James March v. Ronald Wolff, Jr.

In the
United States Court of Appeals
For the Seventh Circuit
____________________
Nos. 25-1755 & 25-1808
J
AMES MARCH,
Plaintiff-Appellant, Cross-Appellee,
v.
T
OWN OF GRAND CHUTE, et al.,
Defendants-Appellees,

and

R
ONALD G. WOLFF, JR.,
Defendant-Appellee, Cross-Appellant.
____________________
Appeals from the United States District Court for the
Eastern District of Wisconsin.
No. 1:23-cv-00656-WCG — William C. Griesbach, Judge.
____________________
A
RGUED MAY 12, 2026 — DECIDED AUGUST 7, 2026
____________________
Before BRENNAN, Chief Judge, and KIRSCH and LEE, Circuit
Judges.
BRENNAN, Chief Judge. Small town politics can get messy.
The town of Grand Chute, Wisconsin, learned this lesson after

2 Nos. 25-1755 & 25-1808
a new faction was elected to its Board of Supervisors. When
the lone holdover from the old board reported new supervi-
sor Ronald Wolff to the Wisconsin Department of Justice, a
long investigation into corruption in Grand Chute ensued.
While this investigation proceeded, the new faction fired
Town Administrator James March, the highest-ranking une-
lected official in Grand Chute. March sued the supervisors
who voted in favor of his termination for retaliating against
him in violation of his First Amendment rights. In his view,
he was fired for participating in the state investigation into
Wolff. The supervisors insist they fired March for political dis-
loyalty and poor job performance.
After March initiated his lawsuit, Wolff was indicted,
tried, and ultimately acquitted on public corruption charges.
Based on information revealed in his criminal case, Wolff filed
a counterclaim, arguing that March deliberately withheld
damaging information to set him up for prosecution.
The district court granted summary judgment to the de-
fendants in both cases, and we affirm. Wolff’s cross-appeal
has little merit. And on March’s First Amendment retaliation
claim, we conclude that the supervisors are entitled to quali-
fied immunity. Because our precedent on firing “policymak-
ing officials” is not a model of clarity, a reasonable official
could think firing March under these circumstances would
not offend the Constitution.
I
The background for this fact-intensive case is presented in
five chapters: the new supervisors’ rise to power; the start of
the Wisconsin DOJ investigation; the fallout from the investi-
gation; March’s firing; and the new information raised in

Nos. 25-1755 & 25-1808 3
Wolff’s criminal case. Because we review cross-motions for
summary judgment, all facts are presented in the light most
favorable to the non-moving party as appropriate. Gill v.
Scholz, 962 F.3d 360, 363 (7th Cir. 2020).
A
The conflict in Grand Chute started as a dispute over tax
policy. For decades, the town’s Board of Supervisors imposed
special assessments—targeted property taxes—to support
certain public works projects. In 2019, the board imposed
hefty special assessments on several citizens to repair dilapi-
dated roads. One resident, Ronald Wolff, faced over $285,000
of these fees.
In response, Wolff organized a coalition to oppose the spe-
cial assessment policies. His faction, self-identified as the
“Doomed Dozen,” filed several lawsuits against the town in
state court. Wolff was a lead plaintiff in one of these suits. He
also paid a substantial portion of the attorney’s fees for the
other lawsuits.
The Doomed Dozen then turned to electoral politics. There
are five elected members of the town board, consisting of one
chairman and four supervisors.
1
Wolff supported the candi-
dacy of Jeffrey Ings for one of the supervisor roles during the
2020 election. Ings unseated an incumbent, an ally of longtime
Chairman Dave Schowalter. In the 2021 election, Wolff him-
self ran, defeating another establishment candidate. Another
ally of the Doomed Dozen, Jason Van Eperen, challenged and

1
Supervisors serve for two-year terms in Grand Chute. But the Town
holds elections every April. Elections for Chairman, as well as Supervisors
#1 and #3, occur in odd years; Supervisors #2 and #4 face reelection in
even-numbered years.

4 Nos. 25-1755 & 25-1808
beat Schowalter in the election for Chairman. Van Eperen
ended Schowalter’s tenure after fifteen years of service and
ten years as Chairman.
After appointing another ally to a vacant seat, the faction
controlled the town government. They held four of the five
seats on the board by October 2021—a major change for a
town that had only two people leave the board since 2007.
The two people most affected by this change were Super-
visor Brad Gehring and Town Administrator James March.
Gehring, the lone holdover from the old board, openly admit-
ted he disagreed with the new faction’s approach to govern-
ing the town. And as Town Administrator, March had
worked closely with the old board since he took the job in
2008. Now, March had to direct the town’s full-time staff to
implement the new board’s political agenda.
B
Controversy arose as soon as the new supervisors took of-
fice. For one, the Board had to settle the special assessment
lawsuits. This was a contentious process, given that Wolff was
now both a board member and a plaintiff suing the town. For
another, Wolff told voters that he would fire many of Grand
Chute’s full-time staff, including Administrator March. A
month before Wolff took office, March expressed concerns to
the town’s human resources department that he and others
would lose their jobs.
The most important feud, though, had to do with Wolff’s
residency. During the new board’s first few months in office,
Supervisor Gehring became convinced that Wolff was not liv-
ing in Grand Chute. Wolff owned a home in the town, but he

Nos. 25-1755 & 25-1808 5
spent substantial time at his house in another town with his
wife, who was sick with ovarian cancer.
So, Gehring initiated a formal complaint about Wolff’s
standing as a supervisor. But his challenge proved unsuccess-
ful. After a public hearing, t he Town Clerk ruled against
Gehring, finding the evidence insufficient to say that Wolff
did not reside at his Grand Chute home.
Undeterred, Gehring turned to higher authorities. In early
September 2021, he reached out to the county district attorney
about Wolff’s residency. The DA escalated the matter to the
Wisconsin Department of Criminal Investigations, which as-
signed Special Agent Jay Yerges to open an investigation.
For six months, Yerges investigated Wolff in secret. He in-
terviewed members of the town staff in their offices. He also
surveilled Wolff’s home, secured warrants to search Wolff’s
trash, and sought access to Wolff’s personal email account.
The new board members did not know about this investi-
gation when it began. That changed in March 2022, when state
authorities executed surprise search warrants at both of
Wolff’s homes and his business. Investigators interviewed
Wolff, his wife, his brother, and the members of his faction on
the board the same day. Each was asked many questions
about Wolff, most of which focused on his residency, his role
in settling the special assessment lawsuits, and whether the
new board members were meeting outside formal settings.
C
Stunned, the new board members spent several weeks try-
ing to figure out what, or who, was motivating the investiga-
tion. Several of them shared their frustration with the lack of
information surrounding the investigation in public. They did

6 Nos. 25-1755 & 25-1808
not know what criminal conduct was being investigated—
aside from Wolff’s residency, which they thought had been
resolved by the Town Clerk. And town staff opposed their ef-
forts to open an official inquiry into the origins of the state
investigation.
Around this time, Ings filed a report with the local FBI of-
fice. His letter reveals how little he knew at the time about the
investigation. Ings relayed that investigators had asked him
about Wolff’s residency, whether Wolff “had bullied anyone
... in order to gain a favorable outcome in his lawsuit media-
tion,” and whether Wolff “threatened to terminate any Grand
Chute Town employees.” The report noted that “these three
topics ... do not appear to merit the involvement of DOJ-level
law enforcement officials.” Ings also noted that Wolff’s war-
rant signaled “residency” and “misconduct in office” as the
reasons for the search. His question: “How did the complaint
grow from a concern about residency to also include miscon-
duct in office?”
The supervisors received no answers. The record shows
that they came to believe the investigation was merely a “po-
litical stunt” orchestrated by Supervisor Gehring to under-
mine their administration. After all, the DOJ had executed the
search warrant two weeks before a hotly contested election,
and local media reported on the supposedly secret raid on
Wolff’s properties almost as soon as it began. The new board
suspected foul play.
D
Nearly two months after the search of Wolff’s properties,
the supervisors learned that Administrator March had been
involved in the investigation. Ings emailed March to ask

Nos. 25-1755 & 25-1808 7
whether town employees had interacted with investigators.
March replied, “I was told that numerous current and former
employees including myself were officially interviewed at
random times ... just like the Town Board members.” He said
the investigators “show[ed] up at random times unan-
nounced.” When Ings followed up, March did not respond.
All the board members aver they had no idea what March
said to the investigators. And they did not learn what he said
until after the investigation concluded in July 2023, nearly two
years after it began.
Meanwhile, relations between March and the new faction
soured even further. Wolff had threatened to fire March on a
regular basis for years; in fall of 2022, the board obtained a
copy of his employment contract and set new performance
goals. The HR director discouraged the board from acting
against March, citing threats of a retaliation lawsuit.
At a town meeting in April 2023, however, the board pub-
licly broached terminating March’s contract. Gehring vehe-
mently opposed, insinuating that the reason his opponents
wanted to fire March was that he was “interviewed and is co-
operating” with the DOJ investigation. This devolved into po-
litical accusations. Van Eperen accused Gehring of having
“motive” and implied that the investigation was meritless;
Gehring responded that “obviously there is something there”
because “the DOJ has not backed off from the investigation.”
Less than a month later, four members of the Board (in-
cluding three members of the Doomed Dozen faction) voted
to terminate March’s employment. Gehring was the lone dis-
senter, saying “this has been the most disgusting thing” he
had seen “in his 30 years of government” experience.

8 Nos. 25-1755 & 25-1808
E
Within a few weeks of being fired, March sued several of
his antagonists and the Town under 42 U.S.C. § 1983. His the-
ory was that the Board retaliated against him for speaking to
the state DOJ during the investigation, violating his First
Amendment rights.
2

After discovery began in this civil case, the State filed a
criminal complaint against Wolff. But it was not based on
anything the Doomed Dozen had thought motivated the in-
vestigation. Rather, Wolff was charged with taking a private
interest in a public contract while in office, a Class I felony.
See Wis. Stat. § 946.13(1)(a).
The charge related to an incident that occurred within the
first six weeks of Wolff’s tenure on the Town Board. Outside
his board duties, Wolff ran Lake Shore Cleaners, Inc., a land-
scaping business which fulfilled many public works contracts
for the town. In May 2021, the town’s primary contractor
reached out to Lake Shore and requested a quote for a project.
After public deliberation, the board voted to approve pay-
ment of over $25,000 to Lake Shore to seed grass at the local
community center. Wolff and Ings abstained from this vote.
At the time, the town’s lawyers and the board discussed
whether the project needed to be bid out. Both Gehring and
March also expressed minor concern about this contract when
they first met with Special Agent Yerges—months after the
contract was approved, but several weeks before Lake Shore
Cleaners completed work on the project and received

2
March also alleged the Town violated his Fourteenth Amendment
Due Process rights. The district court granted summary judgment for the
defendants on that theory, and March did not appeal it.

Nos. 25-1755 & 25-1808 9
payment. Yet the lawyers ultimately signed off, and even
Gehring voted to approve. For his part, March thought a stat-
utory exception to the state law resolved any problems.
This was the only basis for the DOJ’s charge against Wolff.
In an off-the-record meeting between Yerges, Gehring, March,
and the Grand Chute Chief of Police, Yerges explained prose-
cutors would not charge any of the other conduct that had
motivated the investigation due to insufficient evidence. And
after a jury trial Wolff was acquitted of the corruption charge.
The criminal case is relevant here for another reason: it re-
vealed that March was far more involved in the investigation
than he had admitted to the board. March was an active
participant in the investigation, and h e reached out to Yerges
several times on his own when his supervisors expressed
frustration with staff. March was also very critical of Wolff,
Ings, and Van Eperen. During his initial meeting with Yerges,
March noted they were “aligned ideologically” on the town’s
special assessment policy and described “rallies” with their
“coalition of supporters.” March called Wolff “unhinged,”
said he “often lacked sleep, appeared stressed[,] and has a
temper,” and claimed he had detected the smell of alcohol on
Wolff during board meetings. March told Yerges he was being
bullied by Wolff, and he insisted Wolff did not understand
how the town government actually worked. And March said
Supervisor Ings was bitter over failed business ventures in the
town, which gave him “a personal agenda for disapproving
of how the township does business.”
Discovery in this civil case added to the evidence of
March’s involvement. March claims that he was not politically
aligned with Brad Gehring or the old board, as the new fac-
tion believed. But his text messages tell a different story. For

10 Nos. 25-1755 & 25-1808
example, when Gehring won reelection over a member of the
Doomed Dozen faction, March texted him, “Congrats on your
decisive victory! I was able to sleep last night.” And when
Ings shared a memo about town business via text, March
laughed at it—then told Gehring to alert the media because it
“[m]ight be a good story right before the election.” March also
kept Gehring apprised of the status of this lawsuit, including
looping former Chairman Dave Schowalter in so he could in-
form the local media about it.
In light of all this, the defendants now argued that March’s
firing did not violate the First Amendment because he was a
“policymaker” who was aligned with their political enemies
and spoke “in a manner that is critical of superiors or their
stated policies.” Hagan v. Quinn, 867 F.3d 816, 819 (7th Cir.
2017). And Wolff brought a counterclaim, alleging that March
retaliated against him in violation of the First Amendment by
scheming to trick Wolff into committing a crime.
The defendants moved for summary judgment in March’s
lawsuit, as did March on Wolff’s counterclaim. The district
court granted these motions. Relevant here, the court held
that March could not succeed on the merits and that qualified
immunity shielded these defendants from liability. Both par-
ties timely appealed. We review the district court’s resolu-
tions of these motions de novo. Gill, 962 F.3d at 363.
II
March contends the new board faction fired him because
he spoke to law enforcement as part of the state’s criminal in-
vestigation against Wolff, retaliating against him because he
exercised his First Amendment rights. So, he asks us to treat
his case as if he were any other employee of the government.

Nos. 25-1755 & 25-1808 11
But March has made a critical concession: as the town’s
highest unelected official, March qualifies as a “policymaker.”
That brings his case within the ambit of the Elrod–Branti line
of political patronage cases, which permit public officials to
fire policymaking employees for political reasons without of-
fending the First Amendment. See Elrod v. Burns, 427 U.S. 347,
372–75 (1976), and Branti v. Finkel, 445 U.S. 507, 517–19 (1980).
This court’s precedent applying the Elrod–Branti doctrine
to cases involving employee speech has been less than clear.
So, this case is best resolved by granting qualified immunity
to the defendants. To show why, this section proceeds in three
parts. First, we describe the Supreme Court’s political patron-
age cases. Next, we discuss this court’s caselaw extending that
doctrine to speech cases involving policymaking officials. Fi-
nally, we explain why a reasonable official in the defendants’
shoes, looking at governing precedent in May 2023, would see
no constitutional problem with firing March on these facts.
A
To establish a prima facie case of First Amendment retali-
ation, a plaintiff must satisfy three elements: (1) the Constitu-
tion protects his speech or conduct, (2) he suffered a depriva-
tion likely to deter him from exercising his First Amendment
rights, and (3) his speech “was a motivating factor in [his] em-
ployer’s adverse action against” him. Sweet v. Town of Bargers-
ville, 18 F.4th 273, 277–78 (7th Cir. 2021) (citation modified).
Courts normally apply the Pickering–Garcetti framework
when deciding whether a fired employee engaged in pro-
tected speech. Kennedy v. Bremerton Sch. Dist., 597 U.S. 507,
527–28 (2022). A terminated employee must first show he
spoke as a private citizen on a matter of public concern.

12 Nos. 25-1755 & 25-1808
Garcetti v. Ceballos, 547 U.S. 410, 418 (2006). If so, the court
conducts “a delicate balancing of the competing interests sur-
rounding the speech and its consequences.” Id. at 423; Picker-
ing v. Bd. of Educ. of Twp. High Sch. Dist. 205, 391 U.S. 563, 568
(1968).
Some cases, though, do not require the ordinary Pickering–
Garcetti inquiry. Chief among them are the “political patron-
age cases.” In Elrod v. Burns, a plurality of the Supreme Court
concluded that employees “discharged or threatened with
discharge solely because of their partisan political affiliation
or nonaffiliation” can state a claim under the First Amend-
ment. 427 U.S. at 349. A majority of the Court adopted and
refined that rule in Branti v. Finkel, holding that political pat-
ronage firings—terminating employees because they belong
to the “wrong” political party—offend the Constitution. 445
U.S. at 517. And in Rutan v. Republican Party of Illinois, the
Court extended the Elrod–Branti rule to all adverse employ-
ment actions, including “promotion, transfer, recall, and hir-
ing.” 497 U.S. 62, 65 (1990).
But these cases carve out an important exception. Employ-
ers can fire “policymaking” employees for political reasons
without offending the First Amendment. See Elrod, 427 U.S. at
367 (plurality); id. at 375 (Stewart, J., concurring) (agreeing
that the patronage rule applies only to “nonpolicymaking,
nonconfidential government employee[s]”). This exception
comes from a practical concern about making government
work: certain positions require political alignment to imple-
ment the policies of officials elected by the people. Without
such alignment, policymakers could obstruct their superiors,
thwarting representative government. Hanson v. LeVan, 967
F.3d 584, 592 (7th Cir. 2020).

Nos. 25-1755 & 25-1808 13
In short, an employer may not fire an employee because of
his party affiliation unless that employee is a “policymaker.”
The “ultimate inquiry” for identifying a policymaker is
whether “party affiliation is an appropriate requirement for
the effective performance of the public office involved.”
Branti, 445 U.S. at 518. Most courts have construed this defini-
tion broadly, treating bureaucrats of many stripes as “policy-
makers.” See Bogart v. Vermilion County, 909 F.3d 210, 213–14
(7th Cir. 2018) (collecting cases).
B
The Supreme Court has not explained the relationship be-
tween the Pickering–Garcetti and Elrod–Branti lines of cases. It
has hinted at some overlap, see Heffernan v. City of Paterson,
578 U.S. 266, 268, 270–71 (2016), but it has not said so outright.
Over thirty years ago, though, this court made the leap, ex-
tending the logic of the Elrod–Branti policymaker exception to
firings for political speech.
Our starting point was Wilbur v. Mahan, 3 F.3d 214 (7th Cir.
1993). There, an elected deputy sheriff retaliated against an
employee who chose to run against him in the upcoming elec-
tion. Id. at 215. “[A]t first blush,” these facts “present a blatant
case of retaliation for the exercise of the right of free speech.”
Id. Still, the court held that the sheriff was not liable because
this case came “within the scope of the concern that gave rise
to the exceptions” in the Elrod–Branti cases. Id. at 217.
As the court explained, “[a]n elected official is entitled to
insist on the loyalty of his policymaking subordinates,” and
“experience and common sense” showed such “height[s] of
disloyalty” need not be tolerated. Id. at 218–19. It found ab-
surd the idea that the First Amendment “forc[ed] a public

14 Nos. 25-1755 & 25-1808
official to ... retain, in a confidential or policymaking job, per-
sons who are not his political friends and may be his political
enemies.” Id. at 218. Thus, the court inaugurated a new line of
cases permitting government employers to fire policymakers
with impunity. Id. (“Once the employee is classified as ... pol-
icymaking, he can be fired on political grounds even if there
is no evidence that he would not serve ... loyally.”).
For the next several decades, this court reaffirmed this
doctrine but struggled to define its scope. The court bristled
at the thought that employers might fire policymakers for
“speech unrelated to political affiliation or viewpoint.” Bonds
v. Milwaukee County, 207 F.3d 969, 977 (7th Cir. 2000). So, it
eventually concluded that “[e]ven these policymaking em-
ployees ... possess a minimal level of First Amendment pro-
tection against retaliatory dismissal.” Embry v. City of Calumet
City, 701 F.3d 231, 235 (7th Cir. 2012). And it imposed more
rules to that effect. Vargas-Harrison v. Racine Unified Sch. Dist.,
272 F.3d 964, 973 (7th Cir. 2001); Kiddy–Brown v. Blagojevich,
408 F.3d 346, 358 (7th Cir. 2005).
This doctrinal development resulted in a two-step test for
policymaker speech cases. As in all Elrod–Branti cases, the de-
fendant must first show that the employee is a “policymaker.”
Hagan, 867 F.3d at 826. Then, the defendant must show that
the policymaker’s speech was somehow connected to “politi-
cal affiliation or policy views,” or else qualify as “speech on a
matter of public concern in a manner that is critical of superi-
ors or their stated policies.” Id. (citation modified). Though
our doctrine has grown far beyond the original Elrod–Branti
Supreme Court cases, it remains a “logical outgrowth of the
Elrod–Branti line of cases,” even as it is a “corollary to the
standard Pickering analysis.” Hagan, 867 F.3d at 819–20.

Nos. 25-1755 & 25-1808 15
In sum, this court recognizes not one, but two “policy-
maker exceptions.” The first is the standard Elrod–Branti rule
for policymakers, which covers employees fired for their po-
litical affiliations. The second is the “policymaker corollary,”
which covers employees fired for their politically disloyal
speech. Hagan, 867 F.3d at 820.
C
With this background, we turn to March’s First Amend-
ment retaliation theory. This case requires us to decide
whether either of these two policymaker exceptions applies.
As in Bonds v. County of Milwaukee, though, the “unique facts
of this case” place difficult questions about the scope of this
doctrine “in unusually sharp relief.” 207 F.3d at 977. The par-
ties’ approach to this litigation has exposed three ambiguities
in our doctrine impeding resolution of this case.
This uncertainty makes qualified immunity the best way
to resolve this dispute. We will not be the first panel to rely on
qualified immunity in a thorny Elrod–Branti case. See, e.g.,
Mitchell v. Thompson, 18 F.3d 425, 427 (7th Cir. 1994); Wallace
v. Benware, 67 F.3d 655, 656 (7th Cir. 1995); Ryan v. Ill. Dep’t of
Children and Fam. Servs., 185 F.3d 751, 758–60 (7th Cir. 1999);
Wernsing v. Thompson, 423 F.3d 732, 746–47 (7th Cir. 2005);
Matrisciano v. Randle, 569 F.3d 723, 731–36 (7th Cir. 2009). Like
those panels and the district court, we conclude that a reason-
able official, looking at governing precedent, could think fir-
ing March in this situation would not offend the Constitution.
To explain why, this section outlines the qualified immun-
ity test before highlighting the ambiguities in our caselaw
about the policymaker corollary. It concludes by applying the
governing law to this case.

16 Nos. 25-1755 & 25-1808
1
“So long as [a public official’s] conduct does not violate
clearly established law, qualified immunity will shield him
from civil liability.” Stewardson v. Titus, 126 F.4th 1264, 1275
(7th Cir. 2025). The doctrine provides “ample room for mis-
taken judgments and protects all but the plainly incompetent
and those who knowingly violate the law.” Schimandle v.
Dekalb Cnty. Sheriff’s Off., 114 F.4th 648, 655 (7th Cir. 2024) (ci-
tation modified). A plaintiff seeking to defeat a qualified im-
munity defense must make two showings. Villalobos v. Picicco,
168 F.4th 1057, 1062 (7th Cir. 2026). First, he must establish
that the official’s actions violated the Constitution; second, he
must show that the unlawfulness of the conduct was clearly
established at the time of the violation. Id. A court may con-
sider these prongs in any order. Id. at 1063–64.
A plaintiff shows that the law is clearly established by
“pointing to specific cases, not general legal principles.” Id. at
1063. That is so because the Supreme Court requires the
“contours” of a rule to be “so well defined that it is clear to a
reasonable [official] that his conduct was unlawful in the sit-
uation he confronted.” City of Tahlequah v. Bond, 595 U.S. 9, 12
(2021) (citation modified). So, a plaintiff must show that both
the general principle and the facts of an analogous case clearly
establish the law to defeat qualified immunity.
But in some cases, the general principle is not clearly es-
tablished. Cf. Matrisciano, 569 F.3d at 735. In such cases, the
precedent on a doctrine has not been sufficiently developed
“to give fair warning” to officials governed by the rule. Id.
This entitles the defendants to qualified immunity. T here are
also some situations in which governing precedent obfuscates
the principles governing a case precedent, rather than making

Nos. 25-1755 & 25-1808 17
them clear. Cf. Villalobos, 168 F.4th at 1064–65 (a case “muddies
the waters” and “suggests a reasonable officer could think”
his conduct constitutional).
3

2
Keeping these principles in mind, we turn to this case.
There are three primary concerns: the parties strongly dispute
why March was terminated, how “partisan” or “political”
speech must be to trigger this court’s policymaker corollary,
and whether it matters that the defendants did not know the
contents of March’s speech at the time they fired him. We ad-
dress each in turn.
(i)
March insists he was terminated because he spoke to state
authorities during the state’s criminal investigation. The de-
fendants have invoked both the standard Elrod–Branti policy-
maker exception and its corollary. That means they raised two
theories of causation by implication: that March was fired for
his political affiliations as well as his political speech.
Defendants should be cautious about asserting both. To
raise the ordinary Elrod–Branti exception, the Supreme Court
requires defendants to concede that they fired the employee
“solely because of his political beliefs.” Branti, 445 U.S. at 508.
This court’s corollary cases follow suit: a defendant must

3
This court has said its own precedent can clearly establish the law—
even though the Supreme Court has expressly reserved that question.
Compare, e.g., Taylor v. Ways, 999 F.3d 478, 490–92 (7th Cir. 2021) (relying
on this court’s precedent alone to find clearly established law), with Dis-
trict of Columbia v. Wesby, 583 U.S. 48, 66 n.8 (2018) (reserving the question).
For the sake of deciding this case, we assume that Seventh Circuit cases
can both establish and obscure the principles governing a case.

18 Nos. 25-1755 & 25-1808
admit he fired the employee because of politically disloyal
speech. In this way, the doctrine is like an affirmative defense.
The defendant essentially admits, “Yes, I fired this employee
because of his political affiliations or his political speech,
which ordinarily states a claim for First Amendment retalia-
tion. But that does not matter because he is a policymaker.”
This means the corollary is not just a “shorthand for the
Pickering balancing” test, as we have sometimes described it.
Vargas-Harrison, 272 F.3d at 973. It is shorthand for the entire
First Amendment retaliation analysis. That suggests defend-
ants invoking Elrod–Branti should choose their theory of cau-
sation and hold fast. And defendants who wish to present
these arguments in the alternative must proceed with care.
Conflating whether March was fired for his speech or his
political affiliations has important implications. It could cre-
ate a dispute of fact precluding summary judgment. And this
court has not decided whether a case in which both speech
and political affiliations are motivating factors for termination
should proceed under the Supreme Court’s Elrod–Branti test
or under this court’s corollary for policymaker speech. But see
Heffernan, 578 U.S. at 268, 270–71 (hinting that these kinds of
cases might merge into Elrod–Branti).
Defendants hoping to take advantage of these exceptions
must be precise as to why they terminated the employee—
political speech or political affiliations. If they do not, a failure
to specify might require a jury to decide causation.
(ii)
This case also raises a thorny question our caselaw has ex-
pressly avoided: how “partisan” or “political” must the

Nos. 25-1755 & 25-1808 19
employee’s speech be to trigger the policymaker corollary,
which covers an employee fired for his political speech?
Generally, “the policy-maker corollary does not apply ...
when the speech at issue does not implicate the employee’s
politics or substantive policy viewpoints.” Vargas-Harrison,
272 F.3d at 973. But our court’s law has not always been so
consistent. Early on, this court explained that the speech
needed to be connected to “partisan politics.” Marshall v. Por-
ter Cnty. Plan Comm’n, 32 F.3d 1215, 1221 (7th Cir. 1994). Later
cases expressly disagreed, casting the exception in broader
terms. See Matrisciano, 569 F.3d at 731 (“We have not limited
this ‘policy-maker corollary’ to instances where the plaintiff’s
political views led to the adverse action.”).
The most recent precedent emphasizes whether the
speech evinces political disloyalty or criticism for superiors.
See, e.g., Hagan, 867 F.3d at 820; Kiddy–Brown, 408 F.3d at 358;
cf. Warzon v. Drew, 60 F.3d 1234, 1239 (7th Cir. 1995). These
opinions hint that speech critical of superiors but lacking a
political valence could fall within the scope of the corollary.
This rule does not necessarily follow from either Elrod–Branti
themselves or our earlier corollary caselaw. Yet it has been re-
peated several times, adding to the confusion.
At the same time, this court has said in “dicta,” Warzon, 60
F.3d at 1239 n.1, that “speech critical of a superior’s abuse of
office does not come within the policy-maker analysis.” Var-
gas-Harrison, 272 F.3d at 973 n.4. That conclusion derives from
this court’s decision in Marshall v. Porter County Plan Commis-
sion, a case that (as explained below) stands on shaky ground.
More importantly, though, this court has not applied this
principle since Marshall, so its contours are not well defined.

20 Nos. 25-1755 & 25-1808
The speech in this case lands somewhere in the middle.
March’s speech to Special Agent Yerges was not overtly parti-
san—it focused on the alleged misconduct of his superiors.
But March was intensely critical of the new board faction, at-
tacking them personally and questioning whether they had
the judgment or ability to run the town effectively. Nor is
there doubt that his speech was related “to the performance
of [his] duties,” which this court has assumed matters for ap-
plying the corollary. Bonds, 207 F.3d at 979. In the end, our
caselaw does not dictate an obvious answer.
(iii)
One final, critical difficulty is that the defendants did not
know what March said when they fired him. They learned
March spoke to investigators after state law enforcement
searched Wolff’s properties, but they insist they did not know
the content of his speech.
The defendants believe this concession helps them, but it
does not. In all this court’s corollary cases, there has been no
doubt the employers knew what the employee said before ter-
mination. If anything, this admission lends credence to
March’s argument. He contends that simply speaking to law
enforcement, without knowing whether that speech is politi-
cal or otherwise disloyal, cannot come within the scope of the
doctrine.
Still, there are two reasons why this concession does not
hurt the defendants. The first is that this court has not decided
whether the policymaker corollary is a subjective or an objec-
tive inquiry. In other First Amendment contexts, we evaluate
whether the speech is protected as an objective question of
law. Jordan v. Jewel Food Stores, 743 F.3d 509, 515 (7th Cir. 2014).

Nos. 25-1755 & 25-1808 21
In whistleblower cases, courts do not consider what defend-
ants knew about the speech when deciding whether an em-
ployee spoke as a private citizen on a matter of public concern.
E.g., Fehlman v. Mankowski, 74 F.4th 872, 874–75 (7th Cir. 2023),
Metzger v. DaRosa, 367 F.3d 699, 701–02 (7th Cir. 2004). That
matters for the causation inquiry; it does not bear on whether
the speech is protected.
The second reason follows from the first: Supreme Court
precedent provides that actual knowledge of the speech does
not determine the outcome in First Amendment retaliation
cases. In Heffernan v. City of Paterson, a police chief demoted
an officer because he thought that the officer publicly opposed
a mayoral candidate. 578 U.S. at 269. But it was a misunder-
standing—the officer had not actually supported the candi-
date. Id. All the same, the Court held that this incorrect belief
could not serve as a defense to First Amendment liability,
even though the employee “had not in fact engaged in pro-
tected political activity.” Id. at 268.
In reaching that holding, the Court reinvigorated a line of
cases concluding that “the government’s reason for [the ad-
verse employment action] is what counts.” Id. at 273 (citing
Waters v. Churchill, 511 U.S. 661 (1994)). When an “employer
wrongly, though reasonably, believed that the employee had
spoken” in a manner unprotected by the First Amendment,
the Court held in Waters that the employer would be protected
from liability. Id. at 272. All that matters is the employer’s rea-
sonable belief that the employee’s speech did not merit First
Amendment protection and that he “had dismissed the em-
ployee because of that mistaken belief.” Id.
Heffernan expressly invoked the Elrod–Branti line of cases.
578 U.S. at 268, 270. Yet this court’s policymaker cases have

22 Nos. 25-1755 & 25-1808
not considered the importance of Heffernan or its predeces-
sors. Neither did the parties. An employer’s reasonable belief
that a policymaker’s speech falls into the scope of the corol-
lary might therefore serve as a defense to liability. This court
has not decided either way.
3
Based on these uncertainties in the caselaw, the defend-
ants are entitled to qualified immunity. A reasonable official
in their positions would believe that in these circumstances
governing law allowed them to fire March.
The operative question for all the Elrod–Branti policy-
maker cases is whether March is a policymaker. As the high-
est-ranking unelected official in the town, he is. The “test for
whether a position involves policymaking is whether the po-
sition authorizes, either directly or indirectly, meaningful in-
put into government decisionmaking on issues where there is
room for principled disagreement on goals or their imple-
mentation.” Hagan, 867 F.3d at 824 (citation modified). The
Town Administrator’s core duty is to direct the town’s depart-
ments to implement the board’s policy decisions. He was also
an active participant in board meetings, offering his input on
how to solve problems and effectuate policies. It is no wonder,
then, that March concedes he qualifies as a policymaker.
This case therefore begins in the “murky” ambit of Elrod–
Branti. Hanson, 967 F.3d at 597 (citation modified). The only
remaining issue is the reason for his firing—was he let go be-
cause of his speech or his political affiliations? And if fired be-
cause of his speech, was the speech “of the kind that falls
within the scope of the corollary” under this court’s caselaw?
Hagan, 867 F.3d at 826.

Nos. 25-1755 & 25-1808 23
In resolving this issue, we construe the facts in the light
most favorable to March. So, we assume his speech was the
sole factor motivating his termination, even though the record
reveals the reasons for March’s firing were more complex.
4

Granting March this point, his claim still cannot proceed.
As described above, this court had not decided whether
an official’s “factual mistake” serves as a defense to liability
in policymaker cases. Heffernan, 578 U.S. at 272. If the defend-
ants “wrongly, though reasonably, believed” March “had not
engaged in protected speech” when he spoke to law enforce-
ment because it fell into the scope of the policymaker corol-
lary, they might not face liability. Id. (citing Waters, 511 U.S. at
679–80) (emphasis omitted). Based on governing precedent, a
reasonable officer could think he need not wait for smoking-
gun evidence of disloyalty before firing a policymaker.
As the district court explained, it was reasonable for the
defendants to think March’s speech to law enforcement fell
within the scope of the corollary. Based on the questions they
were asked by Special Agent Yerges—their only source of in-
formation about what March might have said—they knew the

4
For example, ample evidence supports the notion that March’s per-
ceived political affiliations were the primary factor motivating his termi-
nation. March has not disputed that the defendants believed he was
“working behind the scenes” with Brad Gehring “to turn Town staff
against” them. And the record shows this was not a post-hoc rationaliza-
tion. Wolff proclaimed in public settings that he believed the new board
had “cut the head of the snake off” (the snake being former Chairman
Dave Schowalter), but “the body of the snake” and the “snake mentality”
remained amongst the town staff, including March. Further, it turns out
the supervisors were correct. Agent Yerges’s report and March’s texts
show that March was aligned against the supervisors, heightening the rea-
sonableness of their belief.

24 Nos. 25-1755 & 25-1808
investigation focused on two pillars of their policy platform:
the special assessment litigation and the termination of town
staff members. See supra at 5–6. They also knew the investiga-
tion centered on Wolff’s residency, a subject they thought was
politicized because their rival, Brad Gehring, initiated those
failed challenges. In these circumstances, it was reasonable for
them to think that March’s interviews involved speech critical
of them and their policies. That is especially so because the
defendants knew someone told investigators about internal
disputes between town staff and the board, an issue that
could not be the basis for criminal charges under Wisconsin
law.
Importantly, the defendants’ conversations with law en-
forcement did not focus on the contract that formed the basis
for Wolff’s criminal charge. The record reveals even March
and Gehring did not place the contract at the top of their list
of concerns. Supra at 8–9. A reasonable official could not pre-
dict that March was speaking about conduct that nobody then
knew was alleged to be criminal.
It is true the defendants knew March was speaking as part
of a criminal investigation. But on its own, this fact is not
enough to change the outcome. The defendants did not know
the purpose of the investigation and were prevented from in-
vestigating further. Van Eperen believed that the investigators
“don’t know what they are looking for” after they questioned
him about myriad topics. Ings instructed his lawyer “to ask,
‘Was a crime committed? If yes, what is the crime? If no, why
was an eight-month investigation conducted? ’” When town
staff would not tell them what the investigation was about,
they relied on the limited information they did have—none of
which would lead a reasonable official to believe March’s

Nos. 25-1755 & 25-1808 25
speech only reported criminal behavior. Cf. Waters, 511 U.S. at
677 (measuring the reasonableness of an employer’s beliefs
based on the extent of their investigation).
Once the defendants discovered March had spoken to law
enforcement, they could reasonably conclude he had criti-
cized their policies. And both Supreme Court and this court’s
precedent did not clearly establish that employers need to
know exactly what a policymaker says to fire him. Nor did it
establish how “political” the speech must be to fall within the
scope of the corollary. So, we must grant qualified immunity
to the defendants, even assuming March’s speech was the sole
cause of his dismissal.
March insists the law on terminating government employ-
ees is clearly established by the Supreme Court’s decision in
Lane v. Franks, 573 U.S. 228 (2014), and this court’s decision in
Marshall v. Porter County Plan Commission. Neither suffices.
In Lane, a government employer fired an employee be-
cause he testified under subpoena in a trial against the em-
ployer. 573 U.S. at 235. March observes that the court did not
consider whether the plaintiff qualified as a policymaker, in-
stead deciding the case by applying the Pickering–Garcetti
framework. Id. at 238. Because March thinks the plaintiff
would be treated as a policymaker under this court’s prece-
dent, he insists Lane clearly establishes the law.
Even if March was correct about that plaintiff being a pol-
icymaker, Lane does not require a ruling for him on qualified
immunity. The problem for March is that the Court has not
decided whether the Elrod–Branti line of cases applies to cases
involving the speech of policymakers. This court made the
leap long ago. That the Supreme Court has not extended the

26 Nos. 25-1755 & 25-1808
Elrod–Branti rationale to cases of employee speech does not
overcome three decades of precedent in this court establish-
ing that either Elrod–Branti or its corollary applies.
With no on-point Supreme Court precedent, we turn to
this court’s decision in Marshall. Like this case, there an exec-
utive secretary of a city board was fired after she criticized her
employers for misconduct in office. 32 F.3d at 1217–18.
Though the secretary qualified as a policymaker, the court de-
cided she could not invoke the protection of Wilbur ( which
had been decided just one year earlier). Because the secre-
tary’s “politics” were not “implicated in the discharge,” and
her speech did not “reflect[] ... her political associations,” the
court concluded that Pickering governed. Id. at 1221.
There are factual distinctions between Marshall and this
case. But more importantly, both this court and the Supreme
Court have chipped away at Marshall’s foundation as prece-
dent. A year after Marshall was decided, the court in Warzon
described its discussion of the policymaker question as
“strictly speaking, ... dicta.” 60 F.3d at 1239 n.1. And, as out-
lined above, Warzon began a process of loosening Marshall’s
requirement that the speech at issue be overtly partisan. Be-
cause “what is ‘political’ or drives voters’ preferences can be
interpreted very broadly and is difficult to assess,” the court
thought it “wise to leave the decision of what type of speech
by policymakers warrants dismissal to the elected official.” Id.
Later cases followed suit, adding confusion by focusing more
on whether the speech is “critical of superiors or their stated
policies.” Hagan, 867 F.3d at 829 (emphasis added).
5

5
Marshall also predates the Court’s decision in Garcetti. If Marshall was
decided today, it would be difficult to tell whether the secretary spoke

Nos. 25-1755 & 25-1808 27
At the least, Marshall does not unambiguously stand for
the general principle that “speech criticizing her employer’s
abuse of office” cannot come within the scope of the policy-
maker corollary. Bonds, 207 F.3d at 979. That is especially true
because Marshall does not explain whether speech touching
on both corruption and political problems can still trigger the
corollary. The speech at issue in Hagan—employees suing
about their employer violating the constitution by denying
them their positions absent due process of law—could be read
as speech “criticizing” the employer’s “abuse of office,” yet
this court treated it as political speech triggering the corollary.
867 F.3d at 829. And the lack of application of this rule in later
cases undermines its value. A reasonable official could be for-
given for thinking this one-off discussion described as “dicta”
did not clearly establish the law. Warzon, 60 F.3d at 1239 n.1.
We conclude that the individual defendants are entitled to
qualified immunity.
6

III
Wolff’s counterclaim is simpler to resolve. After Wolff was
charged in his criminal case, he sued March, alleging that
March deliberately withheld the knowledge that the contract
he signed may be illegal—setting him up for prosecution. In

pursuant to her job duties. That casts doubt on the case’s core holding—
that the secretary’s speech was protected by the First Amendment.
6
March also sued the Town of Grand Chute itself, and municipalities
are not entitled to qualified immunity. Owen v. City of Independence, 445
U.S. 622, 657 (1980). But March did not raise a theory of Monell liability,
either before the district court or on appeal. His only argument involving
the Town was that it violated his Due Process rights—an issue he chose
not to appeal. Supra note 2.

28 Nos. 25-1755 & 25-1808
Wolff’s view, this was retaliation because of his political
views. He also raises a class-of-one theory of liability under
the Equal Protection Clause. But neither claim is viable.
Wolff’s argument is best framed as a claim of retaliatory
prosecution. See Hartman v. Moore, 547 U.S. 250, 259–66 (2006).
To state this claim, Wolff must first establish that March en-
gaged in “state action” and that there was no probable cause
supporting the prosecution. Id.; see also Lindke v. Freed, 601 U.S.
187, 193 (2024). But March himself did not exercise the “power
of the State” to file the criminal charges against Wolff. Lindke,
601 U.S. at 199. And no evidence shows that March colluded
with prosecutors or somehow induced them into bringing
charges. Hartman, 547 U.S. at 252. Wolff has made no effort to
satisfy either of these crucial elements.
As for his class-of-one equal protection theory, Wolff has
not alleged any of the required elements. He must show that
March “intentionally treated [him] differently from others
similarly situated” and that “there is no rational basis for the
difference in treatment.” McDonald v. Village of Winnetka, 371
F.3d 992, 1001 (7th Cir. 2004). To meet that standard, plaintiffs
must point to a comparator who was “similarly situated” and
treated differently. Id. at 1002. Wolff has not done so. He con-
tends that this requirement does not apply to him because
“March’s actions were so uniquely illegitimate and unconsti-
tutional.” But that is simply not how the law operates. All
class-of-one-claimants must point to a comparator; Wolff has
failed to do so, dooming his counterclaim.
IV
Grand Chute has been plagued by lawsuits since 2019. The
district court properly concluded that this set of cases must

Nos. 25-1755 & 25-1808 29
end. As the highest unelected official in the town, March was
a policymaker. So, a reasonable official looking at this court’s
ambiguous caselaw could think it permissible to fire him
without offending the First Amendment, entitling the defend-
ants to qualified immunity.
In another case, this court may need to confront the chal-
lenges posed by our policymaker precedent. But until then,
defendants who invoke either version of the Elrod–Branti doc-
trine should avoid conflating the policymaker exception with
this court’s corollary for speech.
A
FFIRMED