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

Filing 416

AMENDED ORDER Findings of Fact and Conclusions of Law - Defendants' oral motion for judgment on partial findings (Doc. 384 ) is DENIED as moot. The Court finds in favor of Defendants and against Plaintiffs on all claims. The Clerk shall enter judgment accordingly and terminate this case. Signed by Judge Douglas L Rayes on 5/10/18. (DXD)

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

No. CV-16-01065-PHX-DLR
AMENDED FINDINGS OF FACT
AND CONCLUSIONS OF LAW1

v.
Michele Reagan and Mark Brnovich,
Defendants.

14
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
27
28

1

This order amends the Court’s May 8, 2018 order (Doc. 412) to: (1) correct five
non-substantive typographical errors on pages 50 at line 5, 61 at lines 18 and 23, 64 at
line 6, and 69 at line 10 of the original order; and (2) replace the words “qualitative” and
“qualitatively” on pages 56, 58, and 62 of the original order with more accurate and
precise modifiers. The substance of the order remains the same.

1

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

2

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

3

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

4

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

5

and Native American voters. (Doc. 360 at 4-7.)2 Plaintiffs seek (1) a declaration that the

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

claims.3

14

I. PARTIES

15

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

16

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

17

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

18

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

19

committee dedicated to encouraging the election of Democratic Senate candidates to

20

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

21

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

22

public office throughout Arizona.

23

Defendants are Arizona Secretary of State Michele Reagan and Arizona Attorney

24

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

25
26
27
28

2

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.
3
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

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

2

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

3

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

4

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

Secretary Reagan

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

II. OVERVIEW OF TRIAL TESTIMONY

19

A. Plaintiffs’ Expert Witnesses

20

1. Dr. Allan Lichtman

21

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

22

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

23

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

24

Arts and Sciences at American University.

25

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

26

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

27

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

28

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

-3-

He received his B.A. in History from

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

voters.4

11

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

12

registration and turnout data, and scientific surveys.

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

13

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

14

The surveys and data he supplied reveal significant socioeconomic disparities between

15

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

16

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

17

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

18

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

19

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

20

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

21

at 4-5.)

22

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

23

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

24

find Dr. Lichtman’s ultimate opinions useful.

25

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

Dr. Lichtman applied the law as he

26
27
28

4

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

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

2

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

3

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

4

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

5

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

6

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

7

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

8

2. Dr. David Berman

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

89 at 3-4.)

23

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

24

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

25

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

26

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

27

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

28

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

-5-

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

3. Dr. Jonathan Rodden

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

oriented research related to redistricting and representation.

24

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

He also has worked

25

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

26

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

27

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

28

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

-6-

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

(Ex. 95 at 3-4.)

12

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

13

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

14

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

15

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

16

in Pima County.

17

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

18

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

19

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

20

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

21

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

22

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

23

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

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

24

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

25

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

26

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

27

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

28

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

-7-

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

its limitations and affords these opinions moderate weight.

15

B. Plaintiffs’ Lay Witnesses

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26
27
28

C. Defendants’ Expert Witnesses
1. Dr. Donald Critchlow
Dr. Donald Critchlow works at Arizona State University as the Director of the

-8-

1

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

2

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

3

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

4

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

5

to combat discrimination and encourage participation in voting.

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

Court affords little weight to Dr. Critchlow’s opinions

18

Additionally, although Dr. Critchlow acknowledged that

2. Sean Trende

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

-9-

1

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

2

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

3

Court affords Trende’s opinions little weight.

4

3. Dr. M.V. Hood

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 10 -

1

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

2

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

3

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

4

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

5

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

6

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

7

4. Dr. Janet Thornton

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

measurement strategies recommended by Dr. Thornton.

25

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

26

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

27

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

28

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

- 11 -

1

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

2

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

3

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

4

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

5

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

6

therefore affords them less weight.

7

D. Defendants’ Lay Witnesses

8

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

9

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

10

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

11

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

These witnesses

12

E. Witnesses Testifying By Deposition

13

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

14

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

15

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

16

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

17

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

18

outstanding evidentiary matters, in a separate order.

19

III. OVERVIEW OF CHALLENGED ELECTIONS PRACTICES

20

A. H.B. 2023

21

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

22

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

23

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

24

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

25

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

26

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

27

¶ 59.)

28

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

- 12 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

361 ¶¶ 57, 61.)

19

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

20

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

21

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

22

voter’s voted or unvoted early ballot:

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
- 13 -

1

conduct elections pursuant to title 48.

2

2. A family member, household member or caregiver of the
voter. For the purposes of this paragraph:

3

(a) “Caregiver” means a person who provides medical or
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.

4
5
6

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

7
8
9

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

10

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

11
12

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

13

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

14

not entrust other, unauthorized third parties to do so.

15

B. Rejection of OOP Ballots

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

well within the mainstream on this issue.5 Indeed, at no point has the DOJ objected to

28

5

See Ala. Code §§ 17-9-10, -10-2, -10-3; Ark. Code Ann. §§ 7-5-306(b), -

- 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
23
24
25
26
27
28

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), 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

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

2

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

3

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

4

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

5

(2008).

6

Plaintiffs have organizational standing to challenge the election regulations at

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

requirements sufficient to establish standing).

23
24
25
26
27
28

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.

- 16 -

1

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

2

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

3

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

4

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

5

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

6

members to participate in this lawsuit.

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

such violations, A.R.S. § 16-1021.

23

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

24

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

25

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

26

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

27

621(A).

28

Although county officials are responsible for

Though the Court cannot require Secretary Reagan and Attorney General

- 17 -

1

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

2

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

3

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

4

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

5

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

6

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

7

like H.B. 2023.

8

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

9

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

10

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

11

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

12

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

13

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

14

therefore Attorney General Brnovich likely will share enforcement authority in most

15

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

16

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

17

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

18

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

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

19

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

20

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

21

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

22

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

23

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

24

as much.

25

B. Effect of Preliminary Appellate Proceedings

26

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

27

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

28

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

- 18 -

1

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

2

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

3

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

4

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

5

Chief Judge Thomas dissented.

6

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

7

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

8

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

9

for the reasons provided in Chief Judge Thomas’ dissent.6 This preliminary injunction

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

And with that, the Court proceeds to the merits.

24

V. FIRST AND FOURTEENTH AMENDMENTS7

25
26
27
28

6

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.
7
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,

- 19 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

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
45 n.1 (1994).

- 20 -

1
2
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

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.
Under this framework, the degree to which the Court scrutinizes “the propriety of
a state election law depends upon the extent to which a challenged regulation burdens
First and Fourteenth Amendment rights.” Id. A law that imposes severe burdens is
subject to strict scrutiny, meaning it must be narrowly tailored to serve a compelling state
interest. Id. “Regulations imposing . . . [l]esser burdens, however, trigger less exacting
review, and a State’s ‘important regulatory interests’ will usually be enough to justify
‘reasonable, nondiscriminatory restrictions.’” Timmons v. Twin Cities Area New Party,
520 U.S. 351, 358 (1997) (quoting Burdick, 504 U.S. at 434); see also Pub. Integrity
Alliance, 836 F.3d at 1024 (“Applying these precepts, ‘[w]e have repeatedly upheld as
‘not severe’ restrictions that are generally applicable, evenhanded, politically neutral, and
protect the reliability and integrity of the election process.’” (quoting Dudum v. Arntz,
640 F.3d 1098, 1106 (9th Cir. 2011)). Additionally, when applying Anderson/Burdick,
the Court considers the state’s election regime as a whole, including aspects that mitigate
the hardships that might be imposed by the challenged provisions. See Ohio Democratic
Party v. Husted, 834 F.3d 620, 627 (6th Cir. 2016); see also Crawford v. Marion Cty.
Election Bd., 553 U.S. 181, 199 (2008) (considering mitigating aspects of Indiana’s
election laws).
A. Application to H.B. 2023
1. Burden on Voting Rights
At most, H.B. 2023 minimally burdens Arizona voters as a whole. In fact, the vast
majority of Arizona voters are unaffected by the law. Although voting by early mail
ballot has steadily increased in Arizona, in any given election there remains a subset of
voters who choose to vote in person, either early at a designated early voting site or on
Election Day. In-person voters are not impacted by limitations on who may collect early
- 21 -

1

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

2

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

3

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

4

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

5

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

6

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

7

2023 has no impact on these voters.

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 22 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

voting period.8

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

the law support such an argument.

17

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

18

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

19

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

20

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

21
22
23
24
25
26
27
28

8

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

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

2

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

3

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

4

86.

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

increase over the usual burdens of voting.”9 Id. The Court also reasoned that “[t]he

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26
27
28

9

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

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

2

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

3

Arizona voters.

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

challenged law).

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 25 -

1

Id. at 200.

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

election system adequately mitigate those burdens.

16

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

17

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

18

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

19

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

20

experience returning early mail ballots presents special challenges for communities that

21

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

22

socioeconomically disadvantaged voters who lack reliable transportation; voters who

23

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

24

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

25

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

26

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

27

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

28

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

- 26 -

1

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

2

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

3

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

4

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

5

understanding and completing a ballot.

6

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

7

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

8

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

9

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

10

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

11

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

12

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

13

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

14

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

15

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

16

necessary.10

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25
26
27
28

10

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

members, household members, caregivers, or elections officials.

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

daughter-in-law.

11

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

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

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

21

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

22

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

23

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

24

polling place when he votes in person.

25

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

26

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

27

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

28

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

- 28 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

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

11

via a special election board.

12

Democratic Party from assisting her with making those arrangements.

Nothing in H.B. 2023 prevents volunteers from the

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 29 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

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

9

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

10

to a mailbox; she simply must remember to timely mail her ballot.11 (Tr. 128, 130, 133,

11

135, 142.)

12

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

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

the post office on his own.

25

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

26
27
28

11

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

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

2

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

3

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

4

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

5

identifiable segments of the voting public.”

6

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

7

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

8

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

9

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

10

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

11

because the alternatives are particularly difficult.

12

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

2. Burden on Associational Rights

13

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

14

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

15

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

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 31 -

1

1500.)

2

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

3

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

4

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

5

associational rights of groups that encourage and facilitate voting through ballot

6

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

7

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

8

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

9

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

10

n.5 (1984).

11

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

12

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

13

express an idea.”

14

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

15

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

16

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

17

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

18

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

19

argument that collecting and delivering voter registration applications were inherently

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

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

- 32 -

Rather, the First

1
2

completed early ballot and delivering it to the proper place.
Moreover, assuming that H.B. 2023 implicates protected associational rights, it

3

does not impose severe burdens.

4

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

5

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

6

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

7

recorder’s offices, or polling places.

8

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

9

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

10

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

11

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

12

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

13

registration organizations may use to solicit new voters and distribute registration

14

applications.

15

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

16

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

17

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

18

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

19

Nothing in H.B. 2023 prevents Plaintiffs from

See id. at 393 (noting that voter registration

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

3. Justifications

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

integrity.

28

Fraud prevention and preserving public confidence in election integrity are facially

- 33 -

1

important state regulatory interests.

2

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

3

participatory democracy.

4

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

5

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

6

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

7

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

8

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

9

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

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

Voter fraud drives honest citizens out of the democratic

10

clear.”).

11

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

12

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

13

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

14

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

15

collection charged in Arizona.

16

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

17

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

18

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

19

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

20

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

21

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

22

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

23

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

24

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

25

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

26

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

27

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

28

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

- 34 -

1

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

2

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

3

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

4

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

5

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

6

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

7

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

8

polling data indicating that Arizonans lacked confidence in the State’s election system at

9

the time she introduced the bill. (Tr. 1805.)

10

Although there is no direct evidence of ballot collection fraud or of widespread

11

public perception that ballot collection undermined election integrity, Arizona’s

12

legislature is not limited to reacting to problems as they occur, nor is it required to base

13

the laws it passes on evidence that would be admissible in court. See Voting for Am., 732

14

F.3d at 394 (explaining that states “need not show specific local evidence of fraud in

15

order to justify preventative measures”).

16

supporting Arizona’s concerns might be appropriate if H.B. 2023 severely burdened the

17

franchise. But because H.B. 2023’s burdens are at most minimal, the Court’s review is

18

less exacting. Timmons, 520 U.S. at 358.

A more exacting review of the evidence

19

For example, in Crawford the Supreme Court upheld Indiana’s voter identification

20

requirement as a measure designed to prevent in-person voter fraud even though “[t]he

21

record contain[ed] no evidence of any such fraud actually occurring in Indiana at any

22

time in its history.” 553 U.S. at 195. Similarly, in Munro v. Socialist Workers Party, the

23

Supreme Court upheld a Washington law requiring all minor party candidates for partisan

24

office to receive at least one percent of all votes cast during the primary election in order

25

to appear on the general election ballot. 479 U.S. 189 (1986). Washington argued that

26

the law prevented voter confusion from ballot overcrowding by ensuring candidates

27

appearing on the general election ballot had sufficient community support. Id. at 194. In

28

upholding the law, the Supreme Court explained: “We have never required a State to

- 35 -

1

make a particularized showing of the existence of voter confusion, ballot overcrowding,

2

or the presence of frivolous candidates prior to the imposition of reasonable restrictions

3

on ballot access.” Id. at 194-95. Rather, “[l]egislatures . . . should be permitted to

4

respond to potential deficiencies in the electoral process with foresight rather than

5

reactively[.]” Id. at 195; see also Lee v. Va. State Bd. of Elections, 188 F. Supp. 3d 577,

6

609 (E.D. Va. 2016) (“Outlawing criminal activity before it occurs is not only a wise

7

deterrent, but also sound public policy.”), aff’d, 843 F.3d 592 (4th Cir. 2016).

8

Furthermore, many courts have recognized that absentee voting presents a greater

9

opportunity for fraud. See Crawford, 553 U.S. at 225 (Souter, J. dissenting) (noting that

10

“absentee-ballot fraud . . . is a documented problem in Indiana”); Griffin v. Roupas, 385

11

F.3d 1128, 1131 (7th Cir. 2004) (“Voting fraud . . . is facilitated by absentee voting.”);

12

Qualkinbush v. Skubisz, 826 N.E.2d 1181, 1197 (Ill. App. Ct. 2004) (“It is evident that

13

the integrity of the vote is even more susceptible to influence and manipulation when

14

done by absentee ballot.”). Indeed, mail-in ballots by their very nature are less secure

15

than ballots cast in person at polling locations. Accordingly, the Court finds that the

16

regulatory interests Arizona seeks to advance are important.

17

The question then becomes one of means-end tailoring. Because H.B. 2023 does

18

not impose severe burdens, it need not be narrowly tailored to achieve the State’s goals.

19

Nevertheless, the Court still must take into consideration the extent to which Arizona’s

20

important regulatory interests make it necessary to impose those minimal burdens.

21

Burdick, 504 U.S. at 434

22

Plaintiffs contend that H.B. 2023 is not necessary because Arizona law already

23

includes measures designed to ensure the security of early mail ballots, and because H.B.

24

2023 is unlikely to be a useful tool to prevent or deter voter fraud or to preserve public

25

confidence in election integrity.

26

discarding someone else’s ballot all were illegal prior to the passage of H.B. 2023.

27

(Shooter Dep. 51:16-52:5.)

28

knowingly collects voted or unvoted ballots and does not turn those ballots in to an

For example, ballot tampering, vote buying, or

Arizona law has long provided that any person who

- 36 -

1

elections official is guilty of a class 5 felony. A.R.S. § 16-1005. Further, Arizona has

2

long made all of the following class 5 felonies: “knowingly mark[ing] a voted or unvoted

3

ballot or ballot envelope with the intent to fix an election;” “receiv[ing] or agree[ing] to

4

receive any consideration in exchange for a voted or unvoted ballot;” possessing

5

another’s voted or unvoted ballot with intent to sell; “knowingly solicit[ing] the

6

collection of voted or unvoted ballots by misrepresenting [one’s self] as an election

7

official or as an official ballot repository or . . . serv[ing] as a ballot drop off site, other

8

than those established and staffed by election officials;” and “knowingly collect[ing]

9

voted or unvoted ballots and . . . not turn[ing] those ballots in to an election official . . . or

10

any . . . entity permitted by law to transmit post.” A.R.S. §§ 16-1005(a)-(f). The early

11

voting process also includes a number of other safeguards, such as tamper evident

12

envelopes and a rigorous voter signature verification procedure. (Tr. 834-35, 1563-66,

13

1752, 1878, 2209.)

14

Plaintiffs also note that, to the extent Arizona wanted to create a chain of custody

15

for early ballots, the legislature rejected a less restrictive amendment to H.B. 2023

16

proposed by Representative Ken Clark and Senator Martin Quezada, which would have

17

allowed ballot collection if the collector issued a tracking receipt. (Shooter Dep. at

18

50:21-23; Ex. 91 at 12; Ex. 16 at 54.) As enacted, H.B. 2023 is less effective at creating

19

a chain of custody because it allows certain individuals to possess another’s voted early

20

ballot but does not require a record of that collection. (Reagan Dep. 83:25-85:20.) H.B.

21

2023 also is not enforced by county recorders. (Ex. 526 at 5 n.15; Ex. 75.) Instead,

22

county recorders will accept all ballots, even those returned by prohibited possessors

23

under H.B. 2023.

24

Plaintiffs raise fair concerns about whether, as a matter of public policy, H.B.

25

2023 is the best way to achieve Arizona’s stated goals. If H.B. 2023 severely burdened

26

the franchise, and Arizona consequently was required to narrowly tailor the law to

27

achieve compelling ends, Plaintiffs’ arguments would carry more weight. But because

28

H.B. 2023’s burdens are minimal, and the Court’s review consequently less exacting,

- 37 -

1

H.B. 2023’s means-end fit can be less precise.

2

Defendants contend that one of H.B. 2023’s purposes is to reduce the opportunity

3

for early mail ballot fraud by limiting who may possess a voter’s early ballot. They also

4

use the term “fraud” broadly to encompass not just vote tampering, which is amply

5

addressed by other provisions of Arizona law, but also early ballot loss or destruction.

6

By limiting who may possess another’s early ballot, H.B. 2023 reasonably reduces

7

opportunities for early ballots to be lost or destroyed.

8

Although Arizona’s legislature arguably could have addressed this concern

9

through a more narrowly tailored, but also more complex, system of training and

10

registering ballot collectors and requiring tracking receipts or other proof of delivery, the

11

Constitution does not require Arizona to erect such a bureaucracy if the alternative it has

12

chosen is not particularly burdensome. Arizona reasonably chose to limit possession of

13

early ballots to the voter herself, and to a handful of presumptively trustworthy proxies,

14

such as family and household members.

15

recommendation of the bipartisan Commission on Federal Election Reform, chaired by

16

former President Jimmy Carter and former Secretary of State James A. Baker III, which

17

in 2005 wrote:

18

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.

19
20
21
22
23
24

Building Confidence in U.S. Elections § 5.2 (Sept. 2005) (“Carter-Baker Report”),

25

available at https://www.eac.gov/assets/1/6/Exhibit%20M.PDF.12 Though it might not

26
27
28

12

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.

- 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
25
26
27
28

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.

- 39 -

1

in Graham, Greenlee, Cochise, Navajo, Yavapai, and Yuma counties, which have

2

adopted the vote center model.

3

Moreover, the vast majority of in-person voters successfully vote in their assigned

4

precincts, and OOP voting has consistently declined as a percentage of the total ballots

5

cast in Arizona. In the 2008 general election, Arizona voters cast 14,885 OOP ballots out

6

of the 2,320,851 ballots cast statewide, meaning OOP ballots constituted 0.64 percent of

7

all votes cast in that election. In the 2012 general election, Arizona voters cast 10,979

8

OOP ballots out of the 2,323,579 ballots cast statewide, accounting for 0.47 percent of all

9

votes cast. In that same election, 1,542,855 Arizona voters submitted early ballots, and

10

more than 99 percent were counted. In the 2016 general election, Arizona voters cast

11

3,970 OOP ballots out of the 2,661,497 ballots cast statewide, representing 0.15 percent

12

of all votes cast. Since 2008, OOP voting during general presidential elections has

13

declined 73 percent statewide, dropping from 14,885 in 2008 to 3,970 in the 2016

14

election. (Tr. 1927-32; Exs. 578, 581.)

15

OOP voting has declined in midterm elections, as well. In the 2010 general

16

election, Arizona voters cast 4,919 OOP ballots out of the 1,750,840 ballots cast

17

statewide, constituting 0.28 percent of all votes cast. By comparison, in the 2014 general

18

election, Arizona voters cast 3,582 OOP ballots out of the 1,537,671 ballots cast

19

statewide, constituting 0.23 percent of all votes cast. During this same period, the

20

number of registered voters in Arizona increased as follows: 2,987,451 in 2008;

21

3,146,418 in 2010; 3,124,712 in 2012; 3,235,963 in 2014; and 3,588,466 in 2016. (Exs.

22

577, 578.)

23

These trends also hold true at the county level. For example, Maricopa County

24

(Arizona’s most populous) has experienced a consistent decline in the number of OOP

25

ballots, both in terms of raw numbers and as a percentage of the total ballots cast. In the

26

2008 general election, Maricopa County voters cast 9,159 OOP ballots out of the

27

1,380,571 ballots cast countywide, accounting for 0.66 percent of the all votes cast. In

28

the 2012 general election, Maricopa County voters cast 7,529 OOP ballots out of the

- 40 -

1

1,390,836 ballots cast countywide, representing 0.54 percent of all votes. In the 2016

2

general election, Maricopa County voters cast 2,197 OOP ballots out of the 1,608,875

3

ballots cast countywide, representing 0.14 percent of all votes cast. Likewise, in the 2010

4

general midterm election, Maricopa County voters cast 3,527 OOP ballots out of the

5

1,004,125 ballots cast countywide, accounting for 0.35 percent of all votes. In the 2014

6

general election, Maricopa County voters cast 2,781 OOP ballots out of the 877,187

7

ballots cast countywide, constituting 0.32 percent of all votes. Between 2008 and 2016,

8

Maricopa County had a staggering decrease of 76 percent in the raw number of OOP

9

ballots. During this same period, the number of registered voters in Maricopa County

10

increased as follows: 1,730,886 in 2008; 1,851,956 in 2010; 1,817,832 in 2012;

11

1,935,729 in 2014; and 2,161,716 in 2016. (Exs. 579, 582.)

12

Pima County (Arizona’s second most populous) also has experienced a consistent

13

decline in OOP voting. In the 2008 general election, Pima County voters cast 3,227 OOP

14

ballots out of the 397,503 ballots cast countywide, accounting for 0.81 percent of all

15

votes. In the 2012 general election, Pima County voters cast 2,212 OOP ballots out of

16

the 385,725 ballots cast countywide, accounting for 0.57 percent of all votes. In the 2016

17

general election, Pima County voters cast 1,150 OOP ballots out of the 427,102 ballots

18

cast countywide, representing 0.27 percent of all votes. As for Pima County midterm

19

elections, in the 2010 general election Pima County voters cast 641 OOP ballots out of

20

the 318,995 ballots cast countywide, or 0.20 percent of all votes. By comparison, in the

21

2014 general election, Pima County voters cast just 371 OOP ballots out of the 274,449

22

ballots cast countywide, constituting 0.14 percent of the total ballots. The raw number of

23

OOP ballots thus dropped by 64 percent in Pima County between 2008 and 2016. During

24

this same period, the number of registered voters in Pima County increased as follows:

25

498,777 in 2008; 486,697 in 2010; 494,630 in 2012; 497,542 in 2014; and 544,270 in

26

2016. (Exs. 580, 583.)

27

In light of these figures, and much like their H.B. 2023 claim, Plaintiffs’ challenge

28

to Arizona’s treatment of OOP ballots is best described as follows: Arizona’s rejection of

- 41 -

1

OOP ballots has no impact on the vast majority of Arizona voters, though a small subset

2

of voters is affected more often because of their special circumstances. But Plaintiffs’

3

contention that Arizona’s rejection of OOP ballots severely burdens this small subset of

4

voters is unavailing for two independent reasons.

5

First, Plaintiffs do not directly challenge the electoral practices actually

6

responsible for higher rates of OOP voting. For example, high rates of residential

7

mobility are associated with higher rates of OOP voting. Almost 70 percent of Arizonans

8

have changed their residential address in the decade between 2000 and 2010, the second

9

highest rate of any state. The vast majority of Arizonans who moved in the last year

10

moved to another address within their current city of residence and, compared with other

11

states, Arizona has the second highest rate of within-city moves. Most of these within-

12

city moves took place in Maricopa and Pima Counties. (Ex. 95 at 11-12.) Relatedly,

13

rates of OOP voting are higher in neighborhoods where renters make up a larger share of

14

householders. (Ex. 96 at 41.) One significant reason residential mobility tends to result

15

in higher rates of OOP voting is because voters who move sometimes neglect to timely

16

update their voter registration. (See, e.g., Tr. 602-06.) Relatedly, voters registered for

17

PEVL who move and do not update their address information will not have their early

18

ballot forwarded to their new address. Arizona in-person voters are more likely to vote

19

OOP if they have signed up for the PEVL and have moved. (See, e.g., Tr. 124, 987-89.)

20

Additionally, changes in polling locations from election to election, inconsistent

21

election regimes used by and within counties, and placement of polling locations all tend

22

to increase OOP voting rates. (Ex. 95 at 12-15, 26-27, 44-52, 54-58.) In Maricopa

23

County, between 2006 and 2008 at least 43 percent of polling locations changed from one

24

year to the next. Likewise, approximately 40 percent of Maricopa County’s active

25

registered voters’ polling locations changed between 2010 and 2012.

26

Maricopa County polling locations and election regimes continued to occur in 2016,

27

when Maricopa County experimented with 60 vote centers for the presidential preference

28

election, then reverted to a precinct-based system with 122 polling locations for the May

- 42 -

Changes in

1

special election, and then implemented over 700 assigned polling places in the August

2

primary and November general elections. The OOP voting rate was 40 percent higher for

3

voters who had experienced such polling place changes. (Ex. 95 at 14-15, 56-57.)

4

Further, some individual voters testified that they arrived at an incorrect polling place but

5

were not redirected by poll workers to the correct location, nor were the implications of

6

casting a provisional ballot explained. These voters stated that they would have gone to

7

the correct polling location had they been so advised. (Tr. 120, 265-66, 352-54, 493,

8

935-36.)

9

Plaintiffs do not challenge as unconstitutional the manner in which Arizona and its

10

counties allocate or relocate polling places, inform voters of their assigned precincts, or

11

train poll workers. They do not challenge Arizona’s requirement that voters update their

12

voter registrations after moving to a new address. Nor do Plaintiffs challenge Arizona’s

13

use of the precinct-based system, though the logical implication of their argument is that

14

Arizona may utilize a precinct-based system but cannot enforce it as to races for which an

15

OOP voter otherwise would be eligible to vote (usually so-called “top of the ticket”

16

races). (Tr. 1495-96.) Instead, Plaintiffs challenge what Arizona does with OOP ballots

17

after they have been cast. But there is no evidence that it will be easier for voters to

18

identify their correct precincts if Arizona eliminated its prohibition on counting OOP

19

ballots. Though the consequence of voting OOP might make it more imperative for

20

voters to correctly identify their precincts, it does not increase the burdens associated

21

with doing so.

22

Second, the burdens imposed by precinct-based voting—a system which, again,

23

Plaintiffs do not directly challenge—are not severe.

24

requires voters to locate and travel to their assigned precincts, which are ordinary burdens

25

traditionally associated with voting.13

26
27
28

Precinct-based voting merely

See Colo. Common Cause v. Davidson, No.

13

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

- 43 -

1

04CV7709, 2004 WL 2360485, at *14 (Colo. Dist. Ct. Oct. 18, 2004) (“[I]t does not

2

seem to be much of an intrusion into the right to vote to expect citizens, whose judgment

3

we trust to elect our government leaders, to be able to figure out their polling place.”); see

4

also Serv. Emps. Int’l Union Local 1 v. Husted, 698 F.3d 341, 344 (6th Cir. 2012)

5

(explaining that voters cannot be absolved “of all responsibility for voting in the correct

6

precinct or correct polling place by assessing voter burden solely on the basis of the

7

outcome—i.e., the state’s ballot validity determination”).

8
9

Moreover, Arizona does not make it needlessly difficult for voters to find their
assigned precincts.

Indeed, a 2016 Survey of Performance of American Elections

10

(“SPAE”) found that none of the survey respondents for Arizona reported that it was

11

“very difficult” to find their polling places. By comparison, several other states had

12

respondents who reported that it was very difficult to find their polling places. The 2016

13

SPAE also reported that approximately 94 percent of the Arizona respondents thought it

14

was very easy or somewhat easy to find their polling places. (Tr. 1350-51.)

15

In Arizona counties with precinct-based systems, voters generally are assigned to

16

precincts near where they live, and county officials consider access to public

17

transportation when assigning polling places. (Tr. 1570-73.) Arizona voters also can

18

learn of their assigned precincts in a variety of ways. (See Ex. 526 at 11-18.)

19

If precincts or polling places have been altered since the previous election,

20

registered voters are sent a mailing informing them of this fact and of where their new

21

polling places are located. (Tr. 1575-76.) State law requires that election officials send

22

each household with a registered voter who is not on the PEVL a sample ballot at least

23

eleven days prior to election day, A.R.S. § 16-510(C), which contains instructions and

24

identifies their polling location. (Doc. 361 ¶ 52.) The Secretary of State’s Office

25

operates several websites that make voter-specific polling place information available and

26
27
28

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.

- 44 -

1

allow the Secretary’s staff to respond directly to voter inquiries. The Secretary of State’s

2

Office also mails a publicity pamphlet to voters, which includes information on how to

3

locate their correct precincts. This information is provided in English and Spanish. The

4

Secretary also uses social media, town halls, and live events (such as county and state

5

fairs) to register voters and answer questions.

6

In addition, several Arizona counties, including Maricopa and Pima Counties,

7

operate online polling place locators that are available in English and Spanish. Voters

8

also can learn their assigned polling locations by calling the office of the county recorder

9

for the county in which they reside. Counties spread awareness about polling place

10

locations and the consequences of OOP voting through news and social media. This

11

information is communicated in both English and Spanish. Some counties—including

12

the state’s most populous, Maricopa and Pima—post signs at polling places informing

13

voters that OOP ballots will not be counted. (Tr. 1586-88; Ex. 368.) Poll workers also

14

are trained to direct voters who appear at an incorrect polling location to their correct

15

polling location and to notify such voters that their votes will not be counted if they vote

16

with a provisional ballot at the wrong location.

17

The Arizona Citizens Clean Elections Commission (“CCEC”) operates a website

18

in English and Spanish that provides a tool for voters to determine their polling place.

19

The CCEC also engages in advertising to help educate voters on where to vote. Partisan

20

groups, such as the ADP and political campaigns, also help educate voters on how to find

21

their assigned polling places. (Tr. 1575-76.)

22

In sum, Arizona’s rejection of OOP ballots has no impact on the vast majority of

23

voters. Although a small and ever-dwindling subset of voters still vote OOP, how

24

Arizona treats OOP ballots after they have been cast does not make it difficult for these

25

voters to find and travel to their correct precincts. To the extent Plaintiffs’ claim may

26

properly be considered as an indirect challenge to Arizona’s strictly enforced precinct-

27

based system, the burdens imposed on voters to find and travel to their assigned precincts

28

are minimal and do not represent significant increases in the ordinary burdens

- 45 -

1

traditionally associated with voting. Moreover, for those who find it too difficult to

2

locate their assigned precinct, Arizona offers generous early mail voting alternatives.14

3

2. Justifications

4

Weighing against the minimal burdens imposed by precinct-based voting are

5

numerous important state regulatory interests. Precinct-based voting serves an important

6

planning function for Arizona counties by helping them estimate the number of voters

7

who may be expected at any particular precinct, which allows for better allocation of

8

resources and personnel. In turn, orderly administration of elections helps to increase

9

voter confidence in the election system and reduces wait times. (Tr. 1608-10, 1896-913.)

10

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

11

system also ensures that each voter receives a ballot reflecting only the races for which

12

that person is entitled to vote. Precincts must be created, and ballots printed, so that the

13

residential address of every voter is connected to the correct assortment of local elected

14

officials. The system thus promotes voting for local candidates and issues and helps

15

make ballots less confusing by not providing voters with ballots that include races for

16

which they are not eligible to vote. (Ex. 95 at 10; Doc. 361 ¶ 47.)

17

Indeed, other courts have recognized these numerous and significant advantages:

18

[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.

19
20
21
22
23
24
25
26
27
28

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
14

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.

- 46 -

1

Arizona need not reject OOP ballots in their entirety to accomplish these goals. Plaintiffs

2

contend that Arizona can just as easily accomplish these goals and reap these benefits by

3

partially counting OOP ballots, accepting votes in races for which the voter is eligible to

4

vote and rejecting votes in races for which the voter is not.

5

Counting OOP ballots is administratively feasible. Twenty states partially count

6

OOP ballots. (Ex. 94 at 32-33.) These include the neighboring states of California, Utah,

7

and New Mexico. Cal. Elec. Code §§ 14310(a)(3), 14310 (c)(3), 15350; Utah Code Ann.

8

§ 20A-4-107(1)(b)(iii), 2(a)(ii), 2(c); N.M. Stat. Ann § 1-12-25.4(F); N.M. Admin. Code

9

1.10.22.9(N).

Elections administrators in these and other states have established

10

processes for counting only the offices for which the OOP voter is eligible to vote. Some

11

states, such as New Mexico, use a hand tally procedure, whereby a team of elections

12

workers reviews each OOP ballot, determines the precinct in which the voter was

13

qualified to vote, and marks on a tally sheet for that precinct the votes cast for each

14

eligible office.

15

California, use a duplication method, whereby a team of elections workers reviews each

16

OOP ballot, determines the precinct in which the voter was qualified to vote, obtains a

17

new paper ballot for the correct precinct, and duplicates the votes cast on the OOP ballot

18

onto the ballot for the correct precinct. Only the offices that appear on both the OOP

19

ballot and the ballot for the correct precinct are copied. The duplicated ballot then is

20

scanned through the optical scan voting machine and electronically tallied. (Tr. 777-81.)

See N.M. Admin Code 1.10.22.9(H)-(N).

Other states, such as

21

Arizona has a similar duplication procedure that it uses to process certain types of

22

ballots that cannot be read by an optical scan voting machine, such as ballots that are

23

damaged, marked with the wrong color pen, or submitted to the county recorder by a

24

military or overseas voter via facsimile. (Tr. 1564-66; Ex. 455 at 177-78.) Arizona also

25

uses the duplication procedure to process some provisional ballots cast by voters who are

26

eligible to vote in federal elections, but whom Arizona does not permit to vote in state

27

elections. (Ex. 455 at 187.) This duplication procedure takes about twenty minutes per

28

ballot. (Tr. 1604-606.)

- 47 -

1

If strict scrutiny applied and Arizona were required to narrowly tailor its precinct

2

enforcement to achieve compelling state interests, Plaintiffs’ critiques might carry more

3

weight. But in light of the minimal burdens associated with the precinct-based system,

4

Arizona’s policy need not be the narrowest means of enforcement.

5

Moreover, Plaintiffs are incorrect that Arizona can accomplish all of its goals

6

without its strict enforcement regime. If voters in precinct-based counties can have their

7

ballots counted for statewide and countywide races even if they vote in the wrong

8

precinct, they will have far less incentive to vote in their assigned precincts and might

9

decide to vote elsewhere. Other voters might incorrectly believe that that they can vote at

10

any location and receive the correct ballot. Voters might also be nefariously directed to

11

vote elsewhere. North Carolina, for example, has experienced a problem with “political

12

organizations intentionally transporting voters to the wrong precinct.” See N.C. State

13

Conf. of the NAACP v. McCrory, 182 F. Supp. 3d 320, 461 (M.D.N.C. 2016) rev’d on

14

other grounds, 831 F.3d 204 (4th Cir. 2016). This, in turn, would undermine both the

15

ability of Arizona counties to accurately estimate the number of voters who may be

16

expected at any particular precinct and allocate appropriate resources and personnel, and

17

Arizona’s goal of promoting voting for local candidates. Consequently, if OOP ballots

18

are partially counted in Arizona, candidates for local office will have to expend resources

19

to educate voters on why it nevertheless is important to vote within their assigned

20

precincts. Moreover, requiring counties to review all OOP ballots for any given election

21

and determine the specific contests in which each voter was eligible to vote would

22

impose a significant financial and administrative burden on Maricopa and Pima Counties

23

because of their high populations.

24

Plaintiffs’ requested relief essentially would transform Arizona’s precinct-based

25

counties, including its two most populous, into quasi-vote-center counties. But the vote-

26

center model is not appropriate for every jurisdiction.

27

polling places, it can be difficult for counties to predict the number of voters at each vote

28

center.

Compared to precinct-based

Consequently, vote centers can cause voter wait times to increase, with

- 48 -

1

corresponding decreases in turnout, due to the potential for uneven distribution of voters.

2

(Tr. 1607-611, 1896-913.) Plaintiffs’ requested relief therefore would deprive precinct-

3

based counties of the full range of benefits that correspond with the precinct-based

4

system.

5

Precinct-based voting is a quintessential time, place, and manner election

6

regulation. Arizona’s policy to not count OOP ballots is one mechanism by which it

7

enforces and administers this precinct-based system to ensure that it reaps the full extent

8

of its benefits. This policy is sufficiently justified in light of the minimal burdens it

9

imposes.

Accordingly, Arizona’s rejection of OOP ballots does not violate the

10

Fourteenth Amendment.

11

VI. SECTION 2 OF THE VOTING RIGHTS ACT (RESULTS TEST)

12

“Inspired to action by the civil rights movement, Congress responded in 1965 with

13

the Voting Rights Act.” Shelby Cty. v. Holder, 570 U.S. 529, 536 (2013). In its original

14

form, § 2 of the VRA prohibited all states from enacting any “standard, practice, or

15

procedure . . . imposed or applied . . . to deny or abridge the right of any citizen of the

16

United States to vote on account of race or color.” Id. (quoting § 2, 79 Stat. 437).

17

“At the time of passage of the Voting Rights Act of 1965, § 2, unlike other

18

provisions of the Act, did not provoke significant debate in Congress because it was

19

viewed largely as a restatement of the Fifteenth Amendment.” Chisom v. Roemer, 501

20

U.S. 380, 392 (1991). The Supreme Court took a similar view, holding in a 1980

21

plurality opinion that “the language of § 2 no more than elaborates upon that of the

22

Fifteenth Amendment,” and therefore § 2 is violated only if a state enacted the challenged

23

law with the intent to discriminate on account of race or color. City of Mobile v. Bolden,

24

446 U.S. 55, 60-62 (1980) (plurality opinion).

25

In 1982, in response to the Supreme Court’s opinion in Bolden, “Congress

26

substantially revised § 2 to make clear that a violation could be proved by showing

27

discriminatory effect alone[.]” Thornburg v. Gingles, 478 U.S. 30, 35 (1986). In its

28

current form, § 2 provides:

- 49 -

(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).

1
2
3
4

(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
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.

5
6
7
8
9
10

52 U.S.C. § 10301. To succeed on a § 2 claim, a plaintiff now may show either that the

11

challenged law was enacted with the intent to discriminate on account of race or color, or

12

that “under the totality of the circumstances, a challenged election law or procedure ha[s]

13

the effect of denying a protected minority equal chance to participate in the electoral

14

process.” Gingles, 478 U.S. at 44 n.8. “The essence of a § 2 claim” brought under the

15

so-called “effects” or “results test” “is that a certain electoral law, practice, or structure

16

interacts with social and historical conditions to cause an inequality in the opportunities

17

enjoyed by [minority and non-minority] voters to elect their preferred representatives.”

18

Id. at 47.

19

When determining whether, under the totality of the circumstances, a challenged

20

voting practice interacts with social and historical conditions to cause inequality in the

21

electoral opportunities of minority and non-minority voters, courts may consider, as

22

relevant, the following factors derived from the Senate Report accompanying the 1982

23

amendments to the VRA (“Senate Factors”):

24
25
26

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;

27

2. the extent to which voting in the elections of the state or
political subdivision is racially polarized;

28

3. the extent to which the state or political subdivision has
used unusually large election districts, majority vote
- 50 -

1
2
3
4
5
6
7

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;

8

6. whether political campaigns have been characterized by
overt or subtle racial appeals; [and]

9

7. the extent to which members of the minority group have
been elected to public office in the jurisdiction.

10
11

Id. at 36-37 (quoting S. Rep. No. 97-417, at 28-29 (1982)). Courts also may consider

12

“whether there is a significant lack of responsiveness on the part of elected officials to the

13

particularized needs of the members of the minority group,” and “whether the policy

14

underlying the state or political subdivision’s use of such voting qualification,

15

prerequisite to voting, or standard, practice or procedure is tenuous.” Id.; see also

16

Houston Lawyers’ Ass’n v. Attorney Gen. of Tex., 501 U.S. 419, 426-27 (1991)

17

(explaining that courts may consider a state’s “justification for its electoral system”).

18

“[T]here is no requirement that any particular number of factors be proved, or that a

19

majority of them point one way or the other.” Gingles, 478 U.S. at 45.

20

Until relatively recently, “Section 2’s use . . . has primarily been in the context of

21

vote-dilution cases,” which “‘involve challenges to methods of electing representatives—

22

like redistricting or at-large districts—as having the effect of diminishing minorities

23

voting strength.’” League of Women Voters of N.C. v. N.C., 769 F.3d 224, 239 (4th Cir.

24

2014) (quoting Ohio State Conference of N.A.A.C.P. v. Husted, 768 F.3d 524, 554 (6th

25

Cir. 2014), vacated on other grounds by 2014 WL 10384647 (6th Cir. 2014)). Gingles

26

itself was a vote dilution case. “While vote-dilution jurisprudence is well-developed,

27

numerous courts and commentators have noted that applying Section 2’s ‘results test’ to

28

vote-denial claims is challenging, and a clear standard for its application has not been

- 51 -

1

conclusively established.”15 Ohio Democratic Party, 834 F.3d at 636; see also Veasey v.

2

Abbott, 830 F.3d 216, 243-44 (5th Cir. 2016) (“Although courts have often applied the

3

Gingles factors to analyze claims of vote dilution . . . there is little authority on the proper

4

test to determine whether the right to vote has been denied or abridged on account of

5

race.”); League of Women Voters, 769 F.3d at 239 (“[T]here is a paucity of appellate case

6

law evaluating the merits of Section 2 claims in the vote-denial context.”); Ohio State

7

Conference, 768 F.3d at 554 (“A clear test for Section 2 vote denial claims . . . has yet to

8

emerge.”); Simmons v. Galvin, 575 F.3d 24, 41-42 n.24 (1st Cir. 2009) (“‘While Gingles

9

and its progeny have generated a well-established standard for vote dilution, a

10

satisfactory test for vote denial cases under Section 2 has yet to emerge . . . [and] the

11

Supreme Court’s seminal opinion in Gingles . . . is of little use in vote denial cases.’”

12

(quoting Daniel P. Tokaji, The New Vote Denial: Where Election Reform Meets the

13

Voting Rights Act, 57 S.C. L. Rev. 689, 709 (2006))); Janai S. Nelson, The Causal

14

Context of Disparate Vote Denial, 54 B.C. L. Rev. 579, 595 (2013) (“[T]he legal

15

contours of vote denial claims remain woefully underdeveloped as compared to vote

16

dilution claims.”). Indeed, some of the Senate Factors cited by the Gingles Court as

17

relevant to the totality of the circumstances inquiry do not seem particularly germane to

18

vote denial claims. Cf. Gingles, 478 U.S. at 45 (“While the enumerated factors will often

19

be pertinent to certain types of § 2 violations, particularly to vote dilution claims, other

20

factors may also be relevant and considered.”); see Frank v. Walker, 768 F.3d 744, 752-

21

55 (7th Cir. 2014) (questioning usefulness of Senate Factors in vote denial claims); Ohio

22

Democratic Party, 834 F.3d at 638 (explaining that totality of circumstances inquiry in

23

vote denial cases is “potentially informed by the ‘Senate Factors’ discussed in Gingles”

24

(emphasis added)).

25

Several circuit courts that recently have analyzed vote denial claims have adopted

26

the following two-part framework based on the text of § 2 and the Supreme Court’s

27
28

15

A “vote denial” claim generally refers “to any claim that is not a vote dilution
claim.” Ohio State Conference, 768 F.3d at 554

- 52 -

1

guidance in Gingles:

2

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.

3
4
5

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.

6
7
8

League of Women Voters, 769 F.3d at 240 (internal quotations and citations omitted); see

9

also Veasey, 830 F.3d at 244; Ohio Democratic Party, 834 F.3d at 636; Frank, 768 F.3d

10

at 754-55 (adopting two-park framework for sake of argument but expressing skepticism

11

of the second step “because it does not distinguish between discrimination by [the state]

12

from other persons’ discrimination”). The Ninth Circuit likewise endorsed this two-part

13

framework in the context of this case. Feldman v. Ariz. Sec. of State’s Office, 840 F.3d

14

1057, 1070 (9th Cir. 2016); id. at 1091 (Thomas, C.J. dissenting); Feldman, 843 F.3d at

15

367.

16

“The first part of this two-prong framework inquires about the nature of the

17

burden imposed and whether it creates a disparate effect[.]” Veasey, 830 F.3d at 244.

18

Drawing on the Supreme Court’s guidance in Gingles, “[t]he second part . . . provides the

19

requisite causal link between the burden on voting rights and the fact that this burden

20

affects minorities disparately because it interacts with social and historical conditions that

21

have produced discrimination against minorities currently, in the past, or both.” Id. That

22

is, “the second step asks not just whether social and historical conditions ‘result in’ a

23

disparate impact, but whether the challenged voting standard or practice causes the

24

discriminatory impact as it interacts with social and historical conditions.”

25

Democratic Party, 834 F.3d at 638 (emphasis in original).

26
27
28

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
- 53 -

Ohio

1

any evidence that the challenged voting qualification causes
that disparity, will be rejected.

2
3

Gonzales v. Arizona, 677 F.3d 383, 405 (9th Cir. 2012) (internal quotations and citations

4

omitted); see also Smith v. Salt River Project Agric. Improvement & Power Dist., 109

5

F.3d 586, 595 (9th Cir. 1997).

6

A close reading of the decisions of courts that recently have grappled with vote

7

denial claims reveals two important nuances of the results test. The first bears on the

8

meaning of “disparity” as it has been used in vote denial cases. “No state has exactly

9

equal registration rates, exactly equal turnout rates, and so on, at every stage of its voting

10

system.” Frank, 768 F.3d at 754. Perfect racial parity is unlikely to exist in any aspect of

11

a state’s election system, which is to say it is unlikely that minorities and non-minorities

12

will be impacted by laws in perfect proportion to their representation in the overall voting

13

population. Unless the VRA is to be interpreted to sweep away all elections regulations,

14

some degree of disproportionality must be tolerable.

15

Therefore, not every disparity between minority and non-minority voters is

16

cognizable under the VRA. Rather, to be cognizable the disparity must be meaningful

17

enough to work “an inequality in the opportunities enjoyed by [minority as compared to

18

non-minority] voters to elect their preferred representatives.” Gingles, 478 U.S. at 47;

19

see Gonzales, 677 F.3d at 405 (suggesting that disparity must be “relevant”).

20

example, the Seventh Circuit rejected the notion that:

21
22
23
24
25
26

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[.]

27
28

Frank, 768 F.3d at 754.

- 54 -

For

1

The second nuance bears on the definition of “impact” or “effect.”

To be

2

cognizable, the challenged voting practice must “impose a discriminatory burden,”

3

League of Women Voters, 769 F.3d at 240, and not merely result in a “disproportionate

4

impact,” Salt River Project, 109 F.3d at 595. Section 2 “does not sweep away all election

5

rules that result in a disparity in the convenience of voting.” Lee v. Va. State Bd. of

6

Elections, 843 F.3d 592, 601 (4th Cir. 2016). A contrary interpretation would require the

7

Court to accept:

8
9
10
11
12
13
14
15

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.

16

Id.; see also N.E. Ohio Coal. for the Homeless, 837 F.3d at 628 (“A law cannot

17

disparately impact minority voters if its impact is insignificant to begin with.”); Ohio

18

Democratic Party, 834 F.3d at 623 (“[W]hile the challenged regulation may slightly

19

diminish the convenience of registration and voting, it applies even-handedly to all

20

voters, and, despite the change, Ohio continues to provide generous, reasonable, and

21

accessible voting options to all Ohioans. The issue is not whether some voter somewhere

22

would benefit from six additional days of early voting or from the opportunity to register

23

and vote at the same time. Rather, the issue is whether the challenged law results in a

24

cognizable injury under the Constitution or the Voting Rights Act.”); Frank, 843 F.3d at

25

753 (“[U]nless the State of Wisconsin made it ‘needlessly hard’ to obtain the requisite

26

photo identification for voting, this requirement did not result in a ‘denial’ of anything by

27

Wisconsin, as § 2(a) requires.”); Jacksonville Coal. for Voter Prot. v. Hood, 351 F. Supp.

28

2d 1326, 1335 (M.D. Fla. 2004) (“While it may be true that having to drive to an early

- 55 -

1

voting site and having to wait in line may cause people to be inconvenienced,

2

inconvenience does not result in a denial of ‘meaningful access to the political process.’”

3

(quoting Osburn v. Cox, 369 F.2d 1283, 1289 (11th Cir. 2004)); Glover v. S.C.

4

Democratic Party, No. C/A 4-04-CV-2171-25, 2004 WL 3262756, at *6 (D.S.C. Sept. 3,

5

2004) (“[T]he Court does not find that difficulty voting equates with a ‘denial or

6

abridgment’ of the right to vote.”).

7

With these principles in mind, the Court first will apply step one of the two-part

8

vote denial framework to each of the challenged voting practices to determine whether

9

either disparately burdens minority voters. The Court then will discuss step two and the

10
11
12

Senate Factors.
A. Step One (Disparate Impact)
1. H.B. 2023

13

H.B. 2023 is facially neutral. It applies to all Arizonans regardless of race or

14

color. Plaintiffs nonetheless allege that H.B. 2023 disparately burdens Hispanic, Native

15

American, and African American voters as compared to non-minority voters because

16

these groups disproportionately rely on others to collect and return their early ballots.

17

But there are no records of the numbers of people who, in past elections, have relied on

18

now-prohibited third parties to collect and return their early mail ballots, and of this

19

unknown number Plaintiffs have provided no quantitative or statistical evidence

20

comparing the proportion that is minority versus non-minority.

21

This evidentiary hole presents a practical problem. Disparate impact analysis is a

22

comparative exercise. To determine whether a practice disparately impacts minorities,

23

the Court generally must know approximately: (1) how many people will be affected by

24

the practice, and (2) their racial composition. Without this information, it becomes

25

difficult to compare the law’s impact on different demographic populations and to

26

determine whether the disparities, if any, are meaningful. That is, it might be true that

27

minorities broadly have used ballot collection services more often than non-minorities,

28

but the discrepancy might be slight enough that it does not meaningfully deny minorities

- 56 -

1

an equal opportunity to participate in the political process and elect their preferred

2

representatives. See Frank, 768 F.3d at 754.

3

Indeed, the Court is aware of no vote denial case in which a § 2 violation has been

4

found without quantitative evidence measuring the alleged disparate impact of a

5

challenged law on minority voters. Rather, the standards developed for analyzing § 2

6

vote denial cases suggest that proof of a relevant statistical disparity might be necessary

7

at step one, even though it is not alone sufficient to prove a § 2 violation because of the

8

causation requirement at step two.16 See Gonzales, 677 F.3d at 405; Veasey, 830 F.3d at

9

244 (noting that “courts regularly utilize statistical analyses to discern whether a law has

10

a discriminatory impact”).

11

necessity of statistical evidence in disparate impact cases.” Budnick v. Town of Carefree,

12

518 F.3d 1109, 1118 (9th Cir. 2008) (Fair Housing Act); Pottenger v. Potlatch Corp., 329

13

F.3d 740, 749 (9th Cir. 2003) (Age Discrimination in Employment Act); Cooper v. S.

14

Co., 390 F.3d 695, 716 (11th Cir. 2004), overruled on other grounds by Ash v. Tyson

15

Foods, Inc., 546 U.S. 454 (2006) (Title VII); Rollins v. Alabama Cmty. Coll. Sys., No.

16

2:09-CV-636-WHA, 2010 WL 4269133, at *9 (M.D. Ala. Oct. 25, 2010) (Equal Pay

17

Act); Davis v. City of Panama City, Fla., 510 F. Supp. 2d 671, 689 (N.D. Fla. 2007)

18

(Title VII and 42 U.S.C. § 1983).

Further, in other contexts courts have “recognized the

19

The Court is not suggesting that quantitative evidence of a challenged voting

20

practice’s actual effect is needed in vote denial cases. As Chief Judge Thomas noted in

21

his dissent during the preliminary appellate phase of this case, “quantitative evidence of

22

the effect of a rule on voting behavior is only available after an election has occurred, at

23

which point the remedial purpose of the Voting Rights Act is no longer served.”

24
25
26
27
28

16

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.

- 57 -

1

Feldman, 840 F.3d at 1092 n.5 (Thomas, C.J. dissenting). But quantitative evidence of

2

the number of voters who used ballot collection before H.B. 2023’s enactment, together

3

with similar evidence of those voters’ demographics, would permit the Court to

4

reasonably infer how many voters would be affected by H.B. 2023’s limitations in future

5

elections, and whether those voters disproportionately would be minorities.17 As one

6

commentator has argued:

7

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
9
10
11
12
13

Daniel P. Tokaji, Applying Section 2 to the New Vote Denial, 50 Harv. C.R.-C.L. L. Rev.

14

439, 476 (2015).

15

The Court is mindful, however, that no court has explicitly required quantitative

16

evidence to prove a vote denial claim, and a majority of the en banc Ninth Circuit panel

17

reviewing the preliminary phase of this case appears to have rejected such a rule. The

18

Court therefore does not find against Plaintiffs on this basis. Rather, the Court finds that

19

Plaintiffs’ circumstantial and anecdotal evidence is insufficient to establish a cognizable

20

disparity under § 2.

21

To overcome the lack of quantification, Plaintiffs attempt to prove the existence of

22

a meaningful disparity through two general categories of circumstantial evidence. First,

23

lawmakers, elections officials, and community advocates testified that ballot collection

24

tends to be used more by communities that lack easy access to secure, outgoing mail

25

services; the elderly, homebound, and disabled; the poor; those who lack reliable

26
27
28

17

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.

- 58 -

1

transportation; those who work multiple jobs or lack childcare; and less educated voters

2

who are unfamiliar with or more intimidated by the voting process. In turn, data shows

3

that these socioeconomic circumstances are disproportionately reflected in minority

4

communities. (See Ex. 97 at 57; Tr. 59-60, 416-26, 432-39, 629-35, 895-900.) It stands

5

to reason, then, that prior to H.B. 2023’s enactment minorities generally were more likely

6

than non-minorities to give their early ballots to third parties.

7

For example, relative to non-minorities, Hispanics and African Americans are

8

nearly two times more likely to live in poverty, and the poverty rate for Native Americans

9

is over three times higher. (Ex. 93 at 15.) Wages and unemployment rates for Hispanics,

10

African Americans, and Native Americans consistently have exceeded non-minority

11

unemployment rates for the period of 2010 to 2015. (Ex. 91 at 40; Ex. 93 at 15.)

12

According to the 2015 American Community Survey 1-year estimates, unemployment

13

rates were 10.5 percent for African Americans, 7.7 percent for Hispanics, 16.8 percent for

14

Native Americans, and only 5.6 percent for non-minorities. In Arizona, 68.9 percent of

15

non-minorities own a home, whereas only 32.3 percent of African Americans, 49 percent

16

of Hispanics, and 56.1 percent of Native Americans do so. (Ex. 93 at 15, 17; Ex. 98 at

17

33.)

18

Non-minorities remain more likely than Hispanics, Native Americans, and African

19

Americans to graduate from high school, and are nearly three times more likely to have a

20

bachelor’s degree than Hispanics and Native Americans.

21

survey, over 22.4 percent of Hispanics and 11.2 percent of Native Americans rated

22

themselves as speaking English less than “very well,” as compared to only 1.2 percent of

23

non-minorities. (Ex. 93 at 16.) Due to their lower levels of literacy and education,

24

minority voters are more likely to be unaware of certain technical rules, such as the

25

requirement that early ballots be received by the county recorder, rather than merely

26

postmarked, by 7:00 p.m. on Election Day. (Ex. 91 at 38.)

27
28

Additionally, in a recent

As of 2015, Hispanics, Native Americans, and African Americans fared worse
than non-minorities on a number of key health indicators. (Ex. 93 at 18.)

- 59 -

Native

1

Americans in particular have much higher rates of disability than non-minorities, and

2

Arizona counties with large Native American populations have much higher rates of

3

residents with ambulatory disabilities. For example, “17 percent of Native Americans are

4

disabled in Apache County, 22 percent in Navajo County, and 30 percent in Coconino

5

County.” Further, “11 percent [of individuals] have ambulatory difficulties in Apache

6

County, 13 percent in Navajo County, and 12 percent in Coconino County, all of which

7

contain significant Native American populations and reservations.” (Ex. 97 at 60.)

8

Hispanics, Native Americans, and African Americans also are significantly less

9

likely than non-minorities to own a vehicle, more likely to rely upon public

10

transportation, more likely to have inflexible work schedules, and more likely to rely on

11

income from hourly wage jobs. (Ex. 93 at 12-18; Ex. 97 at 51-52; Tameron Dep. 155:5-

12

20; Pstross Dep. 34:11-22.)

13

nonexistent in some minority communities. (Ex. 97 at 57; Ex. 98 at 18; Tr. 506.)

Ready access to reliable and secure mail service is

14

These disparities exist in both urban and rural areas. For example, Representative

15

Charlene Fernandez described a lack of home mail service in rural San Luis, a city that is

16

98 percent Hispanic. Almost 13,000 residents rely on a post office located across a major

17

highway. With no mass transit, a median income of $22,000, and many people not

18

owning cars, receiving and sending mail in San Luis can be more difficult than in other

19

communities. (Tr. 40-46.) A surprising number of voters in the Hispanic community

20

also distrust returning their voted ballot via mail, particularly in low-income communities

21

where mail theft is common. Although a lack of outgoing mail presents a problem for

22

rural minority voters, unsecure mailboxes are an impediment for urban minorities who

23

distrust the mail service and prefer instead to give their ballots to a volunteer. (Tr. 98,

24

238-39, 896-97, 1170; Healy Dep. 97:18-24; Scharff Dep. 92:5-17.)

25

These problems are particularly acute in Arizona’s Native American communities,

26

in which vehicle ownership is significantly lower than non-minority Arizonans. (Ex. 91

27

at 42.)

28

reservations lack access to a vehicle. (Ex. 98 at 16.) Moreover, according to Dr. Rodden,

Between one quarter and one half of all households on Native American

- 60 -

1

“the extent to which rural Native Americans lack mail service is quite striking.” “[T]he

2

majority of Native Americans in non-metropolitan Arizona do not have residential mail

3

service.” “Only 18 percent of Native American registered voters have home mail

4

delivery. . . . The rate at which registered voters have home mail service is over 350

5

percent higher for non-Hispanic whites than for Native Americans.” As such, most

6

Native American registered voters must travel to a town to retrieve their mail, “[y]et rates

7

of vehicle access are quite low.” (Ex. 97 at 57.) On the Navajo Reservation, most people

8

live in remote communities, many communities have little to no vehicle access, and there

9

is no home incoming or outgoing mail, only post office boxes, sometimes shared by

10

multiple families. (Tr. 172-75, 297-98.)

11

There is no home delivery in the Tohono O’odham Nation, where there are 1,900

12

post office boxes and some cluster mail boxes. The postmaster for the Tohono O’odham

13

Nation anecdotally related to Representative Fernandez that she observes residents come

14

to the post office every two or three weeks to get their mail. Due to the lack of

15

transportation, the condition of the roads, and health issues, some go to post office only

16

once per month. (Tr. 52-58, 315-17.)

17

Thus, “for many Native Americans living in rural locations, especially on

18

reservations, voting is an activity that requires the active assistance of friends and

19

neighbors.” (Ex. 97 at 60.) LeNora Fulton—a member of the Navajo Nation, former

20

representative of the Fort Defiance Chapter of the Navajo National Council, member of

21

the Navajo Election Board of Supervisors, and the Apache County Recorder from 2004

22

through 2016, where her responsibilities included overseeing voter registration, early

23

voting and voter outreach—explained that people in the Navajo Nation trust non-family

24

members to deliver their early ballots because “[i]t’s part of the culture. . . . [T]here is a

25

clan system. They may not be related by blood, but they are related by clan. Everyone

26

on the Navajo Nation is related one way or another through the clan system.” Ballot

27

collection and delivery by those with the means to travel “was the standard practice with

28

the Apache County . . . but also with the Nation[.]” “We have many people that would

- 61 -

1

come into our office in St. Johns that help individuals that not are not able to get a ballot,

2

you know, to the office.

3

practice . . . It was a norm for us.” According to Fulton, limiting who may collect and

4

deliver early ballots “would be a huge devastation . . . . The laws are supposed to be

5

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

6

The second category of circumstantial evidence concerns those who tend to offer

7

ballot collection services. Within the last decade, ballot collection has become a larger

8

part of the Democratic Party’s GOTV strategy. The Democratic Party and community

9

advocacy organizations have focused their ballot collection efforts on low-efficacy

10

voters, who trend disproportionately minority. In turn, minorities in Arizona tend to vote

11

for Democratic candidates. (Ex. 93 at 4-6, 11; Tr. 92, 283, 309, 416-26, 632-33, 659,

12

902-03, 1143, 1191-96, 1200, 1407, 1770-71, 1843-44; Healy Dep. 28:15-29:13.)

13

Individuals who have collected ballots in past elections observed that minority voters,

14

especially Hispanics, were more interested in utilizing their services. Indeed, Helen

15

Purcell, who served as the Maricopa County Recorder for 28 years from 1988 to 2016,

16

observed that ballot collection was disproportionately used by Hispanic voters. (Tr. 417-

17

19, 635, 642, 866, 895-900, 931-32, 1039-40, 1071, 1170.)

18

In contrast, the Republican Party has not significantly engaged in ballot collection

19

as a GOTV strategy. The base of the Republican Party in Arizona trends non-minority.

20

On average, non-minorities in Arizona vote 59 percent for Republican candidates, as

21

compared with 35 percent of Hispanic voters. Individuals who have collected ballots in

22

past elections have observed that voters in predominately non-minority areas were not as

23

interested in ballot collection services. (Tr. 430-31, 898, 1170, 1192, 1408; Ex. 91 at 31.)

24

Based on this evidence, the Court finds that prior to H.B. 2023’s enactment

25

minorities generically were more likely than non-minorities to return their early ballots

26

with the assistance of third parties. The Court, however, cannot speak in more specific or

27

precise terms than “more” or “less.”

28

disparities between minorities and non-minorities in Arizona, these disparities are an

Although there are significant socioeconomic

- 62 -

1

imprecise proxy for disparities in ballot collection use. Plaintiffs do not argue that all or

2

even most socioeconomically disadvantaged voters use ballot collection services, nor

3

does the evidence support such a finding. Rather, the anecdotal estimates from individual

4

ballot collectors indicate that a relatively small number of voters have used ballot

5

collection services in past elections.

6

socioeconomically disadvantaged voters, most do not use ballot collection services to

7

vote. Considering the vast majority of Arizonans, minority and non-minority alike, vote

8

without the assistance of third-parties who would not fall within H.B. 2023’s exceptions,

9

it is unlikely that H.B. 2023’s limitations on who may collect an early ballot cause a

10

meaningful inequality in the electoral opportunities of minorities as compared to non-

11

minorities.

It reasonably follows, then, that even among

12

Moreover, H.B. 2023 does not impose burdens beyond those traditionally

13

associated with voting. Although, for some voters, ballot collection is a preferred and

14

more convenient method of voting, H.B. 2023 does not deny minority voters meaningful

15

access to the political process simply because the law makes it slightly more difficult or

16

inconvenient for a small, yet unquantified subset of voters to return their early ballots. In

17

fact, no individual voter testified that H.B. 2023’s limitations on who may collect an

18

early ballot would make it significantly more difficult to vote. The Court therefore finds

19

that Plaintiffs have not carried their burden at step one of the vote denial framework.

20

2. OOP Voting

21

Unlike their H.B. 2023 challenge, Plaintiffs provided quantitative and statistical

22

evidence of disparities in OOP voting through the expert testimony of Dr. Rodden.

23

Because Arizona does not track the racial demographics of its voters, Dr. Rodden used an

24

open-source software algorithm that he found online to predict each voter’s race. (Tr.

25

378-414, 520-24.) Specifically, Dr. Rodden placed the names of individuals who cast

26

ballots in particular elections into Census blocks and tracts, for which racial data is

27

available from the Census, and then combined that information with surname data to

28

estimate the race of each voter. This approach has become common in academic studies,

- 63 -

1

as well as VRA litigation. Dr. Rodden added further precision to his estimates by

2

conditioning them on not just surnames and neighborhood race statistics, but also on

3

additional information found in the voter file, such as the individual’s age and party.18

4

(Ex. 97 at 10-15.)

5

Dr. Rodden’s analysis is credible and shows that minorities are over-represented

6

among the small number of voters casting OOP ballots. For example, in Maricopa

7

County—which accounts for 61 percent of Arizona’s population—non-minority voters

8

accounted for only 56 percent of OOP ballots cast during the 2012 general election,

9

despite casting 70 percent of all in-person votes. In contrast, African American and

10

Hispanic voters made up 10 percent and 15 percent of in-person voters, but accounted for

11

13 percent and 26 percent of OOP ballots, respectively.

12

accounted for 1.1 percent of in-person voters, and 1.3 percent of OOP ballots. Dr.

13

Rodden observed similar results for Pima County. (Ex. 91 at 52; Ex. 95 at 31-33, 37, 43.)

14

Similarly, minority voters cast a disproportionate share of OOP ballots during the

15

2016 general election. In Maricopa County, estimated rates of OOP voting were twice as

16

high for Hispanics, 86 percent higher for African Americans, and 73 percent higher for

17

Native Americans than for their non-minority counterparts. In Pima County, rates of

18

OOP voting were 150 percent higher for Hispanics, 80 percent higher for African

19

Americans, and 74 percent higher for Native Americans than for non-minorities.

20

Moreover, in Pima County the overall rate of OOP voting was higher, and the racial

21

disparities larger, in 2016 than in 2014. Among all counties that reported OOP ballots in

22

the 2016 general election, a little over 1 in every 100 Hispanic voters, 1 in every 100

23
24
25
26
27
28

18

Native American voters

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
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

African-American voters, and 1 in every 100 Native American voters cast an OOP ballot.

2

For non-minority voters, the figure was around 1 in every 200 voters. Racial disparities

3

in OOP voting were found in all counties except La Paz County, which has a small

4

minority population. (Ex. 97 at 3, 19-21, 28-34.)

5

Although Dr. Rodden’s race estimation is credible, his analysis paints an

6

incomplete picture of the practical impact of OOP voting because the majority of

7

Arizonans successfully vote by mail and therefore are unaffected by precinct

8

requirements. For example, in the 2012 general election Arizona voters cast 10,979 OOP

9

ballots out of the 2,323,579 ballots cast statewide, accounting for 0.47 percent of all votes

10

cast. In Maricopa County, 1,390,836 total ballots were cast, of which 7,529 (or 0.54

11

percent) were rejected for being cast out of the voter’s assigned precinct. OOP ballots

12

cast by non-minority voters therefore accounted for only 0.3 percent of all votes cast in

13

Maricopa County during the 2012 election, whereas OOP ballots cast by Hispanic and

14

African American voters accounted only for approximately 0.14 percent and 0.07

15

percent, respectively. These figures dropped substantially in the 2016 general election,

16

during which only 3,970 Arizonans voted OOP out of the 2,661,497 ballots cast

17

statewide, representing only 0.15 percent of all votes cast.

18

1,608,875 total ballots were cast, of which only 2,197 (or 0.14 percent) were cast OOP.

19

(Tr. 1927-32; Exs. 578-79, 581.)

In Maricopa County,

20

Considering OOP ballots represent such a small and ever-decreasing fraction of

21

the overall votes cast in any given election, OOP ballot rejection has no meaningfully

22

disparate impact on the opportunities of minority voters to elect their preferred

23

representatives. To be clear, the Court is not suggesting that the votes of individuals who

24

show up at the wrong precinct are unimportant. But, as a practical matter, the disparity

25

between the proportion of minorities who vote at the wrong precinct and the proportion

26

of non-minorities who vote at the wrong precinct does not result in minorities having

27

unequal access to the political process. See Fed. Judicial Ctr., Reference Manual on

28

Scientific Evidence 252 (3d ed. 2011) (discussing difference between statistical

- 65 -

1

significance and practical significance). No state has exactly equal rates at every stage of

2

its voting system, and in the end the vast majority of all votes in Arizona—cast by

3

minority and non-minority voters alike—are counted. See Frank, 768 F.3d at 754.

4

Moreover, Arizona’s policy to not count OOP ballots is not the cause of the

5

disparities in OOP voting.

6

concentrated in relatively dense precincts that are disproportionately populated with

7

renters and those who move frequently. These groups, in turn, are disproportionately

8

composed of minorities. (Ex. 97 at 16-18.) Because minority voters in Arizona have

9

disproportionately higher rates of residential mobility, they are more likely to need to

10

renew their voter registration and reeducate themselves about their new voting locations.

11

(Ex. 91 at 39; Ex. 93 at 17; Ex. 95 at 4, 7-12; Ex. 98 at 33.)

Dr. Rodden’s analysis confirms that OOP voting is

12

Polling place locations present additional challenges for Native American voters.

13

For example, Navajo voters in Northern Apache County lack standard addresses, and

14

their precinct assignments for state and county elections are based upon guesswork,

15

leading to confusion about the voter’s correct polling place. Additionally, boundaries for

16

purposes of tribal elections and Apache County precincts are not the same. As a result, a

17

voter’s polling place for tribal elections often differs from the voter’s polling place for

18

state and county elections. Inadequate transportation access also can make travelling to

19

an assigned polling place difficult. (Ex. 97 at 51-54; Tr. 299-301.)

20

Plaintiffs, however, do not challenge the manner in which Arizona counties

21

allocate and assign polling places or Arizona’s requirement that voters re-register to vote

22

when they move. Plaintiffs also offered no evidence of a systemic or pervasive history of

23

minority voters being given misinformation regarding the locations of their assigned

24

precincts, while non-minority voters were given correct information. Nor have they

25

shown that precincts tend to be located in areas where it would be more difficult for

26

minority voters to find them, as compared to non-minority voters. To the contrary, there

27

are many ways for voters in Arizona to locate their assigned precincts, and state, county,

28

and local elections officials engage in substantial informational campaigns and voter

- 66 -

1

outreach. Plaintiffs, instead, have challenged what Arizona does with OOP ballots after

2

they have been cast, which does not cause the observed disparities in OOP voting.

3

In sum, Plaintiffs have not carried their burden at step one of the vote denial

4

framework for two independent reasons. First, they have not shown that Arizona’s policy

5

to not count OOP ballots causes minorities to show up to vote at the wrong precinct at

6

rates higher than their non-minority counterparts.

7

account for such a small fraction of votes cast statewide, Plaintiffs have not shown that

8

the racial disparities in OOP voting are practically significant enough to work a

9

meaningful inequality in the opportunities of minority voters as compared to non-

10

minority voters to participate in the political process and elect their preferred

11

representatives.

Second, given that OOP ballots

12

B. Step Two (Senate Factors)

13

Step two of the results test is informed by the Senate Factors and asks whether the

14

disparate burdens imposed by the challenged voting practices are in part caused by or

15

linked to social and historical conditions within the state that have or currently produce

16

discrimination against the affected minorities. The Court does not need to reach this step

17

because Plaintiffs have not shown at step one that the challenged voting practices impose

18

meaningfully disparate burdens on minority voters as compared to non-minority voters.

19

Cf. Ohio Democratic Party, 834 F.3d at 638 (“If this first element is met, the second step

20

comes into play, triggering consideration of the ‘totality of the circumstances,’ potentially

21

informed by the ‘Senate Factors’ discussed in ‘Gingles.’”). Nonetheless, to ensure that

22

the record is fully developed, the Court will address below the evidence pertinent to the

23

Senate Factors. The Court will not discuss factors three and four, however, because they

24

are not germane to the challenged voting practices and there is insufficient evidence to

25

warrant discussion.

26

1. Relevant History of Official Discrimination

27

Arizona has a history of discrimination against Native Americans, Hispanics, and

28

African Americans. Such discrimination began as early as 1912, when Arizona became a

- 67 -

1

state, and continued into the modern era. In 1975, Arizona’s history of discrimination

2

resulted in it becoming one of only nine states to be brought wholly under § 5 of the VRA

3

as a “covered jurisdiction.” In addition to being covered under § 5, it was one of only

4

three states to be covered under § 4(f)(4) of the Act for Spanish Heritage. (Ex. 89 at 5-

5

24; Ex. 91 at 2, 24-30; Ex. 521 at 43-45; Doc. 361 ¶ 42.)

6

When Arizona became a state in 1912, Native Americans were excluded from

7

voting. Even after Congress acknowledged that Native Americans were citizens in 1924,

8

thereby affording them the right to vote, Arizona’s Constitution continued to deny Native

9

Americans that right. (Ex. 89 at 17; Ex. 521 at 43.) It was not until 1948—24 years after

10

federal law allowed Native Americans to vote—that the Arizona Supreme Court found

11

the State’s disenfranchisement of Native Americans was unconstitutional and finally

12

granted Native Americans the right to vote. (Doc. 361 ¶ 17; Ex. 89 at 17; Ex. 521 at 45.);

13

Harrison v. Laveen, 196 P.2d 456, 463 (Ariz. 1948).

14

Despite this ruling, Native Americans, as well as Hispanics and African

15

Americans, continued to face barriers to participation in the franchise. For example, in

16

1912 Arizona enacted an English literacy test for voting.

17

specifically to limit “the ignorant Mexican vote,” but it also had the effect of reducing the

18

ability of African Americans and Native Americans to register and vote, as registrars

19

applied the test to these communities as well. Well into the 1960s, white Arizonans

20

challenged minority voters at the polls by asking them to read and explain literacy cards.

21

(Doc. 361 ¶ 14; Ex. 89 at 14-17; Ex. 521 at 44-45.)

The test was enacted

22

In 1970, Congress amended the VRA to enact a nationwide ban on literacy tests

23

after finding that they were used to discriminate against voters on account of their race or

24

ethnicity. In reaching that finding, Congress cited evidence that showed application of

25

the literacy test had significantly lowered the participation rates of minorities.

26

specifically found that in Arizona “only two counties out of eight with Spanish surname

27

populations in excess of 15% showed a voter registration equal to the state-wide

28

average.” It also noted that Arizona had a serious deficiency in Native American voter

- 68 -

It

1

registrations. Rather than comply with the VRA and repeal its literacy test, Arizona

2

challenged the ban, arguing that it could not be enforced to the extent that it was

3

inconsistent with the State’s literacy requirement. Even after the Supreme Court upheld

4

Congress’s ban, Arizona waited an additional two years to formally repeal its literacy

5

test. (Ex. 89 at 14-18; Ex. 91 at 24); see Or. v. Mitchell, 400 U.S. 112, 118 (1970).

6

The effects of Arizona’s literacy test were compounded by the State’s history of

7

discrimination in the education of its Hispanic, Native American, and African American

8

citizens. (Doc. 361 ¶¶ 2-4; Ex. 89 at 9-10; Ex. 91 at 5.) From 1912 until the Supreme

9

Court’s decision in Brown v. Board of Education, 347 U.S. 483 (1954), segregated

10

education was widespread throughout Arizona and sanctioned by both the courts and the

11

state legislature. (Ex. 521 at 35-39; Ex. 89 at 9-12); see also Dameron v. Bayless, 126 P.

12

273 (Ariz. 1912); Gonzales v. Sheely, 96 F. Supp. 1004, 1008-09 (D. Ariz. 1951)

13

(enjoining segregation of Mexican school children in Maricopa County). In fact, the

14

Tucson Public Schools only recently reached a consent decree with the DOJ over its

15

desegregation plan in 2013. (Ex. 91 at 27.) The practice of segregation also extended

16

beyond schools; it was common place to have segregated public spaces such as

17

restaurants, swimming pools, and theaters. (Ex. 89 at 15; Ex. 521 at 34.) Even where

18

schools were not segregated, Arizona enacted restrictions on bilingual education. As

19

recently as 2000, Arizona banned bilingual education with the passage of Proposition

20

203. (Ex. 89 at 20; Ex. 91 at 47.)

21

Arizona has a record of failing to provide adequate funding to teach its non-

22

English speaking students. This under-funding has taken place despite multiple court

23

orders instructing Arizona to develop an adequate funding formula for its programs,

24

including a 2005 order in which Arizona was held in contempt of court for refusing to

25

provide adequate funding for its educational programs. (Ex. 91 at 46-47); Flores v.

26

Arizona, 405 F. Supp. 2d 1112 (D. Ariz. 2005), vacated, 204 Fed. App’x 580 (9th Cir.

27

2006). “According to the Education Law Center’s latest National Report Card that

28

provided data for 2013, Arizona ranked 47th among the states in per-student funding for

- 69 -

1

elementary and secondary education.” (Ex. 91 at 47.)

2

Along with the State’s hostility to bilingual education, Maricopa County has

3

sometimes failed to send properly translated education materials to its Spanish speaking

4

residents, resulting in confusion and distrust from Hispanic voters. For example, in 2012,

5

Maricopa County misprinted the date of the election on over 2,000 Spanish language

6

information cards and bookmarks, some of which were distributed into the community.

7

(Ex. 89 at 22; Ex. 91 at 51; Healy Dep. 114:1-22.)

8

With that said, discrimination against minorities in Arizona has not been linear.

9

(Ex. 521 at 4.) For example, Arizona was subject to § 5 preclearance requirements until

10

2013. In Shelby, however, the Supreme Court found the formula used to determine

11

which states were subject to preclearance requirements unconstitutional because it was

12

“based on 40–year–old facts having no logical relation to the present day.” 570 U.S. at

13

553-54. Moreover, during the time that Arizona was under preclearance requirements

14

(1975-2013), the DOJ did not issue any objections to any of its statewide procedures for

15

registration or voting.

16

From 1982 to 2002, the DOJ objected to four of Arizona’s statewide redistricting

17

plans. Arizona acted to avoid the politics of racially discriminatory redistricting when, in

18

2000, the Arizona Independent Redistricting Commission (“AIRC”) was formed pursuant

19

to a voter initiative (Proposition 106). The AIRC is composed of two Republicans, two

20

Democrats, and an Independent, and it is tasked with redrawing of legislative and

21

congressional district lines following each decennial Census. According to its enacting

22

constitutional provisions, the AIRC considers the following six criteria when

23

redistricting: (a) equal population; (b) compactness and contiguousness; (c) compliance

24

with the Constitution and the VRA; (d) respect for communities of interest; (e)

25

incorporation of visible geographic features, including city, town and county boundaries,

26

as well as undivided census tracts; and (f) creation of competitive districts where there is

27

no significant detriment to other goals. (Doc. 361 ¶ 44.) The most recent AIRC set a

28

goal to pass preclearance with its first submittal to DOJ. The AIRC did this by ensuring

- 70 -

1

the competitiveness of legislative and congressional districts and ensuring that minorities

2

have the opportunity to elect candidates of their choice. See Harris v. Ariz. Indep.

3

Redistricting Comm’n, -- U.S. --, 136 S. Ct. 1301, 1308 (2016). The Commission

4

succeeded, and the DOJ approved Arizona’s new maps on April 9, 2012 without

5

objection.

6

In sum, “[d]iscriminatory action has been more pronounced in some periods of

7

state history than others . . . [and] each party (not just one party) has led the charge in

8

discriminating against minorities over the years.”

9

objectives are the motivating factor in decisions to take actions detrimental to the voting

10

rights of minorities. “[M]uch of the discrimination that has been evidenced may well

11

have in fact been the unintended consequence of a political culture that simply ignores

12

the needs of minorities.” (Ex. 90 at 8.) Arizona’s recent history is a mixed bag of

13

advancements and discriminatory actions.

14

Sometimes, however, partisan

2. Racially Polarized Voting

15

Arizona has a history of racially polarized voting, which continues today. (Ex. 91

16

at 30-33, 44-45). In the most recent redistricting cycle, experts for the AIRC found that

17

at least one congressional district and five legislative districts clearly exhibited racially

18

polarized voting. (Ex. 91 at 29-33.) Exit polls for the 2016 general election demonstrate

19

that voting between non-minorities and Hispanics continues to be polarized along racial

20

lines. (Ex. 91 at 29-33, 44-45; Ex. 92 at 12, 14; Ex. 94 at 4); see also Gonzalez, 677 F.3d

21

at 407.

22

3. Socioeconomic Effects of Discrimination

23

Racial disparities between minorities and non-minorities in socioeconomic

24

standing, income, employment, education, health, housing, transportation, criminal

25

justice, and electoral representation have persisted in Arizona. (Ex. 89 at 7-8, 12, 23; Ex.

26

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

27

16, 18, 33; Tameron Dep. 155:5-20; Pstross Dep. 34:11-22; Tr. 506.)

28

disparities in transportation, housing, and education are most pertinent to the specific

- 71 -

Of these,

1
2

burdens imposed by the challenged laws.
4. Racial Appeals in Political Campaigns

3

Arizona’s racially polarized voting has resulted in racial appeals in campaigns.

4

For example, when Raul Castro ran for governor in the 1970s, his opponents urged

5

support for the white candidate because “he looked like a governor.” In that same

6

election, a newspaper published a picture of Fidel Castro with a headline that read

7

“Running for governor of Arizona.”

8

Superintendent of Public Education, John Huppenthal “ran an advertisement in which the

9

announcer said that Huppenthal was ‘one of us.’ The announcer noted that Huppenthal

10

voted against bilingual education and ‘will stop La Raza.’” Similarly, when running for

11

governor in 2014, Maricopa County Attorney Andrew Thomas ran an ad describing

12

himself as “the only candidate who has stopped illegal immigration” while

13

“simultaneously show[ing] a Mexican flag with a red strikeout line through it

14

superimposed over the outline of Arizona.” (Ex. 91 at 44.)

15

(Ex. 89 at 19.)

In a 2010 bid for State

Moreover, racial appeals have been made in the specific context of legislative

16

efforts to limit ballot collection.

17

criminalize ballot collection, Republican sponsors and proponents expressed beliefs that

18

ballot collection fraud regularly was occurring but struggled with the lack of direct

19

evidence substantiating those beliefs. In 2014, the perceived “evidence” arrived in the

20

form of a racially charged video created by Maricopa County Republican Chair A.J.

21

LaFaro (the “LaFaro Video”) and posted on a blog. (Ex. 121.) The LaFaro Video

22

showed surveillance footage of a man of apparent Hispanic heritage appearing to deliver

23

early ballots. It also contained a narration of “Innuendos of illegality . . . [and] racially

24

tinged and inaccurate commentary by . . . LaFaro.” (Ex. 91 at 18 n.40; Ex. 524 at 23-

25

24.) LaFaro’s commentary included statements that the man was acting to stuff the ballot

26

box; that LaFaro did not know if the person was an illegal alien, a dreamer, or citizen, but

27

knew that he was a thug; and that LaFaro did not follow him out to the parking lot to take

28

down his tag number because he feared for his life. The LaFaro Video goes on to tell

During the legislative hearings on earlier bills to

- 72 -

1

about ballot parties where people gather en mass and give their un-voted ballots to

2

operatives of organizations so they can not only collect them, but also vote them illegally.

3

(Ex. 91 at 18; Ex. 121.)

4

The LaFaro Video did not show any obviously illegal activity and there is no

5

evidence that the allegations in the narration were true. Nonetheless, it “became quite

6

prominent in the debates over H.B. 2023.” (Tr. 1154.) The LaFaro video also was

7

posted on Facebook and YouTube, shown at Republican district meetings, and was

8

incorporated into a television advertisement—entitled “Do You Need Evidence

9

Terry?”—for Secretary Reagan when she ran for Secretary of State. (Ex. 91 at 18; Ex.

10

107.) In the ad, the LaFaro Video plays after a clip of then-Arizona Attorney General

11

Terry Goddard stating he would like to see evidence that there has been ballot collection

12

fraud. While the video is playing, Secretary Reagan’s narration indicates that the LaFaro

13

Video answers Goddard’s request for evidence of fraud. The LaFaro Video, however,

14

merely shows a man of apparent Hispanic heritage dropping off ballots and not obviously

15

violating any law.19 (Ex. 107.)

16
17

5. Minority Representation in Public Offices
Notwithstanding racially polarized voting and racial appeals, the disparity in the

18

number of minority elected officials in Arizona has declined.

19

recognized for improvements in the number of Hispanics and Native Americans

20

registering and voting, as well as in the overall representation of minority elected officials

21

in the State. (Ex. 521 at 27-28.) “Nonwhites make up 25 percent of Arizona’s elected

22

office holders, compared to 44 percent of the total population. This gives [Arizona] the

23

16th best representation ratio in the country.” (Ex. 524 at 44.)

Arizona has been

24

Nevertheless, Arizona has seen only one Hispanic and one African American

25

elected to statewide office, and Arizona has never elected a Native American to statewide

26
27
28

19

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

office. No Native American or African American has been elected to represent Arizona

2

in the United States House of Representatives. Further, no Hispanic, Native American,

3

or African American has ever served as a United States Senator representing Arizona or

4

as Arizona Attorney General. (Ex. 91 at 45; Ex. 93 at 19-20; Ex. 89 at 19, 22.)

5

6. Lack of Responsiveness to Minority Needs

6

Plaintiffs’ evidence on this factor, presented through the analysis and opinions of

7

Dr. Lichtman, is insufficient to establish a lack of responsiveness on the part of elected

8

officials to particularized needs of minority groups. Dr. Lichtman ignored various topics

9

that are relevant to whether elected officials have shown responsiveness, and he did not

10

conduct research on the issues in Arizona when considering this factor.

11

Notably, the CCEC engages in outreach to various communities, including the

12

Hispanic and Native American communities, to increase voter participation. The CCEC

13

develops an annual voter education plan in consultation with elections officials and

14

stakeholders, and the current Chairman of the CCEC is Steve Titla, an enrolled member

15

of the San Carlos Apache Tribe, who has been particularly vocal in supporting CCEC

16

outreach to Native Americans.

17

7. Justifications for Challenged Provisions

18

Precinct-based voting helps Arizona counties estimate the number of voters who

19

may be expected at any particular precinct, allows for better allocation of resources and

20

personnel, improves orderly administration of elections, and reduces wait times. The

21

precinct-based system also ensures that each voter receives a ballot reflecting only the

22

races for which that person is entitled to vote, thereby promoting voting for local

23

candidates and issues and making ballots less confusing. Arizona’s policy to not count

24

OOP ballots is one mechanism by which it strictly enforces this system to ensure that

25

precinct-based counties maximize the system’s benefits. This justification is not tenuous.

26

As for H.B. 2023, there is no direct evidence that the type of ballot collection

27

fraud the law is intended to prevent or deter has occurred. Although the justifications for

28

H.B. 2023 are weaker than the justifications for the State’s OOP ballot policy, Arizona

- 74 -

1

nonetheless has a constitutionally adequate justification for the law: to reduce

2

opportunities for early ballot loss or destruction.

3

8. Overall Assessment

4

In sum, of the germane Senate Factors, the Court finds that some are present in

5

Arizona and others are not. Plaintiffs have shown that past discrimination in Arizona has

6

had lingering effects on the socioeconomic status of racial minorities. But Plaintiffs’

7

causation theory is too tenuous to support their VRA claim because, taken to its logical

8

conclusion, virtually any aspect of a state’s election regime would be suspect as nearly all

9

costs of voting fall heavier on socioeconomically disadvantaged voters. Such a loose

10

approach to causation, which potentially would sweep away any aspect of a state’s

11

election regime in which there is not perfect racial parity, is inconsistent with the Ninth

12

Circuit’s repeated emphasis on the importance of a “causal connection between the

13

challenged voting practice and a prohibited discriminatory result.” Salt River Project,

14

109 F.3d at 595. For these reasons, the Court concludes that Plaintiffs have not carried

15

their burden at either step of the § 2 results test.

16

VII. FIFTEENTH AMENDMENT/§ 2 (INTENTIONAL DISCRIMINATION)

17

Lastly, Plaintiffs contend that H.B. 2023 violates § 2 and the Fifteenth

18

Amendment because it was enacted with the intent to suppress minority votes. The

19

Fifteenth Amendment provides that “[t]he right of citizens of the United States to vote

20

shall not be denied or abridged by the United States or by any State on account of race,

21

color, or previous condition of servitude,” and authorizes Congress to enforce this

22

mandate “by appropriate legislation.” Section 2 is such legislation. Although Congress

23

amended the VRA in 1982 to add the results test, § 2 continues to prohibit intentional

24

discrimination in a manner coextensive with the Fifteenth Amendment. Consequently,

25

the standards for both the statutory and the constitutional claim overlap.

26

The parties agree that the standard for finding unconstitutional, intentional racial

27

discrimination is governed by the Supreme Court’s decision in Village of Arlington

28

Heights v. Metropolitan Housing Development Corp., 429 U.S. 252 (1977). There, the

- 75 -

1

Supreme Court explained that “official action will not be held unconstitutional solely

2

because it results in a racially disproportionate impact.” Id. at 264-65. Rather, “[p]roof

3

of racially discriminatory intent or purpose is required to show a violation” of the

4

Constitution. Id. at 265.

5

Discriminatory purpose must be “a motivating factor in the decision,” but it need

6

not be the only factor. Id. at 265-66. “Determining whether invidious discriminatory

7

purpose was a motivating factor demands a sensitive inquiry into such circumstantial and

8

direct evidence of intent as may be available.” Id. at 266. “[A]n invidious discriminatory

9

purpose may often be inferred from the totality of the relevant facts, including the fact, if

10

it is true, that the law bears more heavily on one race than another.” Wash. v. Davis, 426

11

U.S. 229, 242 (1976). “But the ultimate question remains: did the legislature enact a law

12

‘because of,’ and not just ‘in spite of,’ its discriminatory effect.” N.C. St. Conf. of

13

NAACP v. McCrory, 831 F.3d 204, 200 (4th Cir. 2016) (quoting Pers. Adm’r of Mass. v.

14

Fenney, 442 U.S. 256, 279 (1979)).

15

To guide this inquiry, the Arlington Heights Court articulated a non-exhaustive list

16

of factors courts should consider. These so-called “Arlington Heights Factors” include:

17

(1) the historical background and sequence of events leading to enactment; (2)

18

substantive or procedural departures from the normal legislative process; (3) relevant

19

legislative history; and (4) whether the law has a disparate impact on a particular racial

20

group. Arlington Heights, 429 U.S. at 266-68. If “racial discrimination is shown to have

21

been a ‘substantial’ or ‘motivating’ factor behind enactment of the law, the burden shifts

22

to the law’s defenders to demonstrate that the law would have been enacted without this

23

factor.” Hunter v. Underwood, 471 U.S. 222, 228 (1985). This same framework applies

24

to § 2 claims based on allegations of discriminatory purpose. See Garza v. Cty. of L.A.,

25

918 F.2d 763, 766 (9th Cir. 1990).

26

Having considered these factors, the Court finds that H.B. 2023 was not enacted

27

with a racially discriminatory purpose.

28

proponents of limitations on ballot collection harbored partisan motives—perhaps

Though some individual legislators and

- 76 -

1

implicitly informed by racial biases about the propensity of GOTV volunteers in minority

2

communities to engage in nefarious activities—the legislature as a whole enacted H.B.

3

2023 in spite of opponents’ concerns about its potential effect on GOTV efforts in

4

minority communities, not because of that effect. Despite the lack of direct evidence

5

supporting their concerns, the majority of H.B. 2023’s proponents were sincere in their

6

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

7

was a necessary prophylactic measure to bring early mail ballot security in line with in-

8

person voting.

9

Beginning with the historical background, H.B. 2023 emerged in the context of

10

racially polarized voting, increased use of ballot collection as a Democratic GOTV

11

strategy in low-efficacy minority communities, and on the heels of several prior efforts to

12

restrict ballot collection, some of which were spearheaded by former Arizona State

13

Senator Don Shooter.20 Due to the high degree of racial polarization in his district,

14

Shooter was in part motivated by a desire to eliminate what had become an effective

15

Democratic GOTV strategy. (Tr. 1061-63, 1200, 1687-88, 2158-62; Ex. 89 at 24; Ex. 91

16

at 52-55; Ex. 92 at 2-10; Ex. 93 at 2; Shooter Dep. at 117:5-16.) Indeed, Shooter’s 2010

17

election was close: he won with 53 percent of the total vote, receiving 83 percent of the

18

non-minority vote but only 20 percent of the Hispanic vote. (Ex. 94 at 4.)

19

Shooter’s efforts to limit ballot collection were marked by unfounded and often

20

farfetched allegations of ballot collection fraud. (Tr. 1064, 2162, 2194, 2205; Ex. 3 at 7-

21

8; Ex. 10 at 3-9; Ex. 25 at 22-23; Ex. 91 at 19-20; Ex. 123.) Though his allegations were

22

demonstrably false, they—along with the racially-tinged LaFaro Video—spurred a larger

23

debate in the legislature about the security of early mail voting as compared to in-person

24

voting.21 (Tr. 1644, 1687, 2158-59, 2161-62; Ex. 10 at 49-53; Ex. 17 at 15-16; Ex. 23 at

25
26
27
28

20

Shooter most recently was a member of the Arizona House of Representatives
but served as a state senator during the relevant time period.
21
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

83-84.)

2

Turning to the relevant legislative history, proponents of H.B. 2023 repeatedly

3

voiced concerns that mail-in ballots were less secure than in-person voting, and that

4

ballot collection created opportunities for fraud. Although no direct evidence of ballot

5

collection fraud was presented to the legislature or at trial, Shooter’s allegations and the

6

LaFaro Video were successful in convincing H.B. 2023’s proponents that ballot

7

collection presented opportunities for fraud that did not exist for in-person voting, and

8

these proponents appear to have been sincere in their beliefs that this was a potential

9

problem that needed to be addressed. (Ex. 17 at 11-13, 17-75, 83-84; Ex. 19 at 56-57;

10

Ex. 21 at 11; Ex. 23 at 36; Tr. 1450, 1805, 1822-23.) Notably, H.B. 2023 found support

11

among some minority officials and organizations.

12

supported by the Arizona Latino Republican Association for the Tucson Chapter, which

13

expressed concerns that elderly people in the Latino community were being taken

14

advantage of by ballot collectors. (Ex. 17 at 71-75.) Likewise, Michael Johnson, an

15

African American who had served on the Phoenix City Council, strongly favored H.B

16

2023 and expressed concern about stories of ballot collectors misrepresenting themselves

17

as election workers. (Ex. 17 at 45-50.) Further, although some Democratic lawmakers

18

accused their Republican counterparts of harboring partisan or racially discriminatory

19

motives, this view was not shared by all of H.B. 2023’s opponents. (Tr. 697.) For

20

example, Representative Fernandez testified that she has no reason to believe H.B. 2023

21

was enacted with the intent to suppress Hispanic voting. (Tr. 83.)

For example, the measure was

22

As for departures from the normal legislative process, Plaintiffs cite two prior

23

efforts to limit ballot collection as examples of procedural discrepancies. First, in 2011

24

Arizona enacted S.B. 1412, which required any person who delivered more than ten early

25

ballots to provide a copy of her photo identification to the receiving elections official. If

26

a ballot collector could not produce a copy of her photo identification, the elections

27

official was directed to record the information from whatever identification that the ballot

28

collector had available. Within 60 days of each election, the Secretary of State was to

- 78 -

1

compile a public statewide report listing the identities and personal information of all

2

ballot collectors. (Ex. 2 at 16-19; Ex. 91 at 6-7.)

3

When S.B. 1412 became law, Arizona still was subject to § 5 preclearance.

4

Accordingly, S.B. 1412 could not go into effect until the law had been precleared by the

5

DOJ or a federal court.

6

preclearance on April 26, 2011, and on June 27, 2011 the DOJ precleared all provisions

7

except for the provision regulating ballot collection. (Ex. 41; Ex. 91 at 6-7.) As to that

8

provision, the DOJ stated that “the information sent [wa]s insufficient to enable us to

9

determine that the proposed changes have neither the purpose nor will have the effect of

10

denying or abridging the right to vote on account of race, color, or membership in a

11

language minority group.” The DOJ asked for more information and stated that “if no

12

response is received within sixty days of this request, the Attorney General may object to

13

the proposed changes.” (Ex. 41.)

14

information, the Attorney General chose to voluntarily withdraw the ballot collection

15

provision on July 28, 2011, rendering the law unenforceable. (Ex. 91 at 6-7; Ex. 42.)

16

“Of 773 preclearance submissions this was one of only 6 that were fully or partially

17

withdrawn in Arizona.”

18

thereafter. (Ex. 5.)

The Arizona Attorney General submitted the law for

Rather than respond to the DOJ’s request for more

(Ex. 91 at 7.)

Arizona formally repealed the law shortly

19

Second, Republican legislators again tried to restrict ballot collection in 2013 with

20

the enactment of H.B. 2305, which banned partisan ballot collection and required other

21

ballot collectors to complete an affidavit stating that they had returned the ballot.

22

Violation of the law was a misdemeanor. H.B. 2305 was passed along nearly straight

23

party lines in the waning hours of the legislative session. (Ex. 7; Ex. 91 at 7-10.) Shortly

24

after its enactment, citizen groups organized a referendum effort and collected more than

25

140,000 signatures to place H.B. 2305 on the ballot for a straight up-or-down vote. (Tr.

26

1071-72; Ex. 91 at 11.) Had H.B. 2305 been repealed by referendum, the legislature

27

could not have enacted related legislation except on a supermajority vote, and only to

28

“further[] the purposes” of the referendum. Ariz. Const. art. 4, pt. 1, § 1(6)(C), (14).

- 79 -

1

Rather than face a referendum, Republican legislators again repealed their own

2

legislation along party lines. The bill’s primary sponsor, Secretary Reagan (who, at the

3

time, was a State Senator), admitted that the legislature’s goal was to break the bill into

4

smaller pieces and reintroduce individual provisions “a la carte.” (Ex. 91 at 11.)

5

Although the circumstances surrounding these prior bills are somewhat suspicious,

6

these departures have less probative value because they involve different bills passed

7

during different legislative sessions by a substantially different composition of

8

legislators. See Burton v. City of Belle Glade, 178 F.3d 1175, 1195 (11th Cir. 1999)

9

(“[W]e fail to see how evidence of . . . a [city’s] prior refusal to annex [a housing

10

project] standing alone establishes any intent, let alone a discriminatory one” for later

11

annexation decision); Kansas City, Mo. v. Fed. Pac. Elec. Co., 310 F.2d 271, 278 (8th

12

Cir. 1962) (noting the “questionable import that the rejection of prior bills may have in

13

determining congressional intent as to subsequently enacted legislation”).

14

Plaintiffs also claim that H.B. 2023 represents a substantive deviation from normal

15

legislative processes because it differs from these prior bills. But the fact that different

16

bills from different sponsors and different legislative sessions did not have the same

17

substance is not alone surprising, nor is it particularly probative of discriminatory intent.

18

Moreover, although Plaintiffs argue that the legislature made H.B. 2023 harsher than

19

previous ballot collection bills by imposing felony penalties, they ignore that H.B. 2023

20

in other respects is more lenient than its predecessors given its broad exceptions for

21

family members, household members, and caregivers.

22

Finally, Plaintiffs highlight the law’s impact on minority voters. As previously

23

noted, ballot collection was used as a GOTV strategy in mostly low-efficacy minority

24

communities, though the Court cannot say how often voters used ballot collection, nor

25

can it measure the degree or significance of any disparities in its usage. The legislature

26

was aware that the law could impact GOTV efforts in low-efficacy minority

27

communities; numerous democratic lawmakers speaking in opposition to the bill

28

expressed concerns that it would adversely impact minority GOTV efforts. (Ex. 17 at 74;

- 80 -

1

Ex. 19 at 17-18, 20, 35-37; Ex. 23 at 89-91; Ex. 25 at 27-28.) But this evidence shows

2

only that the legislature enacted H.B. 2023 in spite of its impact on minority GOTV

3

efforts, not because of that impact. Indeed, proponents of the bill seemed to view these

4

concerns as less significant because of the minimal burdens associated with returning a

5

mail ballot. (See, e.g., Ex. 23 at 81-82.)

6

Based on the totality of the circumstances, Plaintiffs have not shown that the

7

legislature enacted H.B. 2023 with the intent to suppress minority votes. Rather, some

8

individual legislators and proponents were motivated in part by partisan interests.

9

Shooter, for example, first raised concerns about ballot collection after winning a close

10

election. In addition to raising concerns about ballot collectors impersonating election

11

workers, Johnson complained that ballot collection put candidates “who don’t have

12

accessibility to large groups to go out and collect those ballots” at a disadvantage.

13

Likewise, Richard Hopkins, a proponent of the bill and a 2014 Republican candidate for

14

the Arizona House of Representatives, claimed that he lost his election because of “ballot

15

harvesting.” (Ex. 17 at 17, 45-49.) In opposing ballot collection restrictions, Democratic

16

Senator Steve Farley stated “[t]he problem we’re solving is that one party is better at

17

collecting ballots than the other one.” (Ex. 25 at 35.)

18

But partisan motives are not necessarily racial in nature, even though racially

19

polarized voting can sometimes blur the lines.

20

Amendment and § 2 of the VRA—upon which Plaintiffs’ intentional discrimination

21

claims are based—address racial discrimination, not partisan discrimination. That some

22

legislators and proponents harbored partisan interests, rather than racially discriminatory

23

motives, is consistent with Arizona’s history of advancing partisan objectives with the

24

unintended consequence of ignoring minority interests. (Ex. 90 at 8.)

Importantly, both the Fifteenth

25

Moreover, partisan motives did not permeate the entire legislative process.

26

Instead, many proponents acted to advance facially important interests in bringing early

27

mail ballot security in line with in-person voting security, notwithstanding the lack of

28

direct evidence that ballot collection fraud was occurring.

- 81 -

Though Plaintiffs might

1

disagree with the manner in which the legislature chose to address its concerns about

2

early ballot security, “the propriety of doing so is perfectly clear,” and the legislature

3

need not wait until a problem occurs to take proactive steps it deems appropriate.

4

Crawford, 553 U.S. at 196; see also Lee, 188 F. Supp. 3d at 609.

5

The Court therefore finds that the legislature that enacted H.B. 2023 was not

6

motivated by a desire to suppress minority voters. The legislature was motivated by a

7

misinformed belief that ballot collection fraud was occurring, but a sincere belief that

8

mail-in ballots lacked adequate prophylactic safeguards as compared to in-person voting.

9

Some proponents also harbored partisan motives. But, in the end, the legislature acted in

10

spite of opponents’ concerns that the law would prohibit an effective GOTV strategy in

11

low-efficacy minority communities, not because it intended to suppress those votes.

12

VIII. CONCLUSION

13

Plaintiffs have not carried their burden to show that the challenged election

14

practices severely and unjustifiably burden voting and associational rights, disparately

15

impact minority voters such that they have less opportunity than their non-minority

16

counterparts to meaningfully participate in the political process, or that Arizona was

17

motivated by a desire to suppress minority turnout when it placed limits on who may

18

collect early mail ballots. Plaintiffs have raised fair concerns about the wisdom of H.B.

19

2023 and Arizona’s treatment of OOP ballots as matters of public policy. The Court,

20

however, is not charged with second-guessing the prudence of Arizona’s laws. The

21

Court’s authority extends only to determining whether, in exercising its constitutional

22

authority to regulate the times, places, and manner of elections, Arizona has acted within

23

permissible constitutional and statutory bounds. In exercising this duty, the Court also is

24

constrained by decisions of the Supreme Court, including those standing for the

25

proposition that legislatures may act prophylactically rather than upon specific evidence

26

of a documented problem, and those finding that prevention of voter fraud and

27

preservation of public confidence in election integrity are important state interests. See

28

Purcell, 549 U.S. at 4; Crawford, 553 U.S. at 195; Munro, 479 U.S. at 194-95; Eu, 489

- 82 -

1

U.S. at 231. Based on a careful review of the evidence and governing case law, the Court

2

concludes that the challenged provisions contravene neither the Constitution nor the

3

VRA. Therefore,

4

IT IS ORDERED as follows:

5

1.

6

Defendants’ oral motion for judgment on partial findings (Doc. 384) is

DENIED as moot.

7

2. The Court finds in favor of Defendants and against Plaintiffs on all claims.

8

3. The Clerk of the Court shall enter judgment accordingly and terminate this

9
10

case.
Dated this 10th day of May, 2018.

11
12
13
14
15

Douglas L. Rayes
United States District Judge

16
17
18
19
20
21
22
23
24
25
26
27
28

- 83 -