Libertarian Party of Arkansas v. John Thurston

United States Court of Appeals
For the Eighth Circuit
___________________________

No. 19-2503
___________________________

Libertarian Party of Arkansas; Sandra Chaney Richter; Michael Pakko; Ricky
Harrington, Jr.; Christopher Olson; Michael Kalagias

lllllllllllllllllllllPlaintiffs - Appellees

v.

John Thurston, in his official capacity as Secretary of State for the State of Arkansas

lllllllllllllllllllllDefendant - Appellant
____________

Appeal from United States District Court
for the Eastern District of Arkansas - Little Rock
____________

Submitted: December 11, 2019
Filed: June 18, 2020
____________

Before ERICKSON, MELLOY, and KOBES, Circuit Judges.
____________

MELLOY, Circuit Judge.

In Arkansas, new political parties may petition to appear on the next general-
election ballot on a whole-ballot basis—with candidates listed for each elected office
but without meeting individual office-by-office petitioning requirements. Arkansas
recently amended its laws to make the petition requirements for whole-ballot access
more demanding. See 2019 Ark. Acts 164, § 1. The recent amendments moved the
petition deadline farther from the general election and increased the number of
signatures required. The Libertarian Party of Arkansas (“Libertarian Party” or
“Party”) and several of its members brought suit challenging the petition requirements
and seeking preliminary and permanent injunctive relief.

The district court1 found the current petition requirements likely to be
unconstitutional and otherwise found the Libertarian Party entitled to preliminary
relief. Although the district court found a deadline far from the election to be the
most problematic aspect of the current petition requirements, the grant of preliminary
injunctive relief did not amend deadlines. Rather, the district court stated that any
court-imposed changes to the deadlines would be too great an exercise in judicial
legislation and would likely carry unintended consequences. As a result, the district
court ordered the Secretary of State to accept a timely petition with a number of
signatures equal to or greater than required under prior law. The Arkansas Secretary
of State appeals, arguing the district court (1) erred in concluding the Libertarian
Party was likely to prevail on the merits, (2) abused its discretion in otherwise
determining preliminary injunctive relief was appropriate, and (3) abused its
discretion in tailoring relief to the signature requirement when the court’s primary
concern had been the statutory deadline. We affirm.

I.

The current Arkansas statutes governing political parties’ petitions to register
for whole-ballot access impose three requirements that the Libertarian Party
challenges as a collectively unconstitutional burden: (1) a numerosity or “modicum-

1
The Honorable Kristine G. Baker, United States District Judge for the Eastern
District of Arkansas.

-2-
of-support” signature requirement; (2) a limited, rolling window for collecting
signatures; and (3) a deadline for submission.

First, for the 2020 election ballot, Arkansas requires signatures from
approximately 27,000 registered Arkansas voters. See Ark. Code Ann. § 7-7-
205(a)(2). Arkansas phrases the current requirement as a percentage of the total votes
cast in the last gubernatorial election—3%. The parties agree that, based on votes
cast in the last election, this 3% figure amounts to roughly 27,000 signatures.
Because only about half of registered voters actually voted in that last election, the
current 3% figure translates to about 1.5% of the total number of registered voters in
Arkansas. In contrast to the current statute, the prior signature requirement was a
fixed number (10,000) rather than a percentage of voters in a reference election. 2007
Ark. Acts 821, § 1. Neither the current nor the prior versions of the statute restrict
the number of petitions an individual voter may sign nor place geographic limits as
to where in the state the petition-signing voters must reside.

Second, signatures older than ninety days may not be counted. Accordingly,
the petitioning party must strategically time its canvassing efforts to obtain a viable,
rolling 90-day window, submitting its petition at the end of that window. See Ark.
Code Ann. § 7-7-205(a)(4)(B).

Third, signatures must be submitted approximately 425 days prior to the
general election. See Ark. Code Ann. § 7-7-205(a)(6). The statute defines this due
date through references to statutory dates material to the primary process for existing,
recognized major parties. For example, Arkansas Code Section 7-7-205(a)(6) sets the
petition due date as 60 days before the “party filing period.” Section 7-7-203(c)(1),
in turn, sets the “party filing period” as a seven day window beginning at noon on the
first Monday in November of the year preceding the general election. Prior to the
February 2019 statutory amendments, the “party filing period” was in March of the
election year rather than in November prior to the election year. See Ark. Code Ann.

-3-
§ 7-7-203(c)(1) (2018). As such, the recent amendment effectively moved the
petition deadline more than 120 days farther away from the general election.

The “party filing period” is a window for established parties to file for access
to state funding for their party primaries held in March and June of the election year.
See id. § 7-7-201(a) (state funding). Importantly, parties seeking whole-ballot access
via petition do not participate in primaries. Rather, if a new political party is
recognized by successful petitioning, that new party nominates its candidates by
convention and is required to submit its convention-selected nominees, id. § 7-7-
205(c)(2)(A), to the Secretary of State on the day of the established parties’
preferential primary elections, i.e., “on the first Tuesday after the First Monday in
March,” id. § 7-7-203(b)(2). In this way, the deadlines for minor parties to petition
for access and select candidates are tethered to and defined by deadlines for major-
party primaries, even though the petitioning minor parties are not involved in such
primaries. Arkansas asserts that the recent changes to the Arkansas primary schedule
and filing requirements were based at least in part on a desire for Arkansas to hold its
partisan primaries on the national Super Tuesday primary date along with several
other states.

Several additional Arkansas statutory provisions are relevant to the present
case. First, a political party may be recognized without the need for a new petition
if that party successfully petitions for an election and then earns 3% of the vote in the
actual election for governor or president. Id. §§ 7-1-101(27)(A), 7-7-205(c)(4). A
party that succeeds at the alternative path of obtaining 3% of the vote does not select
its candidates through a convention. Rather, such a party may receive state funds for
conducting a primary. Id. Second, independent candidates may petition to be placed
on the ballot for a particular local office with the lesser of 2,000 signatures or 3% of
the “qualified electors in the county, township, or district” for which office is sought.
Id. § 7-7-103(b)(1)(A). For state office or the United States Senate, independent
candidates need the lesser of 10,000 signatures or “3% of the qualified electors of the

-4-
state.” Id. § 7-7-103(b)(1)(B). Finally, political parties who do not seek whole-ballot
access, but only seek to place a candidate on the ballot for president or vice president,
may petition for such access with the signatures of 1,000 registered Arkansas voters.
Id. § 7-8-302(5)(B).

Turning to the facts at hand, the chair of the Libertarian Party in Arkansas
lobbied the legislature in Arkansas arguing against the recent statutory changes.
After failing to stop passage of the amendments, the Libertarian Party began
collecting signatures and also filed suit to challenge the amended requirements. By
the time of the preliminary injunction hearing, the Party had obtained signatures that,
with an assumed rate of validity similar to what the Party’s canvassers had obtained
in past efforts, the Party believed to be in excess of 10,000 valid signatures.2

At the preliminary injunction hearing, the Party offered an affidavit from a
purported expert on minor political parties and testimony from several witnesses: the
Party chair, a former Party candidate, and the Party vice chair. The district court
denied a motion to exclude the affidavit, and the affidavit largely discussed a history
of ballot-law changes in Arkansas and parties’ success or failure under different
iterations of those laws. The State offered testimony from a professional canvassing
consultant, a staff attorney at the Secretary of State’s office, and a political science
professor. The Plaintiffs’ expert’s affidavit was largely consistent with testimony
from the State’s witnesses.

2
Ultimately, after the district court entered its order in the present case, the
Libertarian Party was able to obtain approximately 18,700 signatures. The Secretary
of State initially certified the petition as containing 12,749 valid signatures. This
number was identical to the number the Secretary of State had certified for the
Libertarian Party’s petition as submitted in an effort to gain whole-ballot access in a
prior election. The Libertarian Party noted this fact as suspicious, and the Secretary
of State amended its certification, stating that the current petition included 14,779
valid signatures.

-5-
 The Party chair testified generally about difficulties with obtaining 27,000
signatures. He discussed voters’ disinterest in politics at times far removed from an
election, the relative merits of canvassing when colleges were in session as contrasted
with the summer months, and the cost of past petition drives (always in excess of
$30,000). The Party chair stated the Party could not meet the new level of signatures
in the time required without additional financial input and greatly increased
canvassing efforts. He testified that, as canvassers are required to work faster, the
validity rate for signatures decreases, and the Party’s past petition signature validity
rates had been around 75%. The Party had successfully obtained the required 10,000
signatures in 2012, 2014, 2016, and 2018. In addition, the Party had gained
increasing but still small percentages of votes for statewide offices, including 2.9%
of the vote for governor in the last election, just 0.1% shy of the retention requirement
that would eliminate the need to petition for party status.

The prior Libertarian Party candidate described various difficulties with
canvassing, including voter disinterest and event planners’ disinterest at times far
removed from elections. He also stated the Party was funded by many small
donations and likely could not garner the financial support needed to meet the new
requirements. Finally, he indicated the Party’s resources are largely spent on
canvassing and ballot access fights, leaving little funding for recruiting candidates or
actual campaigning. The vice chair offered testimony consistent with the other Party
witnesses.

The Party’s expert’s affidavit described minor parties’ experience under prior
Arkansas statutes. Between 1971 and 1977 petitions for whole-ballot access in
Arkansas required signatures equal to 7% of the votes in the last gubernatorial
election. Under that 7% requirement, no new party qualified by petition. Between
1977 and 2007, the petition requirement was lower: signatures equal in number to 3%
of the votes cast in the last gubernatorial election with a 150 day window for
collecting signatures. Under that 3% requirement, no new party qualified without

-6-
court intervention. During that same time, independent candidates could petition to
be placed on the ballot for a statewide office with 10,000 signatures, and only two
candidates qualified.3 Finally, from 2007 to 2018 Arkansas statutes provided a
10,000 signature petition requirement for parties seeking whole-ballot access. Under
the 10,000 signature requirement, two elections included two petitioning parties and
the rest included only one: the Green Party qualified by petition in 2008, 2010, 2012,
and 2014, and the Libertarian Party qualified in 2012, 2014, 2016, and 2018. Neither
the Green Party nor the Libertarian Party obtained the necessary 3% in election voting
to qualify for automatic whole-ballot access through any of these several elections.
And, as noted above, under the prior statutes with these lower signature requirements,
the deadline for submitting petitions was generally 120 days or more closer to the
general election than the current requirement.

The State’s consultant referenced canvassing and petition efforts in Arkansas
and in other states for various ballot initiatives and referenda, citing numbers of
signatures obtained and time frames involved. She consistently described referendum
and ballot-initiative campaigns involving greater numbers of signatures than the
numbers at issue in the present case. She described professional canvassers as less
expensive at times far removed from elections. She also described various methods
of combining professional and volunteer canvassers in Arkansas to meet the current
requirements at a cost of $55,000 to $100,000. The Secretary of State staff attorney,

3
Between 1980 and 1996, for the office of president only, Arkansas removed
petition restrictions. This absence of restrictions resulted in ballots with increased
numbers of presidential candidates between 1980 to 1996 as follows: 7, 10, 6, 13, 13.
The parties agree these numbers represented an “overcrowded” ballot for the office
of president. In response Arkansas passed the above-referenced 1,000 signature
requirement to be placed on the ballot as a presidential candidate. After the
imposition of this minimal requirement, the number of candidates on the ballot for
president fell to a level neither party characterizes as overcrowded.

-7-
consistent with Plaintiffs’ expert’s affidavit, described third parties’ success or
failures in gaining ballot access under the older Arkansas requirements.

The State’s political science professor explained Arkansas election law. He
also described studies performed in Alabama and Tennessee and asserted the studies
showed that vote share did not change for minor candidates regardless of whether
they were independent or affiliated with a minor party. Importantly, when asked if
a ballot was overcrowded due to the presence of three or four parties the professor
answered that, no, such a ballot would not be overcrowded. The professor stated that
having a Democrat, a Republican, and a Libertarian on a ballot would not create
overcrowding.

After the hearing, the district court entered a written order granting preliminary
injunctive relief. The court applied Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d
109 (8th Cir. 1981) (en banc), employing the higher threshold for a likelihood of
success on the merits as required for challenges to duly enacted state laws. See
Planned Parenthood Minn., N.D., S.D. v. Rounds, 530 F.3d 725, 732 (8th Cir. 2008).
The court noted that a sliding scale of deference applies to ballot access review under
the First and Fourteenth Amendments, with deferential review applying unless the
statutes at issue impose “a burden of some substance” on the plaintiffs. See Moore
v. Martin, 854 F.3d 1021, 1026 (8th Cir. 2017). If such a burden exists, a form of
strict scrutiny applies and requires that the statute be “narrowly drawn to serve a
compelling state interest.” Id. (quoting Libertarian Party of N.D. v. Jaeger, 659 F.3d
687, 693 (8th Cir. 2011)).

The court addressed a history of court challenges to ballot access laws in
Arkansas, starting with several district court cases that did not involve appeals. The
district court noted that these cases generally suggested that a 10,000 signature
requirement was permissible but a 3% requirement coupled with a deadline in January
of an election year was not. See, e.g., Citizens to Establish a Reform Party in Ark. v.

-8-
Priest, 970 F. Supp. 690, 699 (E.D. Ark. 1996). The court next turned to Eighth
Circuit and Supreme Court cases, compared them to the currently challenged statutory
requirements, and found the current requirements imposed a severe burden such that
strict scrutiny applied.

The court then described the parties’ evidence, finding that the Libertarian
Party had demonstrated a likelihood that they could not meet the current
requirements. The court discounted the testimony of the State’s expert consultant
regarding the feasibility of meeting the current requirements because the past efforts
she described had involved different states, different time frames, different or non-
disclosed budgets, or had dealt with referenda on issues with high levels of voter
interest. The court also found that the State’s expert’s discussion concerning what
the Libertarian Party might do in terms of hired canvassers and volunteers was
unrealistic.

Next, the court noted that the State had failed to identify a compelling state
interest. The court nevertheless suggested several potential interests, finding the
current statute was not narrowly tailored to any of these hypothetical interests. In
particular, the court noted: the absence of any cognizable problem with ballot
crowding in Arkansas; the effectiveness of the prior, less restrictive versions of the
statutes at limiting minor party involvement; and the arbitrariness of the new changes.
Referencing the State’s assertion that the statutory amendments were designed to
place primaries on Super Tuesday, the court noted that it would be a “non-sequitur”
to claim that the desire to move existing parties’ primaries served as a compelling
interest for the new statutory requirements because petitioning parties are not
involved in primaries.

Finally, the court addressed the balance of the Dataphase factors (irreparable
harm, balance of harms, public interest) and found an injunction was appropriate. In
fashioning the preliminary injunction to address the number of signatures rather than

-9-
the statutory deadline, the court restated the three challenged provisions that
collectively served to “severely burden” Plaintiffs’ rights. The court noted that “[t]he
enjoinment of any one of these elements would somewhat reduce the burdens placed
upon plaintiffs’ constitutional rights.” Ultimately, the court ordered the Secretary of
State to accept a petition from the Libertarian Party if it contained 10,000 signatures
and otherwise satisfied the balance of the statutory requirements, stating: “The Court
declines at this stage of the proceeding to alter the deadlines for submitting new party
petitions or the party filing deadline, as such changes might unintentionally affect the
complex primary calendar in the State of Arkansas.”

II.

“Ballot access restrictions implicate . . . the rights of potential candidates for
public office . . . [and] the First and Fourteenth Amendment rights of voters to cast
their ballots for a candidate of their choice and to associate for the purpose of
advancing their political beliefs.” Moore, 854 F.3d at 1025. If a statute restricting
ballot access imposes “a burden of some substance on a plaintiff’s rights” we apply
a form of strict scrutiny and uphold the statute “only if it is ‘narrowly drawn to serve
a compelling interest.’” Id. at 1026 (quoting Jaeger, 659 F.3d at 693); Timmons v.
Twin Cities Area New Party, 520 U.S. 351, 358 (1997) (“Regulations imposing
severe burdens on plaintiffs’ rights must be narrowly tailored and advance a
compelling state interest.”). Because states enjoy broad regulatory power over the
election process, however, “[l]esser burdens . . . trigger less exacting review, and a
State’s ‘important regulatory interests’ will usually be enough to justify ‘reasonable,
nondiscriminatory restrictions.’” Timmons, 520 U.S. at 358 (quoting Burdick v.
Takushi, 504 U.S. 428, 434 (1992)).

In this appeal from a grant of preliminary injunctive relief enjoining the
enforcement of state statute, Dataphase and Rounds govern our review. Dataphase
requires courts to analyze and balance “(1) the threat of irreparable harm to

-10-
[Plaintiffs]; (2) the state of the balance between this harm and the injury that granting
the injunction [would] inflict on [Arkansas]; (3) the probability that [Plaintiffs would]
succeed on the merits; and (4) the public interest.” Dataphase, 640 F.2d at 113.
Under the third factor, “parties seeking to preliminarily enjoin the ‘implementation
of a state statute’ must demonstrate that they are ‘likely to prevail on the merits.’”
Rodgers v. Bryant, 942 F.3d 451, 455 (8th Cir. 2019) (quoting Rounds, 530 F.3d at
731–32)). This heightened standard “reflects the idea that governmental policies
implemented through legislation . . . [and] developed through presumptively reasoned
democratic processes are entitled to a higher degree of deference and should not be
enjoined lightly.” Id. at 455–56 (alteration in original) (citation omitted). Generally,
if a party shows a “likely violation of his or her First Amendment rights, the other
requirements for obtaining a preliminary injunction are . . . deemed to have been
satisfied.” Minn. Citizens Concerned for Life, Inc. v. Swanson, 692 F.3d 864, 870
(8th Cir. 2012) (en banc) (citation omitted). Overall, we review a district court’s
balancing of the Dataphase factors and the manner of relief granted for an abuse of
discretion. Rounds, 530 F.3d at 733; Novus Franchising v. Dawson, 725 F.3d 885,
893 (8th Cir. 2013) (“A district court has broad discretion when ruling on a request
for preliminary injunction, and it will be reversed only for clearly erroneous factual
determinations, an error of law, or an abuse of its discretion.” (citation omitted)).

To address the likelihood of success on the merits, we compare the current
Arkansas petition requirements to requirements the Supreme Court and the Eighth
Circuit have addressed previously. This analysis is necessary to determine what these
courts deem to be substantial burdens (as relevant to determining whether strict
scrutiny applies). This analysis is also relevant to the ultimate question of a
likelihood of success on the merits.

The questions we face, however, do not call for a mere mechanical comparison
of numbers and dates. Anderson v. Celebrezze, 460 U.S. 780, 789 (1983)
(“Constitutional challenges to specific provisions of a State’s election laws . . . cannot

-11-
be resolved by any litmus-paper test . . . .”). Rather several additional non-numerical
factors also guide our analysis. First, some restrictions are acceptable, both in terms
of numerosity and deadlines, even if those restrictions favor the two-party system.
Timmons, 520 U.S. at 367. Second, no one factor stands alone; rather, requirements
must be viewed collectively to assess the overall burden. Green Party of Ark. v.
Martin, 649 F.3d 675, 684 (8th Cir. 2011); Libertarian Party v. Bond, 764 F.2d 538,
541 (8th Cir. 1985). Third, parties’ past success or failure in a particular state under
extant or prior requirements is relevant to show the necessity or burdensomeness of
the restrictions, and such history is also material to the question of whether the
statutes are narrowly tailored to a compelling state interest. Storer v. Brown, 415
U.S. 724, 742 (1974) (“Past experience will be a helpful, if not always an unerring,
guide: it will be one thing if independent candidates have qualified with some
regularity and quite a different matter if they have not.”); Martin, 649 F.3d at 684.
Fourth, the availability of alternative paths to party certification should be considered.
Martin, 649 F.3d at 687. Fifth, the lack or presence of geographic restrictions on the
signature pool is relevant as is the ability of voters to sign multiple petitions and sign
petitions without pledging votes. McLain v. Meier, 851 F.2d 1045, 1049 (8th Cir.
1988) (“McLain II”). Sixth, the existence of similar or different restrictions on
petitioning requirements for party access, independent access, ballot initiative access,
or constitutional amendment access are relevant, in part, when assessing whether
restrictions are narrowly drawn, but differences permissibly may exist between these
schemes given the differences underlying the efforts in each area. See, e.g., Storer,
415 U.S. at 745 (“[T]he political party and the independent candidate approaches to
political activity are entirely different and neither is a satisfactory substitute for the
other.”); Moore, 854 F.3d at 1025 (“Ballot access restrictions implicate not only the
rights of potential candidates . . . but also the First and Fourteenth Amendment rights
of voters to . . . associate for the purpose of advancing their political beliefs.”). And
seventh, the existence of provisions that allow a party to register for individual races,
or that allow individuals to run as independents rather than as party members, do not
relieve the state of its burden to make restrictions on whole-ballot party access

-12-
reasonable. McLain v. Meier, 637 F.2d 1159, 1165 (1980) (“McLain I”) (“A
candidate who wishes to be a party candidate should not be compelled to adopt
independent status in order to participate in the electoral process.”).

Finally, deadlines far before election day are problematic because of the
general disinterest of potential voters so far removed from elections. Id. at 1164
(“[M]ost voters in fact look to third party alternatives only when they have become
dissatisfied with the platforms and candidates put forward by the established political
parties. This dissatisfaction often will not crystalize until party nominees are known.
. . . [I]t is important that voters be permitted to express their support for independent
and new party candidates during the time of the major parties’ campaigning and for
some time after the selection of candidates by party primary.”). Early filing deadlines
pose heavy burdens on minor parties and independent candidates, in part, because
they are likely to impact the ability of parties or independent candidates to procure
signatures. See Andersen, 460 U.S. at 791–92 (“An early filing deadline may have
a substantial impact on independent-minded voters. . . . [I]ssues simply do not remain
static over time. . . . It also burdens the signature-gathering efforts of independents
who decide to run in time to meet the deadline. When the primary campaigns are far
in the future and the election itself is even more remote, the obstacles facing an
independent candidate’s organizing efforts are compounded. Volunteers are more
difficult to recruit and retain, media publicity and campaign contributions are more
difficult to secure, and voters are less interested in the campaign.”).

In general, petition deadlines earlier than spring of election year are
problematic even when coupled with a generally permissible 10,000 signature
requirement. By comparison, Arkansas’s present requirement for 27,000 signatures,
425 days prior to the election and with a rolling 90 day window to obtain the
signatures sets a much higher bar.

-13-
 The parties dispute the importance of numerical guideposts as found in
Supreme Court and Eighth Circuit cases. The Supreme Court in Williams v. Rhodes
held a new-party petitioning regime was unconstitutional based on a deadline of
February in the election year (approximately 250 days prior to election) with a
numerosity requirement equal in number to 15% of the votes cast in the prior
gubernatorial election. 393 U.S. 23, 25–26 (1968). Then, in Jenness v. Fortson, the
Court approved as constitutional a Georgia independent-candidate petition
requirement that demanded a “petition signed by at least 5% of the number of
registered voters at the last general election,” 403 U.S. 431, 432 (1971), by a deadline
approximately 180 days prior to the general election with a 180 day window
permitted for collecting signatures, id. at 432–34. More recently in Anderson, the
Court held an independent-candidate petition requirement unconstitutional even
though the laws at issue required only 5,000 signatures be submitted by March 20 in
the election year and even though the candidate had obtained far more signatures than
required. 460 U.S. at 782–83 & n.1.

The value and importance of Williams in the present analysis is questionable
because additional, uncommon statutory requirements were at issue in that case and
the numerosity requirement was much higher than the current requirements. Jenness
is significant in that the signatures required as a percentage of eligible voters was
greater than in the present case, but the filing deadline was much more forgiving. The
State relies heavily on Jenness noting that the current 3% requirement in the present
case is “effectively” a 1.5% requirement when looking at the entire pool of eligible
voters in Arkansas, as contrasted with the 5% requirement in Jenness. The State also
argues the current signature requirement should be viewed favorably due to the
absence of any geographic restrictions or limits on the number of petitions a voter
may sign. See McLain II, 851 F.2d at 1049. The Libertarian Party counters that
Anderson, coming after Williams and Jenness and requiring only 5,000 signatures by
March 20, demonstrates the Court’s view that pushing the signature requirement too

-14-
far back in the election year is unconstitutional even with an otherwise acceptable
numerosity requirement.

Our own cases appear to comport with the Libertarian Party’s interpretation of
the Court’s guideposts. In McLain I, we held North Dakota’s new-party petition
requirement unconstitutional in that it demanded 15,000 signatures (approx. 3.3% of
eligible voters) by June 1 of the election year. 637 F.2d at 1161, 1165. That deadline
was 90 days before that state’s primary and approximately 150 days before the
general election. We stated, “North Dakota’s filing deadline more than ninety days
in advance of the primary election impresses us as unnecessarily removed from the
time of the major parties’ most active campaigning. . . . Given [the] relatively high
signature requirement, the North Dakota ballot access scheme in its totality is
particularly questionable.” Id. at 1164 (footnote omitted).

Then, in McLain II, we revisited the North Dakota regime after a legislative
change. 851 F.2d at 1047. We approved as constitutional a revised statute that
required only 7,000 signatures with a deadline 200 days before the general election
and 55 days prior to the state’s primaries. Id. There, although not striking the statute
as unconstitutional, we again placed emphasis on the deadline, stating, “The most
troubling aspect of these requirements is that the signatures are due more than 200
days before the November election.” Id. at 1049.

Later, in Martin, our primary focus was not on Arkansas’s statutory petition
requirement. 649 F.3d at 679. Rather, we addressed the alternative path for a party
in Arkansas to gain recognition: securing 3% of the vote in a gubernatorial election.
In rejecting the Green Party’s claim that the election-vote-share retention path was
unconstitutional, we repeatedly emphasized the importance of alternative paths for
new-party recognition and held that the retention requirement was permissible
because, “[a]lternative parties not certified as a political party may secure ballot
access for their entire slate of candidates by filing a petition comprised of the

-15-
signatures of 10,000 registered Arkansas voters, or roughly six-tenths of one percent
of all registered Arkansas voters.” Id. at 685. This “failsafe” alternative path,
however, is precisely the statutory path that Arkansas now has made more demanding
and that the Libertarian Party challenges. Therefore, Arkansas’s recently heightened
requirements for petitioning parties not only call into question the constitutionality
of the petitioning requirements themselves, they call into question the continuing
validity of our decision in Martin.

Finally, in Moore, we recently applied strict scrutiny to the Arkansas
independent-candidate regime. 854 F.3d at 1026–27. We applied strict scrutiny
based on the signature requirement coupled with what we characterized as an early
deadline—the lesser of 10,000 signatures or 3% of the electorate and a deadline of
March 1 of election year (approximately 250 days prior to the election). Id. at 1023.
Although we concluded strict scrutiny applied, we denied summary judgment,
holding a trial was required to determine whether these restrictions were narrowly
drawn to satisfy a compelling state interest. Even so, a dissenting judge on the panel
would have simply struck the scheme as unconstitutional. See id. at 1029–30 (Smith,
C.J., dissenting).

As in Moore, we conclude strict scrutiny applies. The interests to be protected
through an independent candidate petitioning regime, as was at issue in Moore, are
the interests of the voters and the independent candidate. The party petitioning
restrictions at issue in the present case go farther and tread upon the collective
interests of party members in associating to advance political beliefs. See Williams,
393 U.S. at 30 (“[T]he state laws place burdens on two different, although
overlapping, kinds of rights—the right of individuals to associate for the
advancement of political beliefs, and the right of qualified voters, regardless of their
political persuasion, to cast their votes effectively. Both of these rights, of course,
rank among our most precious freedoms.”). Further, the requirements at issue in
Moore were less demanding: a smaller signature requirement and a deadline 250 days

-16-
prior to the election (rather than 425). Moore, 854 F.3d at 1025. We easily conclude
the current statutory requirements impose “a burden of some substance on a plaintiff’s
rights.” Id. at 1026 (quoting Jaeger, 659 F.3d at 693).

In applying strict scrutiny, the district court identified hypothetical interests as
asserted in similar cases including the need to avoid a crowded ballot or voter
confusion and the general need to “protect the integrity of the elections process.”4
Arkansas appears to adopt these alleged interests on appeal. In addition, Arkansas
points to the comparatively low vote total minor parties have attained in prior
elections and describes parties that fail to garner 3% of the vote share as frivolous and
therefore damaging to the integrity of the election.5

We harbor serious doubt that the generalized desire to maintain the integrity
of elections and prevent ballot overcrowding can be viewed as a compelling state
interest when the prior version of the statute undisputedly succeeded at preventing
ballot overcrowding. The evidence at the hearing showed that no party previously

4
Arguably, we could stop our analysis as to the Libertarian Party’s likelihood
of success on the merits with our conclusion that strict scrutiny applies. The district
court correctly noted that the State failed to articulate clearly a compelling state
interest to support the statute. We need not rest our decision on the State’s omission,
however, because, we reach the same result after a more complete analysis addressing
the hypothetical interests that the district court identified, that the state asserts on
appeal, and that largely align with interests asserted by states in other cases.
5
It is doubtful that the repeated failure to attain the retention vote share in an
election is an appropriate basis for characterizing minor parties as frivolous given the
effect that minor vote share additions and detractions might have on the outcome in
close contests between the major parties. We note, however, that Arkansas describes
a particular instance of frivolity in a prior election involving the Libertarian Party that
Arkansas characterizes as harming the general integrity of, and respect for, the
election process. Namely, the Libertarian Party placed an Elvis Presley impersonator
on the ballot for an office and listed the candidate as Elvis Presley.

-17-
achieved access during the years in which Arkansas had a 3% requirement, and even
during the many years in which Arkansas had a 10,000 signature requirement coupled
with a more forgiving deadline, there was no crowding of the ballot. The state’s own
expert witness acknowledged that only one or two new parties had ever qualified by
petition under the 10,000 signature requirement. In fact, he opined that a ballot with
two major parties and two additional parties appearing on a whole-ballot basis would
not be crowded.

Still, assuming a compelling interest exists, and taking the general boundaries
established by the cases discussed above, a regime containing (1) a substantial
signature requirement, (2) a limited rolling window for obtaining signatures, and (3)
a deadline 425 days removed from the general election is not narrowly tailored to a
generalized interest in regulating the integrity of elections. This outcome is clear
when the unprecedented time between the deadline and the election is not based on
anything particular to petitioning parties, but instead is a date adopted by reference
to other deadlines as applicable only to established parties. The asserted desire to
move the primary date for established parties to Super Tuesday is unrelated to the
process for new-party certification or new-party candidate selection via convention.
And while there always will be some degree of arbitrariness in the precise selection
of dates, the deadline in the present case is far beyond anything we previously have
permitted. See McLain II, 851 F.3d at 1050 (“As with any percentage or numerical
requirement, the precise date of the primary is to some extent ‘necessarily arbitrary.’”
(quoting Am. Party of Tex. v. White, 415 U.S. 767, 783 (1974))).

To counter the seemingly overwhelming nature of the law against it, the State
focuses primarily on the signature requirement and ignores the relationship between
early deadlines and the ability to procure signatures. The State relies on
Jenness (which contained a 5% requirement) and several cases from other circuits or
states in which courts refused to strike petitioning regimes with similar or higher
signature requirements. See, e.g., Swanson v. Worley, 490 F.3d 894, 896 (11th Cir.

-18-
2007) (approving a signature requirement equal to 3% of voters from prior election
coupled with a June deadline for independent candidates); Populist Party v.
Herschler, 746 F.2d 656, 660–61 (10th Cir. 1984) (approving an independent
candidate requirement of signatures equal to 5% of voters in last election 45–90 days
prior to an election and 8,000 signatures by June 1 of election year for party
recognition). Here, the 3% requirement from the last election translates into about
1.5% of registered voters, numbers clearly below the figures at issue in the State’s
cases. But, the district court in the present case did not hold as an abstract matter that
Arkansas’s signature requirement, if it stood alone, would present constitutional
difficulties. The district court held the current regime as a whole was
unconstitutionally burdensome. The signature requirement was simply the most
conservative and least invasive path for the district court to follow when fashioning
a preliminary remedy.6

Given the difficulty presented by the precedent and third-party experience
under prior Arkansas statutes, the State also argues that the Libertarian Party failed
to show that the present statutory requirements actually impose a real and personal
burden on the Libertarian Party in this election cycle. This argument is essentially an

6
Further, all of the State’s cases are readily distinguishable as involving much
more forgiving deadlines or other mitigating factors. As accurately noted by Plaintiffs
in their reply brief:

[I]n each of those cases[,] the number or percentage of petition
signatures required had not been increased back to a former requirement
that had never been complied with, the petitioning time in many of the
cases allowed seven months, one year, or an unlimited amount of time
[rather than 90 days], the petition signature deadline was well into the
year of the general election rather than 424 days before the general
election, and the ballot access requirements for new political parties . . .
had actually been complied with.

Reply Br. at 26–27.

-19-
argument that the district court clearly erred in its factual assessment of the evidence
presented at the hearing. It also speaks to the Dataphase factors of irreparable harm
and the balancing of harms. In this respect, the State’s argument comes closer to the
mark.

Plaintiffs did not make an overwhelming showing as to the actual
burdensomeness of the current regime on their own particular ability or inability to
comply. Still, their showing was sufficient, and we find no clearly erroneous
determinations by the district court. Arkansas argues its witnesses described a
feasible path for compliance with a combination of paid and volunteer canvassers and
the use of 90-day windows that would capture many large public events where
signatures are more easily obtained. Arkansas also describes Plaintiffs’ witnesses’
descriptions of actual canvassing efforts as feeble at best. The district court
discounted Arkansas’s expert’s opinion as to the feasible paths to success, finding
that her cited examples of canvassing efforts were not analogous to the present
situation: canvassing for ballot initiatives where interest was high as contrasted with
canvassing for party access at a time far removed from an election. The district court
also credited the testimony of the Libertarian Party officials who (1) described the
burdensome nature of efforts less expensive than those proposed by the State’s expert
and (2) described the relative need for the Party to preserve funds for actual elections
and candidate recruitment.

At the end of the day, Party officials clearly testified that they would not be
able to meet the requirements, and their statements were fully consistent with past
experience. Past experience showed that the 10,000 signature threshold and more
forgiving deadlines were not an insurmountable bar but that no party in Arkansas
could meet the more stringent requirements. Nothing the State presented compelled
rejection of the Party’s evidence, and the district court carefully explained its factual
conclusions. The district court correctly found Plaintiffs were likely to succeed on
the merits.

-20-
 As to the remaining Dataphase factors and the balancing of those factors, we
find no abuse of discretion. The State’s pattern of success at limiting ballot access
with its prior, more forgiving petition requirements shows that any harm caused by
the current limited preliminary injunction is relatively minor in nature. The district
court permissibly found the Party’s inability to gain ballot access to be an irreparable
harm that outweighed the lesser harm to the State in enjoining one portion of the
current requirements. See Wilson v. City of Bel-Nor, 924 F.3d 995, 999 (8th Cir.
2019) (“When a plaintiff has shown a likely violation of his or her First Amendment
rights, the other requirements for obtaining a preliminary injunction are generally
deemed to have been satisfied.” (quoting Minn. Citizens Concerned for Life, 692 F.3d
at 870)).

Looking at the scope and form of the preliminary injunction, the State takes
issue with the details of how the district court structured relief—enjoining the higher
signature requirement rather than addressing a deadline. In arguing that this approach
was impermissible, however, the State overly discounts the flexible nature of
equitable, injunctive relief. Quite prudently, the district court wanted to avoid
unforeseen collateral consequences of altering a complicated deadline scheme. And,
as we noted in Rodgers,“equitable remedies are a special blend of what is necessary,
what is fair, and what is workable.” 942 F.3d at 458 (quoting Lemon v. Kurtzman,
411 U.S. 192, 200 (1973) (plurality opinion) (footnote omitted)). Simply striking the
higher signature provision in favor of the prior, 10,000 signature requirement was
workable and fair. Further, the State’s arguments in this regard fail to acknowledge
the relationship between the signature requirement and the deadlines; they are not
wholly independent requirements that lack effect on one another. All in all, the
district court’s injunction was a much more conservative approach to the problem at
hand than if the district court had altered deadlines. We find no abuse of discretion.

We affirm the judgment of the district court.
______________________________

-21-