Leslie Feldman, et al v. Arizona Sec'y of State's Ofc., et al

Filing

FILED OPINION (SIDNEY R. THOMAS, CARLOS T. BEA and SANDRA S. IKUTA) AFFIRMED. Judge: SRT Dissenting, Judge: CTB Authoring, FILED AND ENTERED JUDGMENT. [10184244]

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 1 of 30

FOR PUBLICATION

FILED

UNITED STATES COURT OF APPEALS

NOV 02 2016

FOR THE NINTH CIRCUIT

LESLIE FELDMAN; LUZ
MAGALLANES; MERCEDEZ HYMES;
JULIO MORERA; CLEO OVALLE;
PETERSON ZAH, Former Chairman and
First President of the Navajo Nation; THE
DEMOCRATIC NATIONAL
COMMITTEE; DSCC, AKA Democratic
Senatorial Campaign Committee; THE
ARIZONA DEMOCRATIC PARTY;
KIRKPATRICK FOR U.S. SENATE;
HILLARY FOR AMERICA,
Plaintiffs-Appellants,
BERNIE 2016, INC.,
Intervenor-PlaintiffAppellant,
v.
ARIZONA SECRETARY OF STATE’S
OFFICE; MICHELE REAGAN, in her
official capacity as Secretary of State of
Arizona; MARICOPA COUNTY BOARD
OF SUPERVISORS; DENNY BARNEY;
STEVE CHUCRI; ANDY KUNASEK;
CLINT HICKMAN; STEVE
GALLARDO, member of the Maricopa
County Board of Supervisors, in their
official capacities; MARICOPA COUNTY

No.

MOLLY C. DWYER, CLERK
U.S. COURT OF APPEALS

16-16865

D.C. No. 2:16-cv-01065-DLR

OPINION

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 2 of 30

RECORDER AND ELECTIONS
DEPARTMENT; HELEN PURCELL, in
her official capacity as Maricopa County
Recorder; KAREN OSBORNE, in her
official capacity as Maricopa County
Elections Director; MARK BRNOVICH,
in his official capacity as Arizona Attorney
General,
Defendants-Appellees,
THE ARIZONA REPUBLICAN PARTY;
DEBBIE LESKO; TONY RIVERO; BILL
GATES; SUZANNE KLAPP
Intervenor-DefendantsAppellees.

Appeal from the United States District Court
for the District of Arizona
Douglas L. Rayes, District Judge, Presiding
Argued and Submitted October 26, 2016
Pasadena, California
Before: Sidney R. Thomas, Chief Judge, and Carlos T. Bea and Sandra S. Ikuta,
Circuit Judges.
Opinion by Judge Bea, Circuit Judge:
In the 1970s1 Arizona enacted a statute which required each voter who votes
in person to cast his or her ballot at the precinct polling station at which the voter

1

Ariz. Rev. Stat. §§ 16-122, 16-135, 16-584 (codified in 1979).

2

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 3 of 30

was registered to vote (the “precinct vote rule”). Since then Arizona has amended
its statutes to adopt voting by mail, so long as the vote is received by election
officials by election day. Arizona has also enacted early in-person voting during
the 27 days preceding election day at designated early voting locations. Further,
on election day, a voter who has received a ballot through the mail may deposit
that ballot at any precinct in the county. But, if one is voting in the traditional inperson manner on election day, the precinct vote rule applies: for the vote to be
valid, one must vote at the assigned polling place or vote center. A vote cast
elsewhere will not be counted.
Feldman and other Appellants2 here challenged the precinct vote rule on the
grounds that it violated the federal Voting Rights Act of 1965 (“VRA”), 52 U.S.C.
§ 10301, and unjustifiably burdened their election rights guaranteed by the
Fourteenth Amendment of the federal Constitution. The district court denied
Feldman’s motion for a preliminary injunction. Feldman brought an emergency

2

The appellants here (plaintiffs below) are Leslie Feldman, Luz Magallanes, Mercedez Hymes, Julio

Morera, and Cleo Ovalle, registered Democratic voters in Maricopa County, Arizona; Peterson Zah, former
Chairman and First President of the Navajo Nation and registered voter in Apache County, Arizona; the Democratic
National Committee; the DSCC, aka Democratic Senatorial Campaign Committee; the Arizona Democratic Party; a
committee supporting the election of Democratic United States Representative Ann Kirkpatrick to U.S. Senate; and
Hillary for America, a committee supporting the election of Hillary Clinton as President of the United States. The
intervenor-plaintiff/appellant is Bernie 2016, Inc., a committee supporting the election of Bernie Sanders as
President of the United States. For convenience, we refer to the appellants as “Feldman.”

3

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 4 of 30

appeal before us.
Feldman’s VRA claim is that the precinct vote rule imposes a discriminatory
burden upon Hispanic, African-American and Native American citizens of Arizona
(“minority voters”) because it affords them less opportunity than have other
members of the electorate to participate in the electoral process and to elect
representatives of their choice. To prove her claim, Feldman proffered expert
witness opinion evidence, some of which was accepted by the district court. Expert
evidence, which the district court found relevant, showed that the share of minority
in-person voters who failed to vote in their precincts exceeded their proportional
representation in the electorate.
We find that the precinct vote rule, as administered by Arizona, probably
does not impermissibly burden minority voters by giving them less opportunity
than non-minorities to participate in the political process. But even assuming,
without deciding, that it imposes a cognizable burden on minority voters, Feldman
has not shown that Arizona’s enactment of the precinct vote rule is linked to social
and historical conditions that have or currently produce racial discrimination
against minority voters. Thus, we find that the district court correctly denied relief
for the claimed violation of the VRA.
Similarly, the district court correctly found that the constitutional violation
4

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 5 of 30

claims failed because the precinct vote rule, when considered together with other
options available to Arizona voters, imposes only a minimal burden upon minority
and majority voters. Such a minimal burden is sufficiently justified by Arizona’s
interests in effective administration of voting in the State.
We affirm.
I.
If an Arizona voter arrives at a polling place on election day to vote, but his
or her name does not appear on the voting register, he or she may still vote, but
only through a provisional ballot. Ariz. Rev. Stat. §§ 16-122, 16-135, 16-584.
This scenario may occur either because the voter recently moved or due to
inaccuracies in the official records. Later, the state reviews all provisional ballots
and counts those votes cast by voters confirmed to be eligible to vote. Id. Arizona
will not count a provisional ballot cast out of the voter’s correct precinct (known as
an “out-of-precinct” or “OOP ballot”). Id. Widely-used early vote by mail
alternatives permit voters to receive ballots by mail several weeks before an
election and cast these ballots through the mail without paying postage or by
dropping them at any polling place in their county on election day. A.R.S. §§ 16542, 16-548. Arizona recently has permitted counties to choose between the

5

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 6 of 30

traditional precinct model and “vote centers,” wherein voters from multiple
precincts can vote at a single location.3 A.R.S. § 16-411.
As noted, Arizona’s precinct vote rule has existed since the 1970s. In the
2012 elections, Arizona election officials determined that 10,979 ballots were cast
OOP and thus not counted, which constituted 0.5% of total ballots cast in the state.4
Feldman submitted an expert report by Dr. Jonathan Rodden, credited by the
district court for the purposes of her motion, which concluded that minorities were
over-represented amongst those who cast OOP ballots in certain Arizona
population centers. Other portions of the factual record are discussed as they
become relevant.

3

In 2011, Arizona revised its election law to permit counties to choose between using “vote centers” or

precincts. A.R.S. § 16-411. The “vote center” approach permits voters from diverse voting precincts in a county to
receive a ballot tailored to include races for which they are eligible and to cast it at a single location. Id. The
“precinct” approach restricts a voter to receiving and casting his in-person ballot at his precinct’s designated polling
place. Under either approach, the ballot received by a voter is tailored to include only those candidates or issues for
which the voter is entitled to vote based on the voter’s claimed place of residency. The ballot will not be counted if
the voter is found to be ineligible.
4

In the 2008 election 14,885 OOP ballots were not counted, constituting 0.6% of total ballots cast. Smaller
numbers of OOP ballots were rejected in the 2010 (0.3% of total ballots cast) and 2014 (0.2% of total ballots cast)
elections.

6

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 7 of 30

In April 2016, Feldman sued Arizona5 challenging its policy of rejecting
OOP ballots. Feldman argued that Arizona’s rejection of OOP ballots pursuant to
the precinct vote rule violates § 2 of the VRA by disparately burdening the
electoral opportunities of Hispanic, Native American, and African American
voters. She also argued the precinct vote rule violates the Fourteenth Amendment
by improperly burdening voting rights and by raising equal protection concerns. In
June 2016, Feldman moved for a preliminary injunction to require Arizona to
count those portions of OOP ballots for which the voter is eligible to vote.6
After full briefing, on October 11, 2016, the district court issued an order
denying the motion for a preliminary injunction because it found Feldman was
unlikely to succeed on the merits of her claims or suffer irreparable harm if an
injunction did not issue. As to the § 2 claim, the district court found that the
disparate burden observed by Dr. Rodden did not constitute a cognizable harm
under the VRA because it did not meaningfully limit minority groups’ access to the
5

The appellees here (defendants below) are the Arizona Secretary of State’s Office; Arizona Secretary of

State Michele Reagan in her official capacity; the Maricopa County Board of Supervisors; members of the Maricopa
County Board of Supervisors Denny Barney, Steve Chucri, Andy Kunasek, Clint Hickman, and Steve Gallardo in
their official capacities; the Maricopa County Recorder and Elections Department; Maricopa County Recorder Helen
Purcell and Maricopa County Elections Director Karen Osbourne in their official capacities; and Arizona Attorney
General Mark Brnovich in his official capacity.
6

For example, most OOP voters would be eligible to vote in the Presidential election even if they were
ineligible to vote for precinct-specific elections because they were not in fact residents of the precinct in which they
erroneously voted.

7

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 8 of 30

political process and was not shown to be linked to historical discrimination in
Arizona. As to the Fourteenth Amendment claim, the district court held that
Arizona’s precinct vote rule constituted a minimal burden on voting because it
simply required that voters appear at the proper polling location on election day
and was justified by the administrative advantages to the State of using a precinct
voting system. The district court also concluded that Feldman was unlikely to
succeed on her equal protection claim because she had not advanced a coherent
theory for it.
Feldman filed a timely notice of interlocutory appeal on October 15, 2016
and on October 18, 2016 filed an emergency motion with this court for an
injunction pending appeal and for an expedited appeal. On October 19, 2016, a
motions panel granted the request for an expedited appeal. The parties were
directed to file simultaneous merits briefs by October 24, 2016, and the appeal was
argued orally on October 26, 2016.7
II.
We incorporate section II of the opinion from the companion appeal that
involves the same parties (Feldman v. Arizona Sec’y of State’s Office, No. 167

In addition to this appeal, Feldman appealed another of the district court’s orders denying a separate

motion to preliminarily enjoin other election practices challenged in the complaint. That appeal has similarly been
expedited and is the subject of a separate disposition. See Feldman v. Arizona Sec’y of State’s Office, No. 16-16698,
— F.3d — (9th Cir. 2016).

8

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 9 of 30

16698, — F.3d — (9th Cir. 2016)) as describing the proper standard of review for
an interlocutory appeal of a denial of a preliminary injunction. Generally, a district
court’s decision regarding preliminary injunctive relief is subject to limited review.
Flexible Lifeline Sys., Inc., v. Precision Lift, Inc., 654 F.3d 989, 993-94 (9th Cir.
2011) (per curiam). The court should be reversed only if it abused its discretion or
based its decision on an erroneous legal standard or on clearly erroneous findings
of fact. FTC v. Enforma Natural Products, 362 F.3d 1204, 1211-12 (9th Cir.
2004).
III.
VRA Claim
We also incorporate section III.A.1 of the opinion from the companion
appeal that involves the same parties (Feldman v. Arizona Sec’y of State’s Office,
No. 16-16698, — F.3d — (9th Cir. 2016)) as describing the proper framework for
analyzing a VRA § 2 claim. This opinion adopts the two-step framework adopted
by a number of our sister circuits for evaluating a VRA § 2 claim. We use this
framework to determine whether the district court properly concluded that Feldman
was unlikely to succeed on the merits of her claim that Arizona’s precinct vote rule
violated § 2 of the VRA. Feldman offers two sets of arguments as to why the

9

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 10 of 30

district court erred in concluding she was unlikely to succeed on the merits of her
VRA § 2 claim. We consider each in turn.
A.
First, she argues that Arizona’s precinct vote rule imposes a discriminatory
burden on minority voters in violation of VRA § 2. Specifically, she asserts that
the report by Dr. Rodden, credited by the district court for the purposes of this
motion, was sufficient to show a cognizable disparate burden under the VRA.
We have grave doubts that the precinct vote rule gives minority voters less
opportunities to participate in the political process than it gives to other, majority
voters. That more minorities vote OOP than is reflective of their proportionate
number in the electorate does not prove that the precinct vote rule denies or
abridges their opportunity to learn of the locations of their precinct polling places
or to get to them in time to vote. There is no evidence in the record that minority
voters were given misinformation regarding the locations of their correct precinct
polling places, while non-minority voters were given correct information.8 Nor
was there evidence that minority voters’ precinct polling places were located where

8

There is evidence in the record that a small number of Spanish language voter communications included

erroneous information, but there is no indication that these were anything but isolated typographical errors and, in
any event, none of these errors gave minority voters incorrect information regarding the location of their polling
places.

10

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 11 of 30

it would be more difficult for minority voters to find them, than were the
corresponding precinct polling places of non-minority voters.
A.R.S. § 16-584 merely imposes a prerequisite to voting: it requires voters to
vote in their precinct or their vote will not be counted. It is analogous to a
registration requirement: If prospective voters do not register to vote, they cannot
vote; their vote will not be counted. Thus, it might seem that the required causal
connection between the voting prerequisite and the observed disproportionate
result simply was not proved. Gonzalez v. Arizona, 677 F.3d 383, 405 (2012)
(“Said otherwise, a § 2 challenge based purely on a showing of some relevant
statistical disparity between minorities and whites, without any evidence that the
challenged voting qualification causes that disparity, will be rejected.”) (internal
quotation marks omitted).
Nonetheless, we recognize that the district court found the results of the
expert witness’s tabulation to have shown disproportionate disqualification of
minority votes, ascribed by the district court in part to the precinct vote rule.
Giving due deference to the district court’s factual findings, we will presume,
without deciding, that Feldman did carry her burden of proof on the first step of
her VRA claim. But, we affirm the district court’s finding that she did not carry
her burden of proof on the second step of her VRA claim: that the disproportionate
11

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 12 of 30

burden, in light of the totality of the circumstances, interacted with racial
discrimination “to cause an inequality in the opportunities enjoyed by [minority]
and [non-minority] voters to elect their preferred representatives.” Thornburg v.
Gingles, 478 U.S. 30, 47 (1986).

B.
Second, Feldman argues that the district court misapplied the Gingles
factors, which are non-exclusive factors incorporated by the Supreme Court to
guide courts’ analyses of whether a voting practice, viewed in light of the totality
of the circumstances, limits minority access to electoral opportunities or the
political process.9 Gingles, 478 U.S. at 44-45. On appeal, Feldman asserts that the
district court erred as “a matter of fact and law” in its analysis of the Gingles
factors. As to the purported error of law, Feldman specifies that: “As a legal
matter, the court erred in holding that establishing a link between a disparate
9

These factors include: “1) the extent of any history of official discrimination in the state or political

subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to
participate in the democratic process; 2) the extent to which voting in the elections of the state or political
subdivision is racially polarized; 3) the extent to which the state or political subdivision has used unusually large
election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that
may enhance the opportunity for discrimination against the minority group; 4) if there is a candidate slating process,
whether the members of the minority group have been denied access to that process; 5) the extent to which members
of the minority group in the state or political subdivision bear the effects of discrimination in such areas as
education, employment and health, which hinder their ability to participate effectively in the political process; 6)
whether political campaigns have been characterized by overt or subtle racial appeals; [and] 7) the extent to which
members of the minority group have been elected to public office in the jurisdiction.” 478 U.S. at 36–37 (internal
quotation marks omitted).

12

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 13 of 30

burden and socioeconomic disparities resulting from discrimination does not
satisfy step two of the test for VRA vote-denial claims.” On the contrary, the
district court indeed applied this correct standard. It found that “Plaintiffs have
only loosely linked the observed disparities in minority OOP voting to social and
historical conditions that have produced discrimination.” The district court simply
found that Feldman had not established a sufficient link. Feldman is left with her
challenge to the district court's factual findings on the second prong, which are
reviewed for clear error. Independent Living Ctr. of S. California, Inc. v. Shewry,
543 F.3d 1050, 1055 (9th Cir. 2008). At oral argument, Feldman's counsel chose
not to press her claim of factual finding error. Nonetheless, we must consider
whether the district court's conclusion that the factual record did not show that the
observed burden “in part [was] caused by or linked to social and historical
conditions that have or currently produce discrimination against members of the
protected class” as reflected in the Gingles factors was clearly erroneous. League
of Women Voters of N.C. v. North Carolina, 769 F.3d 224, 240 (4th Cir. 2014)
(internal quotation marks omitted).
In challenging the district court’s factual findings, Feldman references the
expert report of another expert witness, Dr. Allan Lichtman, as well as evidence
(some of it also from Dr. Lichtman’s report) that minority voters 1) have higher
13

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 14 of 30

rates of residential mobility, 2) have less access to vehicles and hold jobs without
flexible working hours, and 3) cannot inform themselves about voting rules
because of a historical language barrier, as sufficient to satisfy the second prong of
VRA § 2. For the reasons discussed below, the district court’s factual
determination that this evidence failed to establish the requisite link was not clearly
erroneous.
The district court expressly referenced the proper standard for the second
prong of a VRA § 2 analysis: “that burden [identified in the first prong] 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,” which it
recognized can be satisfied in part by turning to the Gingles factors. The district
court then looked to the totality of the circumstances (as required by the statute) to
examine the practical effect of the precinct vote rule on minority voters’ access to
the political process and electoral opportunities. Finally, the district court
examined whether Arizona’s precinct vote rule was caused by or linked to social or
historical conditions that have produced discrimination and found that various
socioeconomic disparities referenced by Feldman that correlated to some degree
with race were alone insufficient to satisfy this second prong.

14

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 15 of 30

On appeal, Feldman cites the report by Dr. Lichtman to establish that the
Gingles factors show that Arizona’s precinct vote rule is linked to or caused by
historical or ongoing discriminatory practices that limit minority access to electoral
opportunities or the political process. Reading his report reveals several
inaccuracies that would clearly justify the district court’s decision not to credit it as
sufficient to satisfy the Gingles factors. First off, Lichtman references the fact that
Arizona was subject to VRA § 5 preclearance requirements until 2013 as evidence
of official discrimination. This assertion, of course, contradicts the clear guidance
of the Supreme Court from Shelby County v. Holder, which found the formula used
to determine which states were subject to preclearance requirements
unconstitutional because it was “based on 40–year–old facts having no logical
relation to the present day.” 133 S. Ct. 2612, 2629 (2013). Then, Lichtman points
to Arizona’s passage of a voter initiative which banned “affirmative action” (racial
preferences) as evidence of a history of “official discrimination.” Ending
preferential treatment of individuals on the basis of their race, as a logical matter,
cannot be considered discrimination and such measures have consequently been
upheld by this and other courts. See Schuette v. Coalition to Defend Affirmative
Action, 134 S. Ct. 1623 (2014) (Michigan ballot Proposal 2); Coalition for Econ.
Equity v. Wilson, 122 F.3d 692 (9th Cir.1997) (California ballot Proposition 209).
15

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 16 of 30

Then, Lichtman points to other actions by Arizona to show a history of official
discrimination, such as reducing the number of polling locations in the 2016
presidential primary election, waiting several years longer than some other states to
declare Martin Luther King Day a state holiday, or promulgating voter
identification laws that were later found to be preempted by federal statute, which
have at best a very tenuous connection to discrimination.
Given such errors, and given substantial rebuttal from other experts (one of
whom stated that Lichtman’s report is “single-minded, conclusory, and one-sided,
and frequently omit[s] mention of contradictory data or important context”),
Lichtman’s report is insufficient to meet the second prong of the VRA test. Even
taken at face value, the Lichtman report fails to show that any burden from the
precinct vote rule on minorities’ opportunity to participate in the political process
is “caused by or linked to social and historical conditions that have or currently
produce discrimination against members of the protected class.” Ohio Democratic
Party v. Husted, 834 F.3d 620, 637 (6th Cir. 2016) (quoting Ohio State Conference
of NAACP v. Husted, 768 F.3d 524, 554 (6th Cir. 2014)). The Lichtman report
does not show how any historical racial discrimination caused more residential
mobility or less access to transportation, which it claims are the background
reasons for more minority voters than non-minority voters voting in the wrong
16

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 17 of 30

precinct. Accordingly, the district court did not clearly err in concluding that
Feldman failed to prove that racial discrimination is a substantial cause of any
socioeconomic disparities alleged to cause more out-of-precinct voting by
minorities.
Moreover, any linkage between racial discrimination, the mobility and
transportation access issues noted by the Lichtman report, and the evidence that
more minority voters than non-minority voters vote outside their precinct, is too
attenuated to support a claim under § 2 of the VRA, which requires a connection
between a challenged voting rule and racial discrimination. As the Sixth Circuit
explained, the second prong of the VRA test considers whether the challenged
practice “has the effect, as it interacts with social and historical conditions, of
causing racial inequality in the opportunity to vote.” Ohio Democratic Party v.
Husted, 834 F.3d at 638 (emphasis omitted); id. (concluding that “the second step
asks not just whether social and historical conditions ‘result in’ a disparate impact,
but whether the challenged voting standard or practice causes the discriminatory
impact as it interacts with social and historical conditions”) (emphasis omitted). As
the district court noted, “if the requisite causal link under § 2 could be established
primarily by pointing to socioeconomic disparities between minorities and whites,
then nearly all voting regulations could conceivably violate the VRA because
17

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 18 of 30

nearly all costs of voting are heavier for socioeconomically disadvantaged voters.”
Indeed, at oral argument, Feldman’s counsel was unable to explain why the same
causal theory used in this case would not equally invalidate states’ registration
requirements if an expert could show that the cost of registration fell more heavily
on minorities than non-minorities, and fewer were registered to vote.
For these reasons, the district court’s decision not to credit this report as
sufficient to prove the existence of the Gingles factors was not “illogical,
implausible, or without support in inferences that may be drawn from the facts in
the record.” United States v. Hinkson, 585 F.3d 1247, 1263 (9th Cir. 2009).
Feldman further argues that evidence showing that minority voters change
residences more often than non-minority voters, are less able to travel to different
polling locations because they are less likely to own vehicles or have jobs with
flexible work schedules, and face language barriers as many do not speak English
should satisfy the fifth Gingles factor. However, the fifth Gingles factor reads:
“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” 478 U.S. at 37 (internal quotation marks omitted) (emphasis
added). The district court accepted that minority voters may be disparately
18

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 19 of 30

burdened by the limitations cited by Feldman, but properly found she had not
shown that these burdens are “the effects of discrimination.”10 Id. Feldman’s brief
cites only to Dr. Lichtman’s report, which the district court could properly discount
for the above-discussed reasons, and a portion of Dr. Rodden’s report discussing
the unrelated question of voter turnout rates to prove that official discrimination
caused the observed disparities. As such, it was not clearly erroneous for the
district court to conclude that Feldman failed to establish that this Gingles factor
turned in her favor.
In addition to the factors listed in Gingles, a court must also consider the
state’s interest in its electoral system, which the Supreme Court has held is “a
legitimate factor to be considered by courts among the ‘totality of the
circumstances’” in determining whether a § 2 violation has occurred. Houston
Lawyers’ Ass’n v. Attorney Gen. of Tex., 501 U.S. 419, 426–27 (1991). Here, the
district court found that there were numerous policy considerations supporting the
use of in-precinct voting requirements. Our sister circuit has also recognized that
“[t]he advantages of the precinct system are significant and numerous.” Sandusky
Cty. Democratic Party v. Blackwell, 387 F.3d 565, 569 (6th Cir. 2004). The
10

We make no claim that the fifth Gingles factor requires proof of intentional discrimination, but only, as

the text of Gingles demands, proof that the adverse socioeconomic conditions be the “effects of discrimination” and
not of something else, such as recent arrival into the workforce. 478 U.S. at 37.

19

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 20 of 30

precinct system (1) “caps the number of voters attempting to vote in the same place
on election day,” (2) “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,” (3) “allows each precinct ballot to list only those votes a citizen may
cast, making ballots less confusing,” (4) “makes it easier for election officials to
monitor votes and prevent election fraud,” and (5) “generally puts polling places in
closer proximity to voter residences.” Id.
As the Gingles factors are non-exclusive, 478 U.S. at 45, we also look to
whether Arizona’s precinct vote rule interacts with racial discrimination to limit
minority voters’ access to the political process or electoral opportunities. The
limited numerical burden of the precinct vote rule (roughly seven thousand nonminority voters and four thousand minority voters, together 0.5% of overall voters)
indicates that it does not practically affect minority access to electoral
opportunities and the political process. The existence of widely-used and
convenient alternatives to in-person voting on election day (such as early voting or
vote-by-mail) further supports this conclusion.
In sum, looking at the totality of the circumstances, the relevant Gingles
factors, the limited number of OOP ballots not counted because of Arizona’s
precinct vote rule, and the convenient alternatives to in-person voting on election
20

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 21 of 30

day together suggest that Feldman will likely not be successful on the second step
of her VRA § 2 claim. That is because we have grave doubts both that "a
searching practical evaluation" would suggest that Arizona's precinct vote rule will
limit minority individuals’ “opportunity to participate in the political processes,"
but even were such opportunity limited, that such limitation was caused by
historical or present official racial discrimination. Gingles, 478 U.S. at 44-46. The
district court’s conclusions here were not clearly erroneous.11
11

The dissent references numerous parts of the district court record but fails to engage with the factual

sufficiency of this material nor does it recognize the deferential standard of review for a district court’s factual
findings. Smith v. Salt River Project Agric. Improvement & Power Dist., 109 F.3d 586, 591 (9th Cir. 1997). For
example, the dissent points to another expert report, by Dr. David Berman, which discusses racial discrimination in
Arizona history as sufficient to satisfy the first Gingles factor. Again, as with the Lichtman report, the conclusions
of Berman’s report are not unrebutted. Appellees’ expert opined that Berman’s description of a “linear trajectory
[from] past discrimination to current legislation concerning voting…is a misuse of historical interpretation.” Much
of Berman’s report focuses on a voting literacy test passed in Arizona in 1912 and a 1928 Arizona Supreme Court
ruling limiting Native Americans’ eligibility to vote in state (as opposed to national) elections, events occurring
roughly one-hundred years ago. Since the 1970s the evidence of racial discrimination in Arizona cited by Dr.
Berman has been notably thin, namely that Arizona had been subject to preclearance requirements (later declared
unconstitutional) and had promulgated voter ID and citizenship voting laws that have later been upheld by this court
or preempted by federal legislation, respectively. See Gonzalez, 677 F.3d at 383. Also, English-only education
initiatives are often widely supported by minority voters because they want to learn English and to our view promote
social inclusion by ensuring all Americans have a common means of communication and improved education. As
with the Lichtman report, these inaccuracies and misstatements would offer substantial justification for the district
court to disregard Berman’s report.
As to the sixth Gingles factor, the dissent claims there was “substantial evidence” of campaigns
characterized by overt or subtle racial appeals apparently on the basis of four discrete campaign statements
referenced in Dr. Lichtman’s problematic report over the course of the past ten years. This is thin evidence indeed.
As to the seventh Gingles factor, the dissent again cites Dr. Berman’s report chronicling a disparity between
the population of certain minority groups in Arizona and their representation in the state legislature. Even assuming
these figures are accurate, they also show that Hispanic representation in the state legislature grew by nearly 40% in

(continued...)
21

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 22 of 30

IV.
Fourteenth Amendment Claim
The district court concluded that Arizona’s precinct vote rule was
constitutionally permissible under the Anderson-Burdick framework because it
imposed a minimal burden on voting and served important regulatory and
administrative interests. Burdick v. Takushi, 504 U.S. 428, 434 (1992) (citing
Anderson v. Celebrezze, 460 U.S. 780, 789 (1983)).
Feldman accepts the Anderson-Burdick framework as the proper analytical
structure, but asserts that the district court did not properly credit the burden that
Arizona’s precinct vote rule and consequent rejection of all OOP ballots actually
places on voting rights. She emphasizes that since 2012, “some 14,500 voters”
11

(...continued)

the past decade. Other evidence in the record shows that Arizona has roughly the same number of Hispanic
legislators as California, even though California has a much larger Hispanic population. Moreover, VRA § 2
expressly provides “[t]hat nothing in this section establishes a right to have members of protected class elected in
numbers equal to their proportion in the population.” 52 U.S.C. § 10301(b).
Taken together, the district court appreciated the thin evidentiary basis (despite the thousands of pages
submitted to the court) supporting Feldman’s claim in making its factual determinations, which we review for clear
error. Salt River Project, 109 F.3d at 591. Most importantly, a mechanical recitation of the Gingles factors misses
the substantive analysis required by VRA § 2: which is "a searching practical evaluation" of the factual record to
determine if a challenged voting regulation will limit minority individuals’ “opportunity to participate in the political
processes." Gingles, 478 U.S. at 44-46. That Arizona requires all non-minority and minority in-person voters to
vote in their assigned precinct or vote center does not undermine minority access to the political process or electoral
opportunities.

22

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 23 of 30

have had their ballots discarded because of this practice, which is substantially
more (both as a proportion and absolute number) than in any other state. She
stresses that Arizona’s precinct vote rule, taken in conjunction with the difficulties
that result from voters’ confusion about polling locations due to frequent changes
and erroneous public communications, together constitute a substantial burden on
voting.
According to the district court’s calculations, Arizona’s OOP ballot policy
leads to 0.5% of total votes cast not being counted at all. Beyond being a very
small proportion of total votes cast, as the district court points out, the practical
burden that Arizona’s precinct vote rule actually imposes is nothing more than
requiring voters to go to the proper polling location or vote by mail, the option on
which roughly 80% of Arizona voters rely. While the dissent makes much of
allegedly inconvenient in-person voting procedures and frequently changing
polling locations, Feldman does not challenge these practices in this lawsuit.
Moreover, it is undisputed that Arizona undertakes substantial outreach (despite
isolated informational errors) to educate voters so they can arrive at the right
polling place. Finally, the Supreme Court has held that substantially more
demanding rules, such as requiring voters to acquire a government identification
card to cast a ballot, do not meaningfully burden voting under the Fourteenth
23

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 24 of 30

Amendment. Crawford v. Marion Cty. Election Bd., 553 U.S. 181, 197-200 (2008)
(controlling opinion of Stevens, J.). Being required to go to the proper polling
location cannot be deemed a burden on voting as this requirement is inherent to the
very essence of in-person voting. See Colo. Common Cause v. Davidson, 2004
WL 2360485, at *14 (Colo. Dist. Ct. Oct. 18, 2004) (“[I]t does not seem to be
much of an intrusion into the right to vote to expect citizens, whose judgment we
trust to elect our government leaders, to be able to figure out their polling place.”);
Serv. Emps. Int’l Union Local 1 v. Husted, 698 F.3d 341, 344 (6th Cir. 2012)
(voters should not be absolved “of all responsibility for voting in the correct
precinct or correct polling place by assessing voter burden solely on the basis of
the outcome—i.e., the state’s ballot validity determination”).
Feldman asserts that the district court erred in its assessment of the burden
by ignoring the fact that thousands of votes are not counted under the precinct vote
rule, and that the burden on the voters casting the OOP ballots is severe because
those voters are disenfranchised entirely. However, this argument ignores that
every voting qualification — including voter registration, for example — will keep
at least one person from casting a ballot that is counted. Under this theory, every
voting qualification would be subjected to strict scrutiny, because the burden
would be “severe” on at least some number of individuals, however small that
24

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 25 of 30

number might be. That result simply can’t be, as the Supreme Court expressly
instructed in Burdick. 504 U.S. at 433.
Given that the district court properly concluded that Arizona’s precinct vote
rule imposed a minimal burden on voting, the Anderson-Burdick framework does
not require a searching inquiry into the justifications for this practice. Pub.
Integrity All., Inc. v. City of Tucson, 2016 WL 4578366, at *3 (9th Cir. Sept. 2,
2016) (en banc). Arizona’s precinct vote rule and resulting policy of rejecting
OOP ballots makes administering a precinct voting system considerably less costly
by not requiring election workers to undertake a time-consuming and uncertain
process of separating eligible from ineligible votes on thousands of OOP ballots
and recounting the potentially valid ones. This, in turn, makes it easier to use
precinct voting systems, which have many practical advantages such as limiting
polling-place wait times and reducing voter confusion by providing only ballots
tailored to the persons and issues on which the voter is entitled to vote. See
Sandusky, 387 F.3d at 569. Moreover, public knowledge that Arizona would try to
identify and count valid portions of OOP ballots could conceivably lead to a
substantial number of voters casting their ballots at improper polling locations if
they were mostly concerned with voting in state-wide or national elections and a
consequent increase in cost and delay in counting those OOP ballots (see infra,
25

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 26 of 30

page 26 et seq.). Further, the failure to enforce the precinct vote rule will depress
voting for local candidates and issues. These reasons are sufficient to bear
Arizona’s burden of asserting interests that outweigh the minimal burden that the
precinct vote rule imposes on voting. The district court should be affirmed on this
point.12
V.
Feldman asserts that restrictions on fundamental voting rights inherently
constitute irreparable harm and that the longevity of an unconstitutional practice
offers no basis to deny a challenge to that practice. Because the district court
properly concluded that Feldman was unlikely to succeed on the merits of her
claims, it was also correct to conclude that she would not suffer irreparable harm
absent an injunction. The district court also found that Feldman’s decision to wait
until decades after the establishment of Arizona’s precinct vote rule also suggested
she would not be irreparably harmed if an injunction did not issue. Feldman’s
suggestion that she had not “delayed” referred only to her actions since filing the
lawsuit in the spring of 2016, but did not address the many years this policy stood
on the books before her lawsuit was filed. Taken together, the district court’s

12

Although Feldman raised an Equal Protection claim to the district court, she did not challenge the district

court’s ruling on this question on appeal and we do not address it here.

26

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 27 of 30

conclusion regarding the lack of irreparable harm should an injunction not issue
was not an abuse of discretion and is affirmed.
VI.
“A plaintiff seeking a preliminary injunction must establish . . . that the
balance of the equities tips in his favor, and that an injunction is in the public
interest.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008).
Alternatively, “a preliminary injunction is also appropriate when a plaintiff raises
‘serious questions’ as to the merits and ‘the balance of hardships tips sharply in
[plaintiff’s] favor.’” Puente Arizona v. Arpaio, 821 F.3d 1098, 1103 n.4 (9th Cir.
2016) (quoting Alliance for the Wild Rockies v. Cottrell, 632 F.3d 1127, 1135 (9th
Cir. 2011)). The balance of the equities and public interest preliminary injunction
factors “merge when the Government is the opposing party.” Nken v. Holder, 556
U.S. 418, 435 (2009).
The Supreme Court has affirmed that “[n]o right is more precious in a free
country than that of having a voice in the election of those who make the laws
under which, as good citizens, we must live.” Burdick, 504 U.S. at 441 (quoting
Wesberry v. Sanders, 376 U.S. 1, 17 (1964)). Nevertheless, the underlying right at
stake must be tempered by the fact that only 11,000 out of more than 2 million
ballots are affected by the law challenged by Feldman. Moreover, Ninth Circuit
27

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 28 of 30

precedent has found that “interference with impending elections is
extraordinary...and interference with an election after voting has begun is
unprecedented.” Southwest Voter Registration Educ. Project v. Shelly, 344 F.3d
914, 919 (9th Cir. 2003); see also Purcell v. Gonzalez, 549 U.S. 1, 4-5 (2006).
Along these lines, in Gonzalez v. Arizona, this court affirmed a denial of a
preliminary injunction of a voter ID law in part because the state had “invested
enormous resources in preparing to apply [the voting law]” and changing election
procedures would cause confusion among election officials and voters. 485 F.3d
1041, 1051-52 (9th Cir. 2007).
The facts of the situation here are analogous. Arizona Elections Director
Eric Spencer affied that instituting a new vote counting procedure “would likely
delay the canvass process, and therefore likely put the counties and the state past
the statutory deadlines.” He also stated, “[t]he elections budgets for counties are
likely already set and do not necessarily include funds to cover the additional labor
and duplicate ballots that would be required to count OOP ballots.” Pima County
Elections Director Brad Nelson explained that partially counting OOP ballots
would take “additional time, manpower, and financial resources” because counties
likely would use a manual approach with four election workers similar to the
method for counting damaged ballots. Nelson estimated the manual approach
28

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 29 of 30

could take up to fifteen minutes per OOP ballot. At oral argument Appellees’
counsel asserted that existing automated vote counting technologies could not
simply be altered to count the votes at the top (for national and state-wide offices)
of OOP ballots because so doing would potentially violate state and federal laws
requiring pre-use testing of all vote counting technologies. Also, if preliminary
injunctive relief is granted, a greater number of voters may decide to vote outside
their precinct or incorrectly believe they can vote at any location, creating further
systemic costs.
These facts suggest that the district court did not abuse its discretion. The
district court was clearly cognizant of the impending election and the cost that
granting preliminary injunctive relief would have both for Arizona’s election
budgets as well as uncertainty in light of last-minute rule changes.
VII.
The district court correctly concluded that Feldman was unlikely to succeed
on the merits of her VRA or Fourteenth Amendment claims. It also properly
concluded that she would not face irreparable harm if an injunction did not issue
and that the balance of equities did not tip in her favor. For these reasons, the
district court did not abuse its discretion in denying Feldman’s motion for a
preliminary injunction.
29

Case: 16-16865, 11/02/2016, ID: 10184244, DktEntry: 33-1, Page 30 of 30

AFFIRMED.

30