GREGORY LUCE and NICHOLAS NEWMAN v. TOWN OF CAMPBELL, WISCONSIN, and TIM KELEMEN

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 15-2627
GREGORY LUCE and NICHOLAS NEWMAN,
Plaintiffs-Appellants,
v.
TOWN OF CAMPBELL, WISCONSIN, and TIM KELEMEN,
Defendants-Appellees.
____________________
Appeal from the United States District Court
for the Western District of Wisconsin.
No. 14-cv-046-wmc — William M. Conley, Judge.
____________________
ARGUED JANUARY 19, 2016 — DECIDED SEPTEMBER 22, 2017
____________________
Before EASTERBROOK, ROVNER, and SYKES, Circuit Judges.
EASTERBROOK, Circuit Judge. Interstate 90 runs through
the Town of Campbell, Wisconsin. The speed limit on I-90 in
the Town is 65 miles per hour. Two streets and one pedestri-
an overpass cross the highway within the Town. A traffic
survey in 2008 found that between 23,000 and 29,000 trucks
and cars pass through the Town on I-90 every day.

-- 1 of 11 --

2 No. 15-2627
Gregory Luce and Nicholas Newman, two members of
the local Tea Party, decided that the pedestrian overpass
would be a good place to draw attention to their views. The
group’s placement of banners bearing messages such as
“HONK TO IMPEACH OBAMA” led the Town’s legislature
to enact an ordinance forbidding all signs, flags, and banners
(other than traffic-control information) on any of the three
overpasses, or within 100 feet of the end of these structures.
The ordinance is content-neutral; it does not matter what
message any privately placed sign bears. Reed v. Gilbert, 135
S. Ct. 2218 (2015). The ordinance is a time, place, and manner
limit, permitting messages to be conveyed anywhere else in
Campbell. But in this suit under 42 U.S.C. §1983 Luce and
Newman contend that the First Amendment (applied to the
states by the Fourteenth) permits them to carry or place ban-
ners and signs everywhere in the Town. The district court dis-
agreed with that contention and granted summary judgment
to the Town. See 113 F. Supp. 3d 1002 (W.D. Wis. 2015). The
court also dismissed a claim against Tim Kelemen, formerly
the Town’s chief of police. 116 F. Supp. 3d 915 (W.D. Wis.
2015). We start with the plaintiffs’ claim against Kelemen,
because his conduct may affect how to understand the gene-
sis and enforcement of the ordinance.
When the Town’s police force began to hand out citations
and escort demonstrators off the pedestrian overpass, they
responded by making video recordings and posting them on
a website. Kelemen did not take kindly to these videos, es-
pecially because one of them showed people being removed
for unfurling a large American flag. Viewers started com-
plaining that the police were mistreating the Tea Party. Kel-
emen then decided to act as a vigilante—as he said in dis-
covery, “It’s just like, you know, you want to mess with us

-- 2 of 11 --

No. 15-2627 3
… we’ll mess with you.” Kelemen decided to “mess with”
Luce by posting his name and email address on websites ca-
tering to gay men and consumers of pornography. That
caused embarrassment to Luce and led to unwanted email
and other attention. Kelemen also posted comments on the
local newspaper’s website accusing Luce of failing to pay his
property taxes and other debts and asserting that his car was
about to be repossessed. Kelemen tried to hide his role—he
signed the comments “Bill O’Reilly”—but his identity even-
tually came out, and Luce sued on a constitutional theory
(that Kelemen was penalizing both the Tea Party’s speech on
the bridge and its videos), plus state tort law.
Kelemen disgraced himself. When what he had done be-
came known, he resigned as police chief. He was prosecuted
for violating Wis. Stat. §947.0125(2)(e) (unlawful use of a
computerized communication system), pleaded no contest,
and received a diversionary disposition. The district court
held, however, that Kelemen had not violated Luce’s rights
under the First Amendment, and it relinquished supple-
mental jurisdiction over the state-law claims.
The court concluded that Kelemen was not engaged in
state action when “messing with” Luce and that the First
Amendment therefore did not apply (for it deals only with
governmental conduct). Acting as a vigilante is not part of a
police officer’s job. Kelemen did some of the dirty work
while on duty and used an office computer for some posts.
But he did not use official information or privileged access to
information. All of the facts he gathered and disclosed about
Luce, such as his physical and email addresses, were availa-
ble to the general public. Anyone else could have done ex-
actly what Kelemen did. And that’s why the district judge

-- 3 of 11 --

4 No. 15-2627
thought that he was acting in a private capacity, off on a lark
and a frolic as some cases say, rather than as a police officer.
The judge held that remedies under state law are the right
response to Kelemen’s misconduct.
A public employee’s acts occur under color of state law
when they relate to official duties. See, e.g., Gibson v. Chicago,
910 F.2d 1510, 1516 (7th Cir. 1990); Hughes v. Meyer, 880 F.2d
967, 971–72 (7th Cir. 1989). Defamation was not among Kel-
emen’s duties. What he did was not even a misguided effort
to perform an official function. His activities could be called
“related” to official duties in the sense that they were de-
signed to injure a person who criticized Kelemen’s imple-
mentation of the Town’s ordinance, but the same could be
said about the misconduct at issue in Honaker v. Smith, 256
F.3d 477 (7th Cir. 2001). There we held that a fire chief was
not acting under color of state law when he burned down
the house of a disgruntled citizen whom the chief had come
to regard as a pest. Arson is not among a fire chief’s duties,
just as defamation is not among a police chief’s. So we agree
with the district court that state law, not §1983, provides the
appropriate remedy for Kelemen’s misconduct. See also, e.g.,
Latuszkin v. Chicago, 250 F.3d 502, 505–06 (7th Cir. 2001);
Pickrel v. Springfield, 45 F.3d 1115, 1118–19 (7th Cir. 1995).
Kelemen’s behavior bears on this federal suit, however,
by undermining his credibility. Much of the information
presented to the Town’s legislature, and to the district court,
about the reason for the ordinance’s enactment came from
Kelemen. He told the legislature, and the judge, that the Tea
Party’s banners caused drivers to pull off the road to take
photographs, produced complaints from drivers about slow
and snarled traffic, and so on. Given Kelemen’s misconduct,

-- 4 of 11 --

No. 15-2627 5
it is not possible (when acting on a motion for summary
judgment) to accept his statements as truthful, even though
there was no directly opposing evidence.
This gives plaintiffs an opening. They recognize that
Campbell’s ordinance is similar to one that was enacted by
the City of Madison, Wisconsin, and sustained against con-
stitutional challenge in Ovadal v. Madison, 469 F.3d 625 (7th
Cir. 2006) (holding that the adoption of an ordinance like
Campbell’s made a constitutional challenge moot by imple-
menting a nondiscriminatory system). See also Ovadal v.
Madison, 416 F.3d 531, 536 (7th Cir. 2005) (remarking that
constitutional problems in Madison’s initial approach to the
subject could be solved if the City “prohibited not just Ova-
dal’s, but all protests and all signs on all Beltline overpass-
es”). But they insist that all time, place, and manner regula-
tions require empirical support and contend that without
Kelemen’s evidence the Town’s ordinance has none.
Plaintiffs offered some evidence of their own, in the form
of a report from traffic engineer Paul Dorothy. He reached
two principal conclusions: first, that 23,000 cars a day is light
traffic, compared with the highway’s design limit; second,
that the presentation of signs and banners on overpasses is
unlikely to cause “long traffic back-ups”, contrary to Kele-
men’s submission. (Kelemen subjectively rated Campbell’s
portion of I-90 as unusually hazardous; Dorothy’s report
shows that this assertion lacks empirical support.) For its
part, the Town offered some evidence independent of Kele-
men’s observation. Officer Casper testified that he observed
a car that had pulled off the road to take pictures of signs on
the overpass. The record contains a photograph of one car

-- 5 of 11 --

6 No. 15-2627
pulled over, with occupants taking pictures. The Town did
not conduct a formal safety evaluation, however.
The paucity of evidence from anyone other than Kelemen
leads us to ask whether record evidence supporting time,
place, and manner restrictions is always essential. Plaintiffs
say yes, relying on decisions such as McCullen v. Coakley, 134
S. Ct. 2518 (2014), and Renton v. Playtime Theatres, Inc., 475
U.S. 41 (1986). McCullen dealt with speech about abortion
and Renton with the location of sexually oriented businesses.
The jurisdictions that enacted those restrictions contended
that those topics required distinctive regulations, and the
Justices wanted some proof.
After Reed v. Gilbert a powerful reason is needed when-
ever a law classifies by speech’s content. See also, e.g., Nor-
ton v. Springfield, 806 F.3d 411 (7th Cir. 2015). Whether or not
the sorts of rules at issue in McCullen and Renton amount to
content discrimination, as Reed understood that phrase, the
Court found each classification sufficiently problematic to
require an extra degree of support. But the Justices have
never suggested that empirical support is required for all
time, place, and manner limits.
Consider, for example, a limit on loud speech or music.
The Supreme Court dealt with such limits in Kovacs v.
Cooper, 336 U.S. 77 (1949), and Ward v. Rock Against Racism,
491 U.S. 781 (1989). Each time the Court sustained the regu-
lation without requiring record evidence about how high
decibel levels affect people subjected to noise. In Clark v.
Community for Creative Non-Violence, 468 U.S. 288 (1984), the
Court rejected a challenge to the Park Service’s ban on sleep-
ing in Park Service units (such as the Mall) in Washington,
D.C., because the ban did not distinguish by the message

-- 6 of 11 --

No. 15-2627 7
anyone proposed to convey. The Justices thought that the
regulation likely reduced congestion and “wear and tear on
park properties” (id. at 299) but relied on their own assess-
ment rather than proof in the record. A dissenting opinion
criticized the Court for not demanding proof, see id. at 311
(Marshall, J., dissenting), but the majority was unmoved.
And it is easy to collect other decisions sustaining time,
place, and manner regulations on the basis of the Justices’
nonempirical assessments. See, e.g., United States v. Kokinda,
497 U.S. 720 (1990) (ban on solicitation on Post Office
grounds); Heffron v. International Society for Krishna Con-
sciousness, Inc., 452 U.S. 640 (1981) (ban on literature dissem-
ination outside designated areas of a state fair).
Metromedia, Inc. v. San Diego, 453 U.S. 490 (1981), offers
another illustration. San Diego forbade many billboards near
its highways. A majority of Justices believed that a ban
would be proper, given the ability of billboards to distract
drivers (and the fact that many billboards are aesthetic disas-
ters); they did not require proof of this effect in the record.
See id. at 508–12 (opinion of White, J., joined by Stewart,
Marshall & Powell, JJ.) (the Court “hesitate[s] to disagree
with the accumulated, common-sense judgments of local
lawmakers” and a ban would be valid for safety and aesthet-
ic reasons), 541 (Stevens, J., agreeing with this conclusion but
dissenting for other reasons), 559–61 (Burger, C.J., agreeing
with this conclusion but dissenting on other grounds), 569–
70 (Rehnquist, J., agreeing with the views expressed by Chief
Justice Burger and Justice Stevens). A plurality of the Court
found the particular statute unconstitutional because it al-
lowed some billboards while forbidding others, discriminat-
ing by content and subject matter. But seven Justices deemed

-- 7 of 11 --

8 No. 15-2627
the rationale for an across-the-board ban adequate, despite
the paucity of record evidence.
The cases we have been discussing do not excuse the ab-
sence of a good reason for regulating; every time, place, and
manner regulation requires that. So if a law were to forbid
the use of a megaphone near Times Square at noon on a
weekday, a court would insist that the city or state have
some evidence to overcome the common understanding that
the din there (and then) is already so great that a megaphone
may be needed for speech to be heard at all; but a limit on
megaphones during concerts in Central Park requires no
such empirical justification, because the potential benefits of
the rule can be appreciated without one.
A regulation of the sort the Town has adopted rests on a
belief that overhead signs and banners will cause at least
some drivers to slow down in order to read what the ban-
ners say, and perhaps to react to them (say, by blowing the
car’s horn in response to “HONK TO IMPEACH OBAMA”).
Stopping to take a picture is just an extreme version of slow-
ing down. Reading an overhead banner requires some of
each driver’s attention, and diverting attention—whether to
banners or to cell phones and texting—increases the risk of
accidents. This effect is well established for cell phones and
texting and is the basis for legislation by many jurisdictions,
uncontested in court as far as we are aware, though talking
and texting are speech.
It does not take a double-blind empirical study, or a line-
ar regression analysis, to know that the presence of overhead
signs and banners is bound to cause some drivers to slow
down in order to read the sign before passing it. When one
car slows suddenly, another may hit it unless the drivers of

-- 8 of 11 --

No. 15-2627 9
the following cars are alert—and, alas, not all drivers are
alert all the time.
Advertising signs well off a freeway don’t have the same
effect. But novel signs directly overhead will affect some
drivers who do not slow for billboards or hotel logos. And
one common finding of empirical research is that when cars
travel at different speeds—as when some slow down and
others don’t—the risk of accidents rises. A report issued by
the Federal Highway Administration summarized this way:
“There is evidence that crash risk is lowest near the average
speed of traffic and increases for vehicles traveling much
faster or slower than average. … When the consequences of
crashes are taken into account, the risk of being involved in
an injury crash is lowest for vehicles that travel near the me-
dian speed[.]” Synthesis of Safety Research Related to Speed and
Speed Management (July 1998), available at
https://www.fhwa.dot.gov/publications/research/safety/9815
4/speed.cfm. The report cites many sources for this conclu-
sion. Plaintiffs’ expert did not consider this source of risk;
Dorothy’s report principally addresses the likelihood that
signs will lead to traffic jams. But collisions, not traffic jams,
are the principal risk when cars move at different speeds.
The assessment in the agency’s report has been subject to
criticism, but it also has been supported by new data. Com-
pare Kara M. Kockleman & Jianming Ma, Freeway Speeds and
Speed Variations Preceding Crashes, Within and Across Lanes, 46
J. Transportation Research Forum 43 (Spring 2007) (not find-
ing evidence that speed variations increase crashes), with
Mohammed Quddus, Exploring the Relationship between Aver-
age Speed, Speed Variation, and Accident Rates Using Spatial Sta-
tistical Models and GIS, 5 J. Transportation Safety & Security

-- 9 of 11 --

10 No. 15-2627
27 (2013) (finding such evidence). We do not try to resolve
this controversy. It is enough to say that a state or local legis-
lature that attempts to reduce the incidence of sudden brak-
ing on a superhighway cannot be thought to be acting irra-
tionally or trying to suppress speech for no good reason.
This is enough to support the district court’s rejection of
plaintiffs’ challenge to the no-signs-on-overpasses rule. But
it does not speak to the 100-foot addition, which the Town
has not even tried to justify, despite the fact that one plaintiff
has filed an affidavit stating that he wants to demonstrate off
the overpass but within the 100-foot limit and has refrained
from doing so only because of the threat of prosecution.
The ordinance forbids a small “For Sale” sign on the front
lawn of any house near the ends of the overpasses. (The par-
ties tell us that two homes are within the 100-foot limits.) It
bans every political sign on a home’s lawn, every balloon
emblazoned “Happy Birthday” for a party in the back yard,
every “Merry Christmas” banner draped over the front door
in December, and every “Open” sign in the door of any shop
near an overpass. These prohibitions apply whether or not
the sign is large enough to attract drivers’ attention.
Time, place, and manner restrictions must serve a “sig-
nificant governmental interest” and be no more extensive
than necessary. See, e.g., Community for Creative Non-Violence,
468 U.S. at 293. It is hard to see why signs off the highway,
and too small to cause drivers to react, should be banned. Cf.
Cincinnati v. Discovery Network, Inc., 507 U.S. 410 (1993) (ban
on handbills distributed through newsracks was not justi-
fied); Ladue v. Gilleo, 512 U.S. 43 (1994) (ban on signs on front
lawns was not justified). Perhaps the Town has some justifi-
cation for the 100-foot rule, but unless it produces one the

-- 10 of 11 --

No. 15-2627 11
district court should ensure that political demonstrations
and other speech that does not jeopardize safety can pro-
ceed.
The judgment of the district court is affirmed, except to
the extent that it rejects plaintiffs’ challenge to the 100-foot
buffer zone. With respect to that issue the judgment is vacat-
ed, and the case is remanded for further proceedings.

-- 11 of 11 --