leading to confusion about the voter’s correct polling place. Additionally, boundaries for

9

purposes of tribal elections and Apache County precincts are not the same. As a result, a

10

voter’s polling place for tribal elections often differs from the voter’s polling place for

11

state and county elections. Inadequate transportation access also can make travelling to

12

an assigned polling place difficult. (Ex. 97 at 51-54; Tr. 299-301.)

13

Plaintiffs, however, do not challenge the manner in which Arizona counties

14

allocate and assign polling places or Arizona’s requirement that voters re-register to vote

15

when they move. Plaintiffs also offered no evidence of a systemic or pervasive history of

16

minority voters being given misinformation regarding the locations of their assigned

17

precincts, while non-minority voters were given correct information. Nor have they

18

shown that precincts tend to be located in areas where it would be more difficult for

19

minority voters to find them, as compared to non-minority voters. To the contrary, there

20

are many ways for voters in Arizona to locate their assigned precincts, and state, county,

21

and local elections officials engage in substantial informational campaigns and voter

22

outreach. Plaintiffs, instead, have challenged what Arizona does with OOP ballots after

23

they have been cast, which does not cause the observed disparities in OOP voting.

24

In sum, Plaintiffs have not carried their burden at step one of the vote denial

25

framework for two independent reasons. First, they have not shown that Arizona’s policy

26

to not count OOP ballots causes minorities to show up to vote at the wrong precinct at

27

rates higher than their non-minority counterparts.

28

account for such a small fraction of votes cast statewide, Plaintiffs have not shown that

- 66 -

Second, given that OOP ballots

1

the racial disparities in OOP voting are practically significant enough to work a

2

meaningful inequality in the opportunities of minority voters as compared to non-

3

minority voters to participate in the political process and elect their preferred

4

representatives.

5

B. Step Two (Senate Factors)

6

Step two of the results test is informed by the Senate Factors and asks whether the

7

disparate burdens imposed by the challenged voting practices are in part caused by or

8

linked to social and historical conditions within the state that have or currently produce

9

discrimination against the affected minorities. The Court does not need to reach this step

10

because Plaintiffs have not shown at step one that the challenged voting practices impose

11

meaningfully disparate burdens on minority voters as compared to non-minority voters.

12

Cf. Ohio Democratic Party, 834 F.3d at 638 (“If this first element is met, the second step

13

comes into play, triggering consideration of the ‘totality of the circumstances,’ potentially

14

informed by the ‘Senate Factors’ discussed in ‘Gingles.’”). Nonetheless, to ensure that

15

the record is fully developed, the Court will address below the evidence pertinent to the

16

Senate Factors. The Court will not discuss factors three and four, however, because they

17

are not germane to the challenged voting practices and there is insufficient evidence to

18

warrant discussion.

19

1. Relevant History of Official Discrimination

20

Arizona has a history of discrimination against Native Americans, Hispanics, and

21

African Americans. Such discrimination began as early as 1912, when Arizona became a

22

state, and continued into the modern era. In 1975, Arizona’s history of discrimination

23

resulted in it becoming one of only nine states to be brought wholly under § 5 of the VRA

24

as a “covered jurisdiction.” In addition to being covered under § 5, it was one of only

25

three states to be covered under § 4(f)(4) of the Act for Spanish Heritage. (Ex. 89 at 5-

26

24; Ex. 91 at 2, 24-30; Ex. 521 at 43-45; Doc. 361 ¶ 42.)

27

When Arizona became a state in 1912, Native Americans were excluded from

28

voting. Even after Congress acknowledged that Native Americans were citizens in 1924,

- 67 -

1

thereby affording them the right to vote, Arizona’s Constitution continued to deny Native

2

Americans that right. (Ex. 89 at 17; Ex. 521 at 43.) It was not until 1948—24 years after

3

federal law allowed Native Americans to vote—that the Arizona Supreme Court found

4

the State’s disenfranchisement of Native Americans was unconstitutional and finally

5

granted Native Americans the right to vote. (Doc. 361 ¶ 17; Ex. 89 at 17; Ex. 521 at 45.);

6

Harrison v. Laveen, 196 P.2d 456, 463 (Ariz. 1948).

7

Despite this ruling, Native Americans, as well as Hispanics and African

8

Americans, continued to face barriers to participation in the franchise. For example, in

9

1912 Arizona enacted an English literacy test for voting.

The test was enacted

10

specifically to limit “the ignorant Mexican vote,” but it also had the effect of reducing the

11

ability of African Americans and Native Americans to register and vote, as registrars

12

applied the test to these communities as well. Well into the 1960s, white Arizonans

13

challenged minority voters at the polls by asking them to read and explain literacy cards.

14

(Doc. 361 ¶ 14; Ex. 89 at 14-17; Ex. 521 at 44-45.)

15

In 1970, Congress amended the VRA to enact a nationwide ban on literacy tests

16

after finding that they were used to discriminate against voters on account of their race or

17

ethnicity. In reaching that finding, Congress cited evidence that showed application of

18

the literacy test had significantly lowered the participation rates of minorities.

19

specifically found that in Arizona “only two counties out of eight with Spanish surname

20

populations in excess of 15% showed a voter registration equal to the state-wide

21

average.” It also noted that Arizona had a serious deficiency in Native American voter

22

registrations. Rather than comply with the VRA and repeal its literacy test, Arizona

23

challenged the ban, arguing that it could not be enforced to the extent that it was

24

inconsistent with the State’s literacy requirement. Even after the Supreme Court upheld

25

Congress’s ban, Arizona waited an additional two years to formally repeal its literacy

26

test. (Ex. 89 at 14-18; Ex. 91 at 24); see Or. v. Mitchell, 400 U.S. 112, 118 (1970).

It

27

The effects of Arizona’s literacy test were compounded by the State’s history of

28

discrimination in the education of its Hispanic, Native American, and African American

- 68 -

1

citizens. (Doc. 361 ¶¶ 2-4; Ex. 89 at 9-10; Ex. 91 at 5.) From 1912 until the Supreme

2

Court’s decision in Brown v. Board of Education, 347 U.S. 483 (1954), segregated

3

education was widespread throughout Arizona and sanctioned by both the courts and the

4

state legislature. (Ex. 521 at 35-39; Ex. 89 at 9-12); see also Dameron v. Bayless, 126 P.

5

273 (Ariz. 1912); Gonzales v. Sheely, 96 F. Supp. 1004, 1008-09 (D. Ariz. 1951)

6

(enjoining segregation of Mexican school children in Maricopa County). In fact, the

7

Tucson Public Schools only recently reached a consent decree with the DOJ over its

8

desegregation plan in 2013. (Ex. 91 at 27.) The practice of segregation also extended

9

beyond schools; it was common place to have segregated public spaces such as

10

restaurants, swimming pools, and theaters. (Ex. 89 at 15; Ex. 521 at 34). Even where

11

schools were not segregated, Arizona enacted restrictions on bilingual education. As

12

recently as 2000, Arizona banned bilingual education with the passage of Proposition

13

203. (Ex. 89 at 20; Ex. 91 at 47.)

14

Arizona has a record of failing to provide adequate funding to teach its non-

15

English speaking students. This under-funding has taken place despite multiple court

16

orders instructing Arizona to develop an adequate funding formula for its programs,

17

including a 2005 order in which Arizona was held in contempt of court for refusing to

18

provide adequate funding for its educational programs. (Ex. 91 at 46-47); Flores v.

19

Arizona, 405 F. Supp. 2d 1112 (D. Ariz. 2005), vacated, 204 Fed. App’x 580 (9th Cir.

20

2006). “According to the Education Law Center’s latest National Report Card that

21

provided data for 2013, Arizona ranked 47th among the states in per-student funding for

22

elementary and secondary education.” (Ex. 91 at 47.)

23

Along with the State’s hostility to bilingual education, Maricopa County has

24

sometimes failed to send properly translated education materials to its Spanish speaking

25

residents, resulting in confusion and distrust from Hispanic voters. For example, in 2012,

26

Maricopa County misprinted the date of the election on over 2,000 Spanish language

27

information cards and bookmarks, some of which were distributed into the community.

28

(Ex. 89 at 22; Ex. 91 at 51; Healy Dep. 114:1-22.)

- 69 -

1

With that said, discrimination against minorities in Arizona has not been linear.

2

(Ex. 521 at 4.) For example, Arizona was subject to § 5 preclearance requirements until

3

2013. In Shelby, however, the Supreme Court found the formula used to determine

4

which states were subject to preclearance requirements unconstitutional because it was

5

“based on 40–year–old facts having no logical relation to the present day.” 570 U.S. at

6

553-54. Moreover, during the time that Arizona was under preclearance requirements

7

(1975-2013), the DOJ did not issue any objections to any of its statewide procedures for

8

registration or voting.

9

From 1982 to 2002, the DOJ objected to four of Arizona’s statewide redistricting

10

plans. Arizona acted to avoid the politics of racially discriminatory redistricting when, in

11

2000, the Arizona Independent Redistricting Commission (“AIRC”) was formed pursuant

12

to a voter initiative (Proposition 106). The AIRC is composed of two Republicans, two

13

Democrats, and an Independent, and it is tasked with redrawing of legislative and

14

congressional district lines following each decennial Census. According to its enacting

15

constitutional provisions, the AIRC considers the following six criteria when

16

redistricting: (a) equal population; (b) compactness and contiguousness; (c) compliance

17

with the Constitution and the VRA; (d) respect for communities of interest; (e)

18

incorporation of visible geographic features, including city, town and county boundaries,

19

as well as undivided census tracts; and (f) creation of competitive districts where there is

20

no significant detriment to other goals. (Doc. 361 ¶ 44.) The most recent AIRC set a

21

goal to pass preclearance with its first submittal to DOJ. The AIRC did this by ensuring

22

the competitiveness of legislative and congressional districts and ensuring that minorities

23

have the opportunity to elect candidates of their choice. See Harris v. Ariz. Indep.

24

Redistricting Comm’n, -- U.S. --, 136 S. Ct. 1301, 1308 (2016). The Commission

25

succeeded, and the DOJ approved Arizona’s new maps on April 9, 2012 without

26

objection.

27

In sum, “[d]iscriminatory action has been more pronounced in some periods of

28

state history than others . . . [and] each party (not just one party) has led the charge in

- 70 -

1

discriminating against minorities over the years.”

2

objectives are the motivating factor in decisions to take actions detrimental to the voting

3

rights of minorities. “[M]uch of the discrimination that has been evidenced may well

4

have in fact been the unintended consequence of a political culture that simply ignores

5

the needs of minorities.” (Ex. 90 at 8.) Arizona’s recent history is a mixed bag of

6

advancements and discriminatory actions.

7

Sometimes, however, partisan

2. Racially Polarized Voting

8

Arizona has a history of racially polarized voting, which continues today. (Ex. 91

9

at 30-33, 44-45). In the most recent redistricting cycle, experts for the AIRC found that

10

at least one congressional district and five legislative districts clearly exhibited racially

11

polarized voting. (Ex. 91 at 29-33.) Exit polls for the 2016 general election demonstrate

12

that voting between non-minorities and Hispanics continues to be polarized along racial

13

lines. (Ex. 91 at 29-33, 44-45; Ex. 92 at 12, 14; Ex. 94 at 4); see also Gonzalez, 677 F.3d

14

at 407.

15

3. Socioeconomic Effects of Discrimination

16

Racial disparities between minorities and non-minorities in socioeconomic

17

standing, income, employment, education, health, housing, transportation, criminal

18

justice, and electoral representation have persisted in Arizona. (Ex. 89 at 7-8, 12, 23; Ex.

19

91 at 39-43; Ex. 93 at 12-18, 21, 24; Ex. 95 at 4, 9-11; Ex. 97 at 46-52, 56-58; Ex. 98 at

20

16, 18, 33; Tameron Dep. 155:5-20; Pstross Dep. 34:11-22; Tr. 506.)

21

disparities in transportation, housing, and education are most pertinent to the specific

22

burdens imposed by the challenged laws.

23

Of these,

4. Racial Appeals in Political Campaigns

24

Arizona’s racially polarized voting has resulted in racial appeals in campaigns.

25

For example, when Raul Castro ran for governor in the 1970s, his opponents urged

26

support for the white candidate because “he looked like a governor.” In that same

27

election, a newspaper published a picture of Fidel Castro with a headline that read

28

“Running for governor of Arizona.”

(Ex. 89 at 19.)

- 71 -

In a 2010 bid for State

1

Superintendent of Public Education, John Huppenthal “ran an advertisement in which the

2

announcer said that Huppenthal was ‘one of us.’ The announcer noted that Huppenthal

3

voted against bilingual education and ‘will stop La Raza.’” Similarly, when running for

4

governor in 2014, Maricopa County Attorney Andrew Thomas ran an ad describing

5

himself as “the only candidate who has stopped illegal immigration” while

6

“simultaneously show[ing] a Mexican flag with a red strikeout line through it

7

superimposed over the outline of Arizona.” (Ex. 91 at 44.)

8
9

Moreover, racial appeals have been made in the specific context of legislative
efforts to limit ballot collection.

During the legislative hearings on earlier bills to

10

criminalize ballot collection, Republican sponsors and proponents expressed beliefs that

11

ballot collection fraud regularly was occurring but struggled with the lack of direct

12

evidence substantiating those beliefs. In 2014, the perceived “evidence” arrived in the

13

form of a racially charged video created by Maricopa County Republican Chair A.J.

14

LaFaro (the “LaFaro Video”) and posted on a blog. (Ex. 121.) The LaFaro Video

15

showed surveillance footage of a man of apparent Hispanic heritage appearing to deliver

16

early ballots. It also contained a narration of “Innuendos of illegality . . . [and] racially

17

tinged and inaccurate commentary by . . . LaFaro.” (Ex. 91 at 18 n.40; Ex. 524 at 23-

18

24.) LaFaro’s commentary included statements that the man was acting to stuff the ballot

19

box; that LaFaro did not know if the person was an illegal alien, a dreamer, or citizen, but

20

knew that he was a thug; and that LaFaro did not follow him out to the parking lot to take

21

down his tag number because he feared for his life. The LaFaro Video goes on to tell

22

about ballot parties where people gather en mass and give their un-voted ballots to

23

operatives of organizations so they can not only collect them, but also vote them illegally.

24

(Ex. 91 at 18; Ex. 121.)

25

The LaFaro Video did not show any obviously illegal activity and there is no

26

evidence that the allegations in the narration were true. Nonetheless, it “became quite

27

prominent in the debates over H.B. 2023.” (Tr. 1154.) The LaFaro video also was

28

posted on Facebook and YouTube, shown at Republican district meetings, and was

- 72 -

1

incorporated into a television advertisement—entitled “Do You Need Evidence

2

Terry?”—for Secretary Reagan when she ran for Secretary of State. (Ex. 91 at 18; Ex.

3

107.) In the ad, the LaFaro Video plays after a clip of then-Arizona Attorney General

4

Terry Goddard stating he would like to see evidence that there has been ballot collection

5

fraud. While the video is playing, Secretary Reagan’s narration indicates that the LaFaro

6

Video answers Goddard’s request for evidence of fraud. The LaFaro Video, however,

7

merely shows a man of apparent Hispanic heritage dropping off ballots and not obviously

8

violating any law.18 (Ex. 107.)

9
10

5. Minority Representation in Public Offices
Notwithstanding racially polarized voting and racial appeals, the disparity in the

11

number of minority elected officials in Arizona has declined.

12

recognized for improvements in the number of Hispanics and Native Americans

13

registering and voting, as well as in the overall representation of minority elected officials

14

in the State. (Ex. 521 at 27-28.) “Nonwhites make up 25 percent of Arizona’s elected

15

office holders, compared to 44 percent of the total population. This gives [Arizona] the

16

16th best representation ratio in the country.” (Ex. 524 at 44.)

Arizona has been

17

Nevertheless, Arizona has seen only one Hispanic and one African American

18

elected to statewide office, and Arizona has never elected a Native American to statewide

19

office. No Native American or African American has been elected to represent Arizona

20

in the United States House of Representatives. Further, no Hispanic, Native American,

21

or African American has ever served as a United States Senator representing Arizona or

22

as Arizona Attorney General. (Ex. 91 at 45; Ex. 93 at 19-20; Ex. 89 at 19, 22.)

23

6. Lack of Responsiveness to Minority Needs

24

Plaintiffs’ evidence on this factor, presented through the analysis and opinions of

25

Dr. Lichtman, is insufficient to establish a lack of responsiveness on the part of elected

26
27
28

18

Notably, LaFaro was not called as a witness in this case, Defendants do not rely
on the LaFaro Video as evidence of fraud, and, despite the implications of her campaign
advertisement, Secretary Reagan testified in deposition that “I have never accused anyone
collecting ballots as doing fraudulent activities[.]” (Reagan Dep. 91:2-3.)

- 73 -

1

officials to particularized needs of minority groups. Dr. Lichtman ignored various topics

2

that are relevant to whether elected officials have shown responsiveness, and he did not

3

conduct research on the issues in Arizona when considering this factor.

4

Notably, the CCEC engages in outreach to various communities, including the

5

Hispanic and Native American communities, to increase voter participation. The CCEC

6

develops an annual voter education plan in consultation with elections officials and

7

stakeholders, and the current Chairman of the CCEC is Steve Titla, an enrolled member

8

of the San Carlos Apache Tribe, who has been particularly vocal in supporting CCEC

9

outreach to Native Americans.

10

7. Justifications for Challenged Provisions

11

Precinct-based voting helps Arizona counties estimate the number of voters who

12

may be expected at any particular precinct, allows for better allocation of resources and

13

personnel, improves orderly administration of elections, and reduces wait times. The

14

precinct-based system also ensures that each voter receives a ballot reflecting only the

15

races for which that person is entitled to vote, thereby promoting voting for local

16

candidates and issues and making ballots less confusing. Arizona’s policy to not count

17

OOP ballots is one mechanism by which it strictly enforces this system to ensure that

18

precinct-based counties maximize the system’s benefits. This justification is not tenuous.

19

As for H.B. 2023, there is no direct evidence that the type of ballot collection

20

fraud the law is intended to prevent or deter has occurred. Although the justifications for

21

H.B. 2023 are weaker than the justifications for the State’s OOP ballot policy, Arizona

22

nonetheless has a constitutionally adequate justification for the law: to reduce

23

opportunities for early ballot loss or destruction.

24

8. Overall Assessment

25

In sum, of the germane Senate Factors, the Court finds that some are present in

26

Arizona and others are not. Plaintiffs have shown that past discrimination in Arizona has

27

had lingering effects on the socioeconomic status of racial minorities. But Plaintiffs’

28

causation theory is too tenuous to support their VRA claim because, taken to its logical

- 74 -

1

conclusion, virtually any aspect of a state’s election regime would be suspect as nearly all

2

costs of voting fall heavier on socioeconomically disadvantaged voters. Such a loose

3

approach to causation, which potentially would sweep away any aspect of a state’s

4

election regime in which there is not perfect racial parity, is inconsistent with the Ninth

5

Circuit’s repeated emphasis on the importance of a “causal connection between the

6

challenged voting practice and a prohibited discriminatory result.” Salt River Project,

7

109 F.3d at 595. For these reasons, the Court concludes that Plaintiffs have not carried

8

their burden at either step of the § 2 results test.

9

VII. FIFTEENTH AMENDMENT/§ 2 (INTENTIONAL DISCRIMINATION)

10

Lastly, Plaintiffs contend that H.B. 2023 violates § 2 and the Fifteenth

11

Amendment because it was enacted with the intent to suppress minority votes. The

12

Fifteenth Amendment provides that “[t]he right of citizens of the United States to vote

13

shall not be denied or abridged by the United States or by any State on account of race,

14

color, or previous condition of servitude,” and authorizes Congress to enforce this

15

mandate “by appropriate legislation.” Section 2 is such legislation. Although Congress

16

amended the VRA in 1982 to add the results test, § 2 continues to prohibit intentional

17

discrimination in a manner coextensive with the Fifteenth Amendment. Consequently,

18

the standards for both the statutory and the constitutional claim overlap.

19

The parties agree that the standard for finding unconstitutional, intentional racial

20

discrimination is governed by the Supreme Court’s decision in Village of Arlington

21

Heights v. Metropolitan Housing Development Corp., 429 U.S. 252 (1977). There, the

22

Supreme Court explained that “official action will not be held unconstitutional solely

23

because it results in a racially disproportionate impact.” Id. at 264-65. Rather, “[p]roof

24

of racially discriminatory intent or purpose is required to show a violation” of the

25

Constitution. Id. at 265.

26

Discriminatory purpose must be “a motivating factor in the decision,” but it need

27

not be the only factor. Id. at 265-66. “Determining whether invidious discriminatory

28

purpose was a motivating factor demands a sensitive inquiry into such circumstantial and

- 75 -

1

direct evidence of intent as may be available.” Id. at 266. “[A]n invidious discriminatory

2

purpose may often be inferred from the totality of the relevant facts, including the fact, if

3

it is true, that the law bears more heavily on one race than another.” Wash. v. Davis, 426

4

U.S. 229, 242 (1976). “But the ultimate question remains: did the legislature enact a law

5

‘because of,’ and not just ‘in spite of,’ its discriminatory effect.” N.C. St. Conf. of

6

NAACP v. McCrory, 831 F.3d 204, 200 (4th Cir. 2016) (quoting Pers. Adm’r of Mass. v.

7

Fenney, 442 U.S. 256, 279 (1979)).

8

To guide this inquiry, the Arlington Heights Court articulated a non-exhaustive list

9

of factors courts should consider. These so-called “Arlington Heights Factors” include:

10

(1) the historical background and sequence of events leading to enactment; (2)

11

substantive or procedural departures from the normal legislative process; (3) relevant

12

legislative history; and (4) whether the law has a disparate impact on a particular racial

13

group. Arlington Heights, 429 U.S. at 266-68. If “racial discrimination is shown to have

14

been a ‘substantial’ or ‘motivating’ factor behind enactment of the law, the burden shifts

15

to the law’s defenders to demonstrate that the law would have been enacted without this

16

factor.” Hunter v. Underwood, 471 U.S. 222, 228 (1985). This same framework applies

17

to § 2 claims based on allegations of discriminatory purpose. See Garza v. Cty. of L.A.,

18

918 F.2d 763, 766 (9th Cir. 1990).

19

Having considered these factors, the Court finds that H.B. 2023 was not enacted

20

with a racially discriminatory purpose.

21

proponents of limitations on ballot collection harbored partisan motives—perhaps

22

implicitly informed by racial biases about the propensity of GOTV volunteers in minority

23

communities to engage in nefarious activities—the legislature as a whole enacted H.B.

24

2023 in spite of opponents’ concerns about its potential effect on GOTV efforts in

25

minority communities, not because of that effect. Despite the lack of direct evidence

26

supporting their concerns, the majority of H.B. 2023’s proponents were sincere in their

27

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

28

was a necessary prophylactic measure to bring early mail ballot security in line with in-

Though some individual legislators and

- 76 -

1

person voting.

2

Beginning with the historical background, H.B. 2023 emerged in the context of

3

racially polarized voting, increased use of ballot collection as a Democratic GOTV

4

strategy in low-efficacy minority communities, and on the heels of several prior efforts to

5

restrict ballot collection, some of which were spearheaded by former Arizona State

6

Senator Don Shooter.19 Due to the high degree of racial polarization in his district,

7

Shooter was in part motivated by a desire to eliminate what had become an effective

8

Democratic GOTV strategy. (Tr. 1061-63, 1200, 1687-88, 2158-62; Ex. 89 at 24; Ex. 91

9

at 52-55; Ex. 92 at 2-10; Ex. 93 at 2; Shooter Dep. at 117:5-16.) Indeed, Shooter’s 2010

10

election was close: he won with 53 percent of the total vote, receiving 83 percent of the

11

non-minority vote but only 20 percent of the Hispanic vote. (Ex. 94 at 4.)

12

Shooter’s efforts to limit ballot collection were marked by unfounded and often

13

farfetched allegations of ballot collection fraud. (Tr. 1064, 2162, 2194, 2205; Ex. 3 at 7-

14

8; Ex. 10 at 3-9; Ex. 25 at 22-23; Ex. 91 at 19-20; Ex. 123.) Though his allegations were

15

demonstrably false, they—along with the racially-tinged LaFaro Video—spurred a larger

16

debate in the legislature about the security of early mail voting as compared to in-person

17

voting.20 (Tr. 1644, 1687, 2158-59, 2161-62; Ex. 10 at 49-53; Ex. 17 at 15-16; Ex. 23 at

18

83-84.)

19

Turning to the relevant legislative history, proponents of H.B. 2023 repeatedly

20

voiced concerns that mail-in ballots were less secure than in-person voting, and that

21

ballot collection created opportunities for fraud. Although no direct evidence of ballot

22

collection fraud was presented to the legislature or at trial, Shooter’s allegations and the

23

LaFaro Video were successful in convincing H.B. 2023’s proponents that ballot

24

collection presented opportunities for fraud that did not exist for in-person voting, and

25
26
27
28

19

Shooter most recently was a member of the Arizona House of Representatives
but served as a state senator during the relevant time period.
20
Although the video referenced by various proponents of ballot collection
limitations was not always identified as such, it is plain from their descriptions that they
were describing the LaFaro video.

- 77 -

1

these proponents appear to have been sincere in their beliefs that this was a potential

2

problem that needed to be addressed. (Ex. 17 at 11-13, 17-75, 83-84; Ex. 19 at 56-57;

3

Ex. 21 at 11; Ex. 23 at 36; Tr. 1450, 1805, 1822-23.) Notably, H.B. 2023 found support

4

among some minority officials and organizations.

5

supported by the Arizona Latino Republican Association for the Tucson Chapter, which

6

expressed concerns that elderly people in the Latino community were being taken

7

advantage of by ballot collectors. (Ex. 17 at 71-75.) Likewise, Michael Johnson, an

8

African American who had served on the Phoenix City Council, strongly favored H.B

9

2023 and expressed concern about stories of ballot collectors misrepresenting themselves

10

as election workers. (Ex. 17 at 45-50.) Further, although some Democratic lawmakers

11

accused their Republican counterparts of harboring partisan or racially discriminatory

12

motives, this view was not shared by all of H.B. 2023’s opponents. (Tr. 697.) For

13

example, Representative Fernandez testified that she has no reason to believe H.B. 2023

14

was enacted with the intent to suppress Hispanic voting. (Tr. 83.)

For example, the measure was

15

As for departures from the normal legislative process, Plaintiffs cite two prior

16

efforts to limit ballot collection as examples of procedural discrepancies. First, in 2011

17

Arizona enacted S.B. 1412, which required any person who delivered more than ten early

18

ballots to provide a copy of her photo identification to the receiving elections official. If

19

a ballot collector could not produce a copy of her photo identification, the elections

20

official was directed to record the information from whatever identification that the ballot

21

collector had available. Within 60 days of each election, the Secretary of State was to

22

compile a public statewide report listing the identities and personal information of all

23

ballot collectors. (Ex. 2 at 16-19; Ex. 91 at 6-7.)

24

When S.B. 1412 became law, Arizona still was subject to § 5 preclearance.

25

Accordingly, S.B. 1412 could not go into effect until the law had been precleared by the

26

DOJ or a federal court.

27

preclearance on April 26, 2011, and on June 27, 2011 the DOJ precleared all provisions

28

except for the provision regulating ballot collection. (Ex. 41; Ex. 91 at 6-7.) As to that

The Arizona Attorney General submitted the law for

- 78 -

1

provision, the DOJ stated that “the information sent [wa]s insufficient to enable us to

2

determine that the proposed changes have neither the purpose nor will have the effect of

3

denying or abridging the right to vote on account of race, color, or membership in a

4

language minority group.” The DOJ asked for more information and stated that “if no

5

response is received within sixty days of this request, the Attorney General may object to

6

the proposed changes.” (Ex. 41.)

7

information, the Attorney General chose to voluntarily withdraw the ballot collection

8

provision on July 28, 2011, rendering the law unenforceable. (Ex. 91 at 6-7; Ex. 42.)

9

“Of 773 preclearance submissions this was one of only 6 that were fully or partially

10

withdrawn in Arizona.”

11

Rather than respond to the DOJ’s request for more

thereafter. (Ex. 5.)

(Ex. 91 at 7.)

Arizona formally repealed the law shortly

12

Second, Republican legislators again tried to restrict ballot collection in 2013 with

13

the enactment of H.B. 2305, which banned partisan ballot collection and required other

14

ballot collectors to complete an affidavit stating that they had returned the ballot.

15

Violation of the law was a misdemeanor. H.B. 2305 was passed along nearly straight

16

party lines in the waning hours of the legislative session. (Ex. 7; Ex. 91 at 7-10.) Shortly

17

after its enactment, citizen groups organized a referendum effort and collected more than

18

140,000 signatures to place H.B. 2305 on the ballot for a straight up-or-down vote. (Tr.

19

1071-72; Ex. 91 at 11.) Had H.B. 2305 been repealed by referendum, the legislature

20

could not have enacted related legislation except on a supermajority vote, and only to

21

“further[] the purposes” of the referendum. Ariz. Const. art. 4, pt. 1, § 1(6)(C), (14).

22

Rather than face a referendum, Republican legislators again repealed their own

23

legislation along party lines. The bill’s primary sponsor, Secretary Reagan (who, at the

24

time, was a State Senator), admitted that the legislature’s goal was to break the bill into

25

smaller pieces and reintroduce individual provisions “a la carte.” (Ex. 91 at 11.)

26

Although the circumstances surrounding these prior bills are somewhat suspicious,

27

these departures have less probative value because they involve different bills passed

28

during different legislative sessions by a substantially different composition of

- 79 -

1

legislators. See Burton v. City of Belle Glade, 178 F.3d 1175, 1195 (11th Cir. 1999)

2

(“[W]e fail to see how evidence of . . . a [city’s] prior refusal to annex [a housing

3

project] standing alone establishes any intent, let alone a discriminatory one” for later

4

annexation decision); Kansas City, Mo. v. Fed. Pac. Elec. Co., 310 F.2d 271, 278 (8th

5

Cir. 1962) (noting the “questionable import that the rejection of prior bills may have in

6

determining congressional intent as to subsequently enacted legislation”).

7

Plaintiffs also claim that H.B. 2023 represents a substantive deviation from normal

8

legislative processes because it differs from these prior bills. But the fact that different

9

bills from different sponsors and different legislative sessions did not have the same

10

substance is not alone surprising, nor is it particularly probative of discriminatory intent.

11

Moreover, although Plaintiffs argue that the legislature made H.B. 2023 harsher than

12

previous ballot collection bills by imposing felony penalties, they ignore that H.B. 2023

13

in other respects is more lenient than its predecessors given its broad exceptions for

14

family members, household members, and caregivers.

15

Finally, Plaintiffs highlight the law’s impact on minority voters. As previously

16

noted, ballot collection was used as a GOTV strategy in mostly low-efficacy minority

17

communities, though the Court cannot say how often voters used ballot collection, nor

18

can it measure the degree or significance of any disparities in its usage. The legislature

19

was aware that the law could impact GOTV efforts in low-efficacy minority

20

communities; numerous democratic lawmakers speaking in opposition to the bill

21

expressed concerns that it would adversely impact minority GOTV efforts. (Ex. 17 at 74;

22

Ex. 19 at 17-18, 20, 35-37; Ex. 23 at 89-91; Ex. 25 at 27-28.) But this evidence shows

23

only that the legislature enacted H.B. 2023 in spite of its impact on minority GOTV

24

efforts, not because of that impact. Indeed, proponents of the bill seemed to view these

25

concerns as less significant because of the minimal burdens associated with returning a

26

mail ballot. (See, e.g., Ex. 23 at 81-82.)

27

Based on the totality of the circumstances, Plaintiffs have not shown that the

28

legislature enacted H.B. 2023 with the intent to suppress minority votes. Rather, some

- 80 -

1

individual legislators and proponents were motivated in part by partisan interests.

2

Shooter, for example, first raised concerns about ballot collection after winning a close

3

election. In addition to raising concerns about ballot collectors impersonating election

4

workers, Johnson complained that ballot collection put candidates “who don’t have

5

accessibility to large groups to go out and collect those ballots” at a disadvantage.

6

Likewise, Richard Hopkins, a proponent of the bill and a 2014 Republican candidate for

7

the Arizona House of Representatives, claimed that he lost his election because of “ballot

8

harvesting.” (Ex. 17 at 17, 45-49.) In opposing ballot collection restrictions, Democratic

9

Senator Steve Farley stated “[t]he problem we’re solving is that one party is better at

10
11

collecting ballots than the other one.” (Ex. 25 at 35.)
But partisan motives are not necessarily racial in nature, even though racially

12

polarized voting can sometimes blur the lines.

13

Amendment and § 2 of the VRA—upon which Plaintiffs’ intentional discrimination

14

claims are based—address racial discrimination, not partisan discrimination. That some

15

legislators and proponents harbored partisan interests, rather than racially discriminatory

16

motives, is consistent with Arizona’s history of advancing partisan objectives with the

17

unintended consequence of ignoring minority interests. (Ex. 90 at 8.)

Importantly, both the Fifteenth

18

Moreover, partisan motives did not permeate the entire legislative process.

19

Instead, many proponents acted to advance facially important interests in bringing early

20

mail ballot security in line with in-person voting security, notwithstanding the lack of

21

direct evidence that ballot collection fraud was occurring.

22

disagree with the manner in which the legislature chose to address its concerns about

23

early ballot security, “the propriety of doing so is perfectly clear,” and the legislature

24

need not wait until a problem occurs to take proactive steps it deems appropriate.

25

Crawford, 553 U.S. at 196; see also Lee, 188 F. Supp. 3d at 609.

Though Plaintiffs might

26

The Court therefore finds that the legislature that enacted H.B. 2023 was not

27

motivated by a desire to suppress minority voters. The legislature was motivated by a

28

misinformed belief that ballot collection fraud was occurring, but a sincere belief that

- 81 -

1

mail-in ballots lacked adequate prophylactic safeguards as compared to in-person voting.

2

Some proponents also harbored partisan motives. But, in the end, the legislature acted in

3

spite of opponents’ concerns that the law would prohibit an effective GOTV strategy in

4

low-efficacy minority communities, not because it intended to suppress those votes.

5

VIII. CONCLUSION

6

Plaintiffs have not carried their burden to show that the challenged election

7

practices severely and unjustifiably burden voting and associational rights, disparately

8

impact minority voters such that they have less opportunity than their non-minority

9

counterparts to meaningfully participate in the political process, or that Arizona was

10

motivated by a desire to suppress minority turnout when it placed limits on who may

11

collect early mail ballots. Plaintiffs have raised fair concerns about the wisdom of H.B.

12

2023 and Arizona’s treatment of OOP ballots as matters of public policy. The Court,

13

however, is not charged with second-guessing the prudence of Arizona’s laws. The

14

Court’s authority extends only to determining whether, in exercising its constitutional

15

authority to regulate the times, places, and manner of elections, Arizona has acted within

16

permissible constitutional and statutory bounds. In exercising this duty, the Court also is

17

constrained by decisions of the Supreme Court, including those standing for the

18

proposition that legislatures may act prophylactically rather than upon specific evidence

19

of a documented problem, and those finding that prevention of voter fraud and

20

preservation of public confidence in election integrity are important state interests. See

21

Purcell, 549 U.S. at 4; Crawford, 553 U.S. at 195; Munro, 479 U.S. at 194-95; Eu, 489

22

U.S. at 231. Based on a careful review of the evidence and governing case law, the Court

23

concludes that the challenged provisions contravene neither the Constitution nor the

24

VRA. Therefore,

25

IT IS ORDERED as follows:

26

1.

27
28

Defendants’ oral motion for judgment on partial findings (Doc. 384) is

DENIED as moot.
2. The Court finds in favor of Defendants and against Plaintiffs on all claims.

- 82 -

1
2
3

3. The Clerk of the Court shall enter judgment accordingly and terminate this
case.
Dated this 8th day of May, 2018.

4
5
6
7
8

Douglas L. Rayes
United States District Judge

9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

- 83 -