Roy L. Bourgeois v. Bobby Peters

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT
__________________ FILED
U.S. COURT OF APPEALS
No. 02-16886 ELEVENTH CIRCUIT
October 15, 2004
__________________
THOMAS K. KAHN
CLERK
D.C. Docket No. 02-00171 CV-CDL-4

ROY L. BOURGEOIS,
JEFF WINDER,
BECKY JOHNSON,
ERIC LECOMPTE,
SCHOOL OF THE AMERICAS WATCH,

Plaintiffs-Appellants,

versus

BOBBY PETERS,
WILLIE L. DOZIER
CONSOLIDATED GOVERNMENT OF COLUMBUS, GEORGIA,

Defendants-Appellees.

____________________

Appeal from the United States District Court
for the Middle District of Georgia
_____________________

(October 15, 2004)
Before TJOFLAT, BIRCH and GOODWIN*, Circuit Judges.

TJOFLAT, Circuit Judge:

The plaintiffs in this case are an organization called “School of the

Americas Watch” (“SAW”) and several of its members, including SAW’s founder,

Rev. Roy Bourgeois. The group engages in various forms of nonviolent protest,

seeking to pressure the federal government to cut funding to the Western

Hemisphere Institute for Security Cooperation, better known as the “School of the

Americas” (SOA). The SOA is run by the United States Army and housed at Fort

Benning, Georgia. It trains military leaders from other countries throughout the

Western Hemisphere in combat and various counterinsurgency techniques. SAW

contends that the SOA bolsters military dictatorships by training their leaders how

to kill, to torture, and otherwise to suppress their citizens.

As part of its ongoing efforts to shut down the SOA, the SAW engages in an

annual protest each November on property open to the public immediately outside

of Fort Benning. Approximately 15,000 people attend the demonstration each

year. Throughout the thirteen-year history of these protests, no weapons have ever

been found at the protest site, and no protestor has ever been arrested for an act of

*
Honorable Alfred T. Goodwin, United States Circuit Judge for the Ninth Circuit, sitting
by designation.

2
violence. Each year, however, a small number of protestors violate 18 U.S.C. §

1382 by entering onto Fort Benning and attempting to march to the SOA,1 which

is actually located a few miles inside the base.

In November 2002, a week before that year’s protest, the City of Columbus

(“the City”) instituted a policy requiring everyone wishing to participate in the

protest to submit to a magnetometer (essentially, a metal detector) search at a

checkpoint “a couple of long city blocks” away from the SAW protest site. If the

magnetometer indicated the presence of metal as a protester was walking through

it, police would physically search that individual’s person and belongings. The

police estimated that protestors “would probably have to arrive . . . an hour and a

half, maybe 2 hours, ahead of time” to get through the metal detector checkpoints

1
The law provides:

Whoever, within the jurisdiction of the United States, goes upon
any military, naval, or Coast Guard reservation, post, fort, arsenal,
yard, station, or installation, for any purpose prohibited by law or
lawful regulation; or

Whoever reenters or is found within any such reservation, post,
fort, arsenal, yard, station, or installation, after having been
removed therefrom or ordered not to reenter by any officer or
person in command or charge thereof—

Shall be fined under this title or imprisoned not more than six
months, or both.

18 U.S.C. § 1382.

3
to the protest site.

The City contends that its decision to conduct mass searches was based on

several factors. First, the Department of Homeland Security threat assessment

level was “elevated,” indicating a “significant” risk of attack. Second, protestors

in previous years had demonstrated a history of “lawlessness” because many of

them engaged in frenzied dancing, did not immediately disburse at the end of the

scheduled protest, and “formed a ‘global village’ from large debris.” In addition,

some of them ignited a smoke bomb, and a few entered onto Fort Benning in a

peaceful march to the SOA. Finally, SAW had invited several “affinity

groups”—in particular, the Anarchists—to attend the protest that had allegedly

instigated violence at other, unrelated protests such as the one that led to a riot in

Seattle during a 1999 meeting of the World Trade Organization.2

SAW immediately sought a temporary restraining order and preliminary

injunction from the United States District Court for the Middle District of

Georgia,3 alleging that the searches violated protestors’ First and Fourth

2
The proposed magnetometer searches, which do nothing more than detect the presence
of metal, would have done little, if anything, to deter or to interdict any of this conduct.
3
The complaint also sought damages, but since the searches had not occurred at the time
of the trial on the merits, the damages issue was not yet ripe. Although SAW is free to file a new
§ 1983 suit for damages (subject, of course, to the statute of limitations), we have no occasion to
consider the damages claim from their original complaint.

4
Amendment rights. Two days later, the court held a hearing on preliminary relief,

which the parties agreed to consolidate with the trial on the merits. The court

refused to enter an injunction and instead dismissed the complaint. As a result, the

City conducted the magnetometer searches as planned. SAW appeals the denial of

a permanent injunction against the magnetometer searches. While this appeal was

pending, in November 2003, SAW again held a protest, and the City again

conducted magnetometer searches.

We conclude that these searches violate the First and Fourth Amendments

to the Constitution. Part I of this opinion explains why this case falls within the

exception to the mootness doctrine for issues that are “capable of repetition, yet

evading review.” Part II discusses how the City of Columbus search policy

contravenes the Fourth Amendment, while Part III sets forth the reasons why the

searches violate the First Amendment. Part IV briefly concludes.

I.

Before reaching the merits of this matter, we must first determine whether

we may exercise jurisdiction over it. Because Article III of the Constitution limits

the jurisdiction of federal courts to “cases and controversies,” Nat’l Adver. Co. v.

City of Ft. Lauderdale, 934 F.2d 283, 285-86 (11th Cir. 1991), we cannot entertain

this appeal unless an actual dispute continues to exist between the parties. If the

5
matter has become moot, we must vacate the district court’s ruling and dismiss the

case. See De La Teja v. United States, 321 F.3d 1357, 1364 (11th Cir. 2003)

(“[W]hen an issue in a case becomes moot on appeal, the court not only must

dismiss as to the mooted issue, but also vacate the portion of the district court’s

order that addresses it.”). The City points out that the original complaint sought

an injunction against mass searches at SAW’s 2002 protest. Because that protest

occurred over two years ago, they argue, this matter is moot. The point is well-

taken. “Past injury from alleged unconstitutional conduct does not in itself show a

present case or controversy regarding injunctive relief, if unaccompanied by

current adverse effects.” Lynch v. Baxley, 744 F.2d 1452, 1456 (11th Cir. 1984).

There is an exception to this general rule, however. We may entertain a

moot case if it arises from a situation that is “capable of repetition, yet evading

review.” Alabama Disabilities Advocacy Program v. J.S. Tarwater Devel., 97 F.3d

492, 496 n.1 (11th Cir. 1996) (“[E]ven if the appeal would otherwise be moot, this

case is an appropriate one to decide on the merits because the challenged action is

capable of repetition, yet evading review.”). This standard is satisfied where “(1)

the challenged action was in its duration too short to be fully litigated prior to its

cessation or expiration, and (2) there was a reasonable expectation that the same

complaining party would be subjected to the same action again.” Weinstein v.

6
Bradford, 423 U.S. 147, 149, 96 S. Ct. 347, 349, 46 L. Ed. 2d 350 (1975) (per

curiam). As a final requirement, if there exists some alternative vehicle through

which a particular policy may effectively be subject to a complete round of

judicial review, then courts will not generally employ this exception to the

mootness doctrine. See Sierra Club v. EPA, 315 F.3d 1295, 1303 n.11 (11th Cir.

2002) (“The lawfulness of the Extension Policy can be reviewed in any challenge

to EPA’s approval of [permanent air quality standards]. Thus, even if the

allegedly improper reliance on the Extension Policy [in promulgating interim

quality standards] could be repeated, it would not evade review, and the exception

to the mootness doctrine does not apply.”). The mass searches at issue in this case

satisfy all three prongs of this test.

First, the time between the City’s decision to institute magnetometer

searches at an SAW protest in a given year and the protest itself is too short to

allow full consideration by the district court, this court, and possibly the Supreme

Court. In 2002, the City announced that it was mandating magnetometer searches

barely two weeks before the protest; a two week period is clearly insufficient to

allow meaningful judicial review. Even if SAW were to file a preemptive suit for

an injunction a full year prior to its next rally, that period of time could well prove

insufficient for a complete round of judicial review. The instant appeal, for

7
example, has been pending in this court close to a year.

We have never definitively stated the amount of time the federal judiciary

must have to adjudicate a dispute that would constitute sufficient opportunity for

review. Among the time periods that we have held to be insufficient for full

judicial review of a dispute are twenty-four hours,4 an “[e]lection period[]” of

unspecified length,5 and four months.6 Roe v. Wade, 410 U.S. 113, 93 S. Ct. 705,

35 L. Ed. 2d 147 (1973), sets a useful benchmark in this area; the Supreme Court

there held that a 266-day pregnancy is insufficient time to allow for full appellate

review. See Dow Jones & Co. v. Kaye, 256 F.3d 1251, 1257 (11th Cir. 2001).

Guided by these precedents, as well as our experience with appellate dockets, we

conclude that one year is an insufficient amount of time for a district court, circuit

court of appeals, and Supreme Court to adjudicate the typical case. Consequently,

if this issue arises again regarding a future protest, it is likely to evade review

because the protest will occur before the parties have a final ruling on the merits

4
Doe v. O’Brien, 329 F.3d 1286, 1293 (11th Cir. 2003) (holding that because a period of
24 hours was insufficient to allow judicial review of challenged governmental actions, an
aggrieved party had standing to challenge them under the “capable of repetition yet evading
review” doctrine).
5
Fla. Right to Life v. Lamar, 273 F.3d 1318, 1324 n.6 (11th Cir. 2001) (“Election
periods are too short to fully litigate the constitutionality of [a law] before a given election
ends.”).
6
Sierra Club v. Martin, 110 F.3d 1551, 1554 (11th Cir. 1997) (“[T]he four-month term
of the preliminary injunction was too short to allow for appellate review prior to its expiration.”).

8
from a court of last resort.

SAW also satisfies the second prong of the “capable of repetition yet

evading review” test because it has a “reasonable expectation” that the City will

continue to implement mass, warrantless, suspicionless magnetometer searches at

SAW protests in future years. Christian Coalition v. Cole, 355 F.3d 1288, 1291

(11th Cir. 2004) (“Only when the defendant can demonstrate that there is no

reasonable expectation that the wrong will be repeated are federal courts precluded

from deciding the case on mootness grounds.” (quotation marks omitted)). This

standard requires more than a “mere possibility” that the conduct at issue will

recur, but far less than absolute certainty. Najjar v. Ashcroft, 273 F.3d 1330, 1340

(11th Cir. 2001); C & C Products, Inc. v. Messick, 700 F.2d 635, 637 (11th Cir.

1983) (“A controversy is not capable of repetition if there is only ‘a mere physical

or theoretical possibility’ of recurrence.” (citation omitted)). The threat of future

injury may not be “conjectural” or “hypothetical.” O’Shea v. Littleton, 414 U.S.

488, 494, 94 S. Ct. 669, 675, 38 L. Ed. 2d 674 (1974). The City bears a “heavy

burden” in demonstrating that SAW’s expectation that the searches will continue

is fanciful or unreasonable. United States v. W.T. Grant Co., 345 U.S. 629, 633,

73 S. Ct. 894, 897, 97 L. Ed. 1303 (1953). For this “reasonable expectation”

standard to be met, future recurrences must involve substantially the same parties.

9
See, e.g., Adler v. Duval Cty. Sch. Bd., 112 F.3d 1475, 1478 (11th Cir. 1997)

(“Because the complaining students have graduated from high school, there is no

reasonable expectation that they will be subjected to the same injury again.”);

accord Soliman v. United States, 296 F.3d 1237, 1243 (11th Cir. 2002)

(“Furthermore, since [the plaintiff] has been removed from the United States and

there is absolutely no reason to believe that he will again be detained or force-fed

under the same circumstances, the narrow exception for cases that are capable of

repetition yet evading review does not apply.”).

“Past wrongs do constitute evidence bearing on whether there is a real and

immediate threat of repeated injury which could be adverted by the issuing of an

injunction.” Lynch, 744 F.2d at 1456. The City has argued for over two years that

its search policy is constitutional and has continued to implement it even in the

face of ongoing litigation. Moreover, there has been no suggestion that the

upcoming 2004 protests will be exempt from the search policy.7 Finally, the basis

7
Indeed, even if the City declared that it would not conduct a search this year, that would
not in itself be enough to avoid the “capable of repetition” doctrine. “Where a defendant
voluntarily ceases challenged conduct, the case is not moot because nothing would prevent the
defendant from resuming its challenged action.” Sierra Club v. EPA, 315 F.3d 1295, 1303 (11th
Cir. 2003). “[V]oluntary cessation of a challenged practice renders a case moot only if there is
no ‘reasonable expectation’ that the challenged practice will resume after the lawsuit is
dismissed.” Jews for Jesus v. Hillsborough Cty. Aviation Auth., 162 F.3d 627, 629 (11th Cir.
1998).

10
for exception in this case is not predicated upon an unlikely future confluence of

events, such as that in C & C Products, 700 F.2d at 638 (refusing to resolve a

mooted discovery dispute under the “capable of repetition yet evading review”

doctrine because the plaintiff “would be subjected to [future] disclosure of

materials covered by a protective order only if it is involved in another lawsuit in

which a protective order has been entered and a nonparty moves to modify the

order”). This case also stands in stark contrast to Christian Coalition v. Cole,

wherein the government officials facing the prospect of injunction expressly

guaranteed that they would not reinstate their unconstitutional policy, in light of

an intervening Supreme Court decision. 355 F.3d 1288, 1292-93 (11th Cir. 2004)

(“[T]he [plaintiff] has every reason to believe that the [defendant’s] representation

is genuine, and can reasonably expect that the [defendant] will not [reinstate its

unconstitutional policy].”). Consequently, we find that SAW’s fear of future

magnetometer searches is reasonable.

The City contends that a Fourth Amendment case like this can almost never

fit into the “capable of repetition, yet evading review” exception to the mootness

doctrine because the decision to implement magnetometer searches is made anew

each year, based on circumstances as they exist at that time. While this is a valid

point, it does not make this case moot because all of the circumstances that the

11
City deemed sufficient to implement the search in 2002 continue to exist today,

and by all indications will continue to persist into the indefinite future.

As noted earlier, the decision to institute magnetometer searches was based

on several factors: (1) events at prior protests, (2) the affinity groups likely to get

involved, and (3) the “elevated” status of the Department of Homeland Security’s

threat advisory system. Because all of these circumstances continue to exist and

are likely to do so for the foreseeable future, it is quite likely that the City will

arrive at the same conclusion regarding future protests and implement mass

searches. Thus, SAW may reasonably expect that its future protests will continue

to be subject to these searches.

The requirement that the dispute involve the same parties is satisfied in this

case. Because this case involves the likelihood of future searches by the City at

SAW’s annual protest, and many of the plaintiffs are likely to attend these future

protests, this requirement is easily met.

Finally, there is no other available method for obtaining a complete round of

judicial review of magnetometer searches at upcoming protests. As discussed

above, a new lawsuit aimed specifically at an upcoming search is likely to take as

long as the instant one and be similarly mooted. For these reasons, we find that

we may exercise jurisdiction over the instant dispute, because while moot, it is

12
capable of repetition yet evading review.

II.

The plaintiffs’ first contention is that the mass, suspicionless, warrantless

magnetometer searches violate their Fourth Amendment right to be free of

“unreasonable searches and seizures.” We agree. The City makes several

arguments in defense of its searches; we explore each in turn.

A.

The City’s brief begins with the bold declaration that “[l]ocal governments

need an opinion that, without question, allows non-discriminatory, low-level

magnetometer searches at large gatherings.” Appellees’ Brief at 13. Citing

nothing more than a single case from 1980, the City contends that “[p]ost

September 11, 2001, this Court can determine [that] the preventive measure of a

magnetometer at large gatherings is constitutional as a matter of law.” Id. (citing

Donovan v. Dewey, 452 U.S. 594, 606 (1980)).8

8
The City’s reference to September 11 is largely new on appeal. It did not attempt to
establish by testimony or other evidence that the searches were justified by fear of a terrorist
attack. The City’s counsel made an oblique reference to that day very briefly during closing
argument before the district court: “We’ve got new terror alerts happening everyday [sic]. We
had another one last night. Everybody heard on CNN, for four different cities and hospitals in
the U.S., and they don’t know who’s coming to this event and they can’t control who’s coming.”
V2-R15-159. This passing mention of terror is the only time the issue was even plausibly put
before the district court.
That mention may have resonated with the district court, which referenced September 11
while issuing its decision from the bench:

13
 This argument is troubling. While the threat of terrorism is omnipresent, we

cannot use it as the basis for restricting the scope of the Fourth Amendment’s

protections in any large gathering of people. In the absence of some reason to

believe that international terrorists would target or infiltrate this protest, there is no

basis for using September 11 as an excuse for searching the protestors.

Even putting aside the City’s ill-advised and groundless reference to

September 11, its demand for the unbridled power to perform “magnetometer

searches at [all] large gatherings” is untenable. The text of the Fourth Amendment

contains no exception for large gatherings of people. It cannot be argued that the

Framers simply failed to foresee the possibility of large protests of this character.

The Assembly Clause of the First Amendment, expressly guaranteeing “the right

of the people peaceably to assemble,” U.S. Const. amend. I, demonstrates the

Framers’ commitment to protect individuals exercising this fundamental right

from governmental interference. The City’s request for the broad authority to

[Law enforcement officers] should be commended for their efforts
in a difficult, often impossible job, particularly given the post
September 11 environment. They are criticized when their actions
appear to tilt too much in favor of public safety and infringe upon
fundamental rights, and they are criticized when they do not go far
enough and a tragedy results.

V2-R15-162. Nothing in the substance of the court’s reasoning, however, reflected a concern
that the demonstration at issue in this case was actually a potential source or target of terrorist
threats.

14
conduct mass, suspicionless, warrantless searches is similarly bereft of any support

from either the Supreme Court or the Eleventh Circuit.

As SAW points out, under the City’s theory,

mass suspicion-less [sic] searches could be implemented
for every person who attends any large event including: a
high school graduation, a church picnic, a public concert
in the park, an art festival, a Fourth of July parade,
sporting events such as a marathon, and fund-raising
events such as the annual breast cancer walk. And if the
government began to pick and choose amongst [sic]
these groups, viewpoint discrimination would likely
result.

Reply Brief of Appellants at 4.

The City’s position would effectively eviscerate the Fourth Amendment. It

is quite possible that both protestors and passersby would be safer if the City were

permitted to engage in mass, warrantless, suspicionless searches. Indeed, it is

quite possible that our nation would be safer if police were permitted to stop and

search anyone they wanted, at any time, for no reason at all. Cf. Terry v. Ohio,

392 U.S. 1, 88 S. Ct. 1868, 20 L. Ed. 2d 889 (1968) (requiring that police

demonstrate individualized suspicion that a suspect is armed before frisking him).

Nevertheless, the Fourth Amendment embodies a value judgment by the Framers

that prevents us from gradually trading ever-increasing amounts of freedom and

privacy for additional security. It establishes searches based on evidence—rather

15
than potentially effective, broad, prophylactic dragnets—as the constitutional

norm.

We also reject the notion that the Department of Homeland Security’s threat

advisory level somehow justifies these searches. Although the threat level was

“elevated” at the time of the protest, “[t]o date, the threat level has stood at yellow

(elevated) for the majority of its time in existence. It has been raised to orange

(high) six times.” Wikipedia, Homeland Security Advisory System, available at

http://en.wikipedia.org/wiki/Department_of_Homeland_Security_Advisory_Syste

m (last referenced Aug. 16, 2004). Given that we have been on “yellow alert” for

over two and a half years now, we cannot consider this a particularly exceptional

condition that warrants curtailment of constitutional rights. We cannot simply

suspend or restrict civil liberties until the War on Terror is over, because the War

on Terror is unlikely ever to be truly over. September 11, 2001, already a day of

immeasurable tragedy, cannot be the day liberty perished in this country.

Furthermore, a system that gave the federal government the power to determine

the range of constitutionally permissible searches simply by raising or lowering

the nation’s threat advisory system would allow the restrictions of the Fourth

Amendment to be circumvented too easily. Consequently, the “elevated” alert

status does not aid the City’s case.

16
 B.

The City, quoting the district court, next contends that the search is

permissible as a “special needs” search because its purpose is “not to detect

unlawful activity or criminal wrongdoing, but . . . [to] detect[] dangerous devices

to ensure the safety of participants, spectators, and law enforcement.” Appellees’

Brief at 10. The Supreme Court has held that warrantless, suspicionless searches

are constitutionally permissible in certain narrow cases where they are meant to

further “special needs, beyond the normal need for law enforcement.” Vernonia

Sch. Dist. 47J v. Acton, 515 U.S. 646, 653, 115 S. Ct. 2386, 2391, 132 L. Ed. 2d

564 (1995) (quotation marks and citation omitted).

The City contends that the searches here were not intended to further the

City’s interest in law enforcement, but instead to help to achieve its “special need”

to keep the protestors and others safe by detecting weapons and contraband. The

City of Columbus and State of Georgia have enacted a variety of laws against the

possession or use of certain kinds of weapons, smoke bombs, and incendiary

devices to achieve this goal of public safety. As the City admits, many arrests

under these laws occurred as a result of these searches. In a case such as this,

where the very purpose of a particular law (such as the law banning the possession

of certain dangerous items) is to protect the public, and the government protects

17
the public by enforcing that law, it is difficult to see how public safety could be

seen as a governmental interest independent of law enforcement; the two are

inextricably intertwined.

Under the City’s rationale, a search intended to enforce a given law would

be permissible so long as the government officially maintained that its purpose

was to secure the objectives that motivated the law’s enactment in the first place

(e.g., public safety) rather than simply to enforce that law for its own sake. Such a

distinction is untenable. Moreover, it is difficult to conceptualize what the

government’s interest in “enforcing a law for its own sake” would be, if not to

secure the benefits of having that law enforced. Given “[t]he extensive

involvement of law enforcement and the threat of prosecution” in this search, and

our inability to tease out a rationale totally independent of the City’s interest in

law enforcement, we find that the search does not fall within the special needs

doctrine. Ferguson v. City of Charleston, 532 U.S. 67, 83-84 & n.20, 121 S. Ct.

1281, 1291 & n.20, 149 L. Ed. 2d 205 (2001); cf. Michigan Dep’t of State Police

v. Sitz, 496 U.S. 444, 454-55 (1990) (upholding highway sobriety checkpoints run

by police who arrested drunk drivers).

Even putting aside this difficulty, the type of search at issue here does not

fall within any of the “special needs” exceptions recognized by the Supreme

18
Court. Both due to the potentially unlimited sweep of the “special needs”

standard, as well as to the concerns discussed in Part II.A supra, we decline to take

it upon ourselves to craft another exception to the Fourth Amendment’s general

requirement of individualized suspicion.

C.

The City’s final argument is that this search is constitutionally permissible

because it is “reasonable.” A necessary ancillary to this argument is that the

Fourth Amendment permits all reasonable searches, whether or not the officials

conducting them have either a warrant, probable cause, or indeed any degree of

individualized suspicion. The City focuses too much on the grammatical

construction of the first half of the amendment, however. As the Supreme Court

reminds us in Chimel v. California, 395 U.S. 752, 765, 89 S. Ct. 2034, 2041, 23 L.

Ed. 2d 685 (1969), discussions of reasonableness “must be viewed in the light of

established Fourth Amendment principles.”

As the Court has repeatedly emphasized, “[T]he most basic constitutional

rule in this area is that ‘searches conducted outside the judicial process, without

prior approval by judge or magistrate, are per se unreasonable under the Fourth

Amendment—subject only to a few specially established and well-delineated

exceptions,’” Coolidge v. New Hampshire, 403 U.S. 443, 454-55, 91 S. Ct. 2022,

19
2032, 29 L. Ed. 2d 564 (1971) (quoting Katz v. United States, 389 U.S. 347, 357,

88 S. Ct. 507, 514, 19 L. Ed. 2d 576 (1967)), which are “jealously and carefully

drawn.” Jones v. United States, 357 U.S. 493, 499, 78 S. Ct. 1253, 1257, 2 L. Ed.

2d 1514 (1958) ; accord Schneckloth v. Bustamonte, 412 U.S. 218, 219, 93 S. Ct.

2041, 2043, 36 L. Ed. 2d 854 (1973); Swint v. City of Wadley, 51 F.3d 988, 995

(11th Cir. 1995). Similarly, “[i]n enforcing the Fourth Amendment’s prohibition

against unreasonable searches and seizures, the Court has insisted upon probable

cause as a minimum requirement for a reasonable search permitted by the

Constitution.” Chambers v. Maroney, 399 U.S. 42, 51, 90 S. Ct. 1975, 1980, 26 L.

Ed. 2d 419 (1970). “A search or seizure is ordinarily unreasonable in the absence

of [at least some] individualized suspicion of wrongdoing.” City of Indianapolis

v. Edmond, 531 U.S. 32, 37, 121 S. Ct. 447, 451, 148 L. Ed. 2d 333 (2000).

Conducting an ad hoc analysis of the reasonableness of the search based on

the judge’s personal opinions about the governmental and privacy interests at

stake, instead of applying the Supreme Court’s well-established per se rules

regarding warrants, prior judicial scrutiny of proposed searches, probable cause,

and individualized suspicion ignores these crucial Fourth Amendment principles.

The need to apply these per se rules reaches all searches, whether of the home,

office, person, or other location. See, e.g., O’Rourke v. Hayes, 378 F.3d 1201,

20
1206 (11th Cir. 2004) (“[The Fourth Amendment’s] protection extends to any area

in which an individual has a reasonable expectation of privacy. Offices and other

workplaces are among the areas in which individuals may enjoy such a reasonable

expectation of privacy.” (quotation marks and citations omitted)).9

In general, warrantless searches are permissible only where an individual

has a substantially reduced expectation of privacy. That expectation of privacy

has both a subjective and objective component. That is, a person must both assert

or otherwise exhibit a belief in a right to privacy in the object of the search, and

9
The Supreme Court conducted ad hoc balancing in Terry v. Ohio, 392 U.S. 1, 88 S. Ct.
1868, 20 L. Ed. 2d 889 (1968), a case involving pat-downs that the City contends is similar
to—and, indeed, involved allegedly more intrusive searches than—the instant case. We note that
the searches in this case could quite easily be as intrusive or more intrusive than those in
Terry because if a magnetometer reveals the presence of metal, the individual and his belongings
would be physically searched, and he could likely be asked to remove his shoes, belt, or other
articles of clothing. In Terry, in contrast, nothing other than a cursory pat-down could occur
unless an officer actually felt something believed to be a weapon.
Putting aside this quibble, the Court employed balancing in Terry because the searches
there involved “necessarily swift action predicated upon the on-the-spot observations of the
officer on the beat—which historically has not been, and as a practical matter could not be,
subjected to the warrant procedure.” Id. at 20, 88 S. Ct. at 1879. The searches at issue here
involve no such long historical pedigree. Moreover, the 12-year history of SAW protests prior to
the creation of this policy—as well as the long history of large and sometimes raucous public
gatherings without such searches in this case—casts doubt on any claim that officers could not
“as a practical matter”conduct their jobs without suspicionless magnetometer searches. Finally,
Terry did away with the warrant and probable cause requirements for pat-down searches, but not
an individualized suspicion requirement. Id. at 26, 88 S. Ct. at 1883 (“[T]here must be a
narrowly drawn authority to permit a reasonable search for weapons for the protection of the
police officer, where he has reason to believe that he is dealing with an armed and dangerous
individual.”). Our conclusion in this case is entirely consistent with this rationale, and as
discussed at the end of this Subpart, leaves officers with a variety of alternatives approved by the
Supreme Court for ensuring that large public gatherings are safe.

21
that expectation must be one that society is prepared to accept as reasonable.

Katz, 389 U.S. at 361, 88 S. Ct. at 516 (Harlan, J., concurring). Situations in

which such expectations are reduced include automobile searches,10 searches

incident to arrest,11 border searches,12 and searches of open fields,13 items in plain

view believed to contain contraband,14 and prisoners’ cells.15

The logic underlying recognition of these enclaves of diminished protection

simply does not apply here. In each of those situations, some unique and

10
Pennsylvania v. Labron, 518 U.S. 938, 940, 116 S. Ct. 2485, 2487, 135 L. Ed. 2d 1031
(1996) (noting that an “individual’s reduced expectation of privacy in an automobile” is a
justification for the automobile exception to the warrant requirement).
11
United States v. Robinson, 414 U.S. 218, 237, 94 S. Ct. 467, 477, 38 L. Ed. 2d 427
(1973) (Powell, J., concurring) (“[A]n individual lawfully subjected to a custodial arrest retains
no significant Fourth Amendment interest in the privacy of his person.”); see also Mincey v.
Arizona, 437 U.S. 385, 391, 98 S. Ct. 2408, 2413, 57 L. Ed. 2d 290 (1978) (accepting the
proposition that “one who is legally taken into custody has a lessened right of privacy in his
person”).
12
United States v. Montoya de Hernandez, 473 U.S. 531, 539, 105 S. Ct. 3304, 3309, 87
L. Ed. 2d 381 (1985) (noting that “the expectation of privacy [is] less at the border than in the
interior” of the country).
13
Oliver v.United States, 466 U.S. 170, 180, 104 S. Ct. 1735, 1742, 80 L. Ed. 2d 214
(1984) (“[T]he common law implies, as we reaffirm today, that no expectation of privacy
legitimately attaches to open fields”).
14
See, e.g., New York v. Class, 475 U.S. 106, 114, 106 S. Ct. 960, 966, 89 L. Ed. 2d 81
(1986) (“[B]ecause of the important role played by the VIN in the pervasive governmental
regulation of the automobile and the efforts by the Federal Government to ensure that the VIN is
placed in plain view, we hold that there was no reasonable expectation of privacy in the VIN.”)
15
Hudson v. Palmer, 468 U.S. 517, 526, 104 S. Ct. 2194, 3200, 82 L. Ed. 2d 393 (1984)
(“[S]ociety is not prepared to recognize as legitimate any subjective expectation of privacy that a
prisoner might have in his prison cell.”).

22
identifiable heightened state or diminished private interest was a catalytic

ingredient in the “reasonable expectations” formula. Because the Fourth

Amendment protects people rather than places, for example, individuals have

reduced privacy expectations in their automobiles and open fields. See Labron,

518 U.S. at 940, 116 S. Ct. at 2487 (noting that individuals have a reduced privacy

expectation in an automobile, “owing to its pervasive regulation”); Oliver, 466

U.S. at 180, 104 S. Ct. at 1742 (explaining that open fields do not share the

characteristics of the “sanctity of a man’s home and the privacies of life” (quoting

Boyd v. United States, 116 U.S. 616, 630, 6 S. Ct. 524, 532, 29 L. Ed. 746 (1886)

(quotation marks omitted))). Transborder travel and arrests are two contexts in

which compelling state interests may diminish legitimate privacy expectations.

See Montoya de Hernandez, 473 U.S. at 539, 105 S. Ct. at 3309 (“These cases

reflect longstanding concern for the protection of the integrity of the border. The

concern is, if anything, heightened by the veritable national crisis in law

enforcement caused by smuggling of illicit narcotics . . . .”); Robinson, 414 U.S. at

234, 94 S. Ct. at 476 (“The justification or reason for the authority to search

incident to a lawful arrest rests quite as much on the need to disarm the suspect in

order to take him into custody as it does on the need to preserve evidence on his

person for later use at trial.”).

23
 In this case, ordinary citizens seek to assemble and exercise free speech

rights in a public place. Under such circumstances, participants might have a

diminished privacy interest in their conversations. See Katz, 389 U.S. at 361, 88

S. Ct. at 516 (Harlan, J., concurring) (“[C]onversations in the open would not be

protected against being overheard, for the expectation of privacy under the

circumstances would be unreasonable.”). In their persons and property, however,

individuals “are not shorn of all Fourth Amendment protection when they step

from their homes onto the public sidewalks.” Delaware v. Prouse, 440 U.S. 662,

663, 99 S. Ct. 1391, 1401, 59 L. Ed. 2d 660 (1979). Nothing unique to this

gathering indicates that reasonable privacy expectations ought to diminish. It is a

large gathering, but the line-drawing problems that attend any attempt to decide

when a crowd is “too big” for Fourth Amendment purposes should be self-evident.

The SAW protest historically has been perhaps an unusually expressive gathering,

but it has been unblemished by violence. As a result, the individuals attending the

rally retained a legitimate expectation of privacy in their person and possessions.

Consequently, the main justification for doing away with a warrant requirement is

inapplicable to this case.

“Exigent circumstances” may also excuse the warrant requirement in some

cases, but not here. Such searches are permitted when “the inevitable delay

24
incident to obtaining a warrant must give way to an urgent need for immediate

action.” United States v. Satterfield, 743 F.2d 827, 844 (11th Cir. 1984). Even a

search under exigent circumstances must be supported by probable cause. See

United States v. Santa, 236 F.3d 662, 668 (11th Cir. 2000) (“A warrantless search

is allowed, however, where both probable cause and exigent circumstances

exist.”); United States v. Pantoja-Soto, 739 F.2d 1520, 1523 (11th Cir. 1983)

(“When exigent circumstances coexist with probable cause, the Fourth

Amendment has been held to permit warrantless searches and seizures.”). The

mass searches here were not. Moreover, a person walking to a rally hardly seems

like the type of “exigent circumstance” the Supreme Court had in mind in crafting

this doctrine. Consequently, the searches here are invalid for two reasons: they do

not fall within any recognized exception to the warrant requirement, and they were

not supported by probable cause, which is a requirement for most warrantless

searches.

Our holding here does not leave police without any means of promoting

public safety and detecting wrongdoing. First, if they have “reasonable suspicion”

that anyone is carrying a weapon, they may conduct an ordinary Terry stop. If

they have probable cause to believe anyone is carrying a weapon, they may either

conduct a full-fledged exigent-circumstances search, see United States v. Banshee,

25
91 F.3d 99, 101 (11th Cir. 1996) (“[B]ased upon the inconsistent statements and

the bulge in [the defendant’s] mid-section, we find that [the officer] had probable

cause to believe a search would uncover evidence of a crime. We also find that

there were exigent circumstances excusing the need for a warrant.” (citations

omitted)), or arrest the individual and conduct a search incident to arrest, id.

(“[T]he search could be considered a lawful search incident to an arrest.

Specifically, we find that the bulge in [the defendant’s] mid-section, coupled with

the inconsistent statements, were sufficient grounds for [the officer] to conclude

that [the defendant] was committing a crime.”). Armed with these alternatives, the

City cannot plausibly claim that it has no alternative but to accost innocent people

who show no indication of possessing contraband or weapons. We therefore find

that the mass, warrantless, suspicionless search policy violated the Fourth

Amendment.

III.

The City’s search policy also violates the First Amendment in five ways.

First, it is a burden on free speech and association imposed through the exercise of

a government official’s unbridled discretion; restrictions on First Amendment

rights may not be left to an executive agent’s uncabined judgment. Second, the

searches were a form of prior restraint on speech and assembly; to participate in

26
the protest, individuals had to receive the prior permission of officers manning the

checkpoints. Third, the search policy was implemented based on the content of

the protestors’ speech. Fourth, even assuming the searches were implemented

exclusively for content-neutral reasons, they were impermissible because they did

not constitute reasonable time, place, and manner limitations, which are the only

permissible content-neutral burdens that may be placed upon free speech and

association. Finally, even putting aside First Amendment analysis, the search

policy constitutes an “unconstitutional condition;” protestors were required to

surrender their Fourth Amendment rights (as discussed in Part II) in order to

exercise their First Amendment rights. We explore each of these deficiencies in

turn.

A.

The first crucial problem with the City’s decision to search the protestors

entering the SAW protest area is that it was not made according to any set,

objective, neutral criteria. The Supreme Court has repeatedly held that burdens on

First Amendment rights imposed by executive officials are invalid “in the absence

of narrowly drawn, reasonable and definite standards for the officials to

follow . . . . [A restriction is invalid when] [n]o standards appear anywhere; no

narrowly drawn limitations; no circumscribing of this absolute power.” Neimotko

27
v. Maryland, 340 U.S. 268, 271, 71 S. Ct. 325, 327, 95 L. Ed. 2d 267 (1951).

Even if a particular restriction or condition is an otherwise permissible content-

neutral regulation of the time, place, or manner of speech, it is unconstitutional if a

government official has unbridled discretion to apply it. See Burk v. Augusta-

Richmond County, 365 F.3d 1247, 1256 (11th Cir. 2004) (“Even a facially

content-neutral time, place, and manner regulation may not vest public officials

with unbridled discretion over permitting decisions.”).

Although this doctrine originated with cases involving grants of power to

executive officials to determine whether or not to grant licenses to engage in

expression at all, see Lovell v. Griffin, 303 U.S. 444, 58 S. Ct. 666, 82 L. Ed. 2d

949 (1938) (invalidating an ordinance requiring individuals to obtain permits

before circulating leaflets), it has subsequently been held to apply to a wider range

of burdens on expression. For example, in Secretary of State of Maryland v.

Joseph H. Munson, Co., 467 U.S. 947, 104 S. Ct. 2839, 81 L. Ed. 2d 786 (1984),

the Supreme Court considered a local ordinance that permitted door-to-door

solicitation for charitable causes, but empowered a local official to burden such

speech by granting him complete discretion over whether to require a solicitor to

certify that 75% or more of collected funds actually went to charitable purposes

rather than administrative costs. One of the reasons (albeit not the primary one)

28
the Court invalidated the ordinance was that it “place[d] discretion in the hands of

an official to grant or deny a license [based on the exemption],” thereby creating a

“threat of censorship that by its very existence chills speech.” Id. at 964 n.12, 104

S. Ct. at 2850 n.12. Similarly, in Lakewood v. Plain Dealer Pub. Co., 486 U.S.

750, 108 S. Ct. 2138, 100 L. Ed. 2d 771 (1988), the Supreme Court addressed an

ordinance that permitted the distribution of newspapers, but empowered the Mayor

to burden such expression by granting him complete discretion over whether

newsstands could be used to aid the process. The Court invalidated the ordinance,

holding, “[I]n the area of free expression a licensing statute placing unbridled

discretion in the hands of a government official or agency constitutes a prior

restraint and may result in censorship.” Id. at 757, 108 S. Ct. at 2144.

Along these same lines, in Burk, we struck down an ordinance giving a

county attorney unbridled discretion in determining whether speakers must

indemnify the county against damages caused by their activities. 365 F.3d at 1256

(“We readily conclude that the indemnification provision in the . . . [o]rdinance

fails to provide adequate standards [for the county attorney to follow].”); see also

Saia v. New York, 334 U.S. 558, 68 S. Ct. 1148, 92 L. Ed. 1574 (1948)

(invalidating an ordinance that permitted political communications, but gave an

executive official unbridled discretion to burden them by deciding whether to

29
permit certain candidates to use loudspeakers).

In this case, the City’s search policy unquestionably placed a burden on the

speech and assembly rights of individuals wishing to participate in SAW’s protest

against the SOA. Protestors were inconvenienced by having to wait in line to be

searched in order to enter the protest area. They were then subjected to a

magnetometer, and possibly physical, search of their person and possessions.

Those who refused to submit to magnetometer searches were prevented from

speaking altogether.

The decision to implement this search policy was an exercise of the

apparently unbridled discretion of Police Chief W.L. Dozier. The City does not

provide us with any generally applicable state laws or local ordinances that set

forth the circumstances under which individuals attending a large public gathering

will be searched. The decision to implement searches instead appears to be left to

the Chief’s personal discretion, to be based on whatever factors he deems

appropriate at any given point in time. The First Amendment does not permit the

government to place burdens on speech and assembly in such an unprincipled, ad

hoc manner.

The Chief contends that he decided to search the protestors because he felt

there was probable cause to believe violence might erupt. SAW’s unblemished

30
thirteen-year history of nonviolence at its protests forces us to question this self-

serving claim. Even taking his explanation at face value, however, it entirely

misses the point. The problem is not that the Chief applied an inappropriate

standard in deciding whether to implement this search policy. Instead, the

problem is that there were no objective, established standards for the Chief to

utilize in making this decision other than those he happened to deem relevant. In

this respect, this case is quite similar to Forsyth County v. Nationalist Movement,

505 U.S. 123, 112 S. Ct. 2395, 120 L. Ed. 2d 101 (1992), in which the Supreme

Court invalidated an ordinance that permitted a county official to decide how

much to charge for a permit to hold a public gathering. The Court noted,

The decision how much to charge for police protection
or administrative time—or even whether to charge at
all—is left to the whim of the administrator. There are no
articulated standards either in the ordinance or in the
county’s established practice. The administrator is not
required to rely on any objective factors. He need not
provide any explanation for his decision, and that
decision is unreviewable.

Id. at 133, 112 S. Ct. at 2403. Because there are no established standards, nothing

prevents the Chief from applying one standard to the SAW protest and an entirely

different standard to other public gatherings (including those sponsored by

organizations with which he might be more sympathetic). Cf. id. (“Nothing in the

31
law or its application prevents the official from encouraging some views and

discouraging others through the arbitrary application of fees.”).

We emphasize that, in establishing such a general policy for determining the

specific occasions on which mass searches may be implemented, legislatures or

municipal governing bodies must establish specific criteria susceptible to judicial

review. They may not simply craft ordinances permitting mass searches “when

public safety so requires” or “when the Chief shall deem it advisable.” Such

general language does not meaningfully constrain the Chief’s discretion. Instead,

the ordinance must include specific, non-conclusory factors to guide these

determinations. See Neimotko, 340 U.S. at 271, 71 S. Ct. at 327 (holding that

when executive officials are given discretion in implementing restrictions on

speech, there must be “narrowly drawn, reasonable and definite standards for the

officials to follow”); see, e.g., Shuttlesworth v. Birmingham, 394 U.S. 147, 150,

89 S. Ct. 935, 938, 22 L. Ed. 2d 162 (1969) (holding that an ordinance directing

public officials to grant licenses where the proposed expression is consistent with

“public welfare, peace, safety, health, decency, good order, morals or

convenience” is too amorphous meaningfully to cabin their discretion).

Even if this search would have been constitutional had it been conducted

32
under a general policy that appropriately limited the Chief’s discretion,16 the

Chief’s unilateral, unprincipled decision to implement the search outside the

context of such a general policy would still have been improper. See Miami

Herald Pub. Co. v. Hallandale, 734 F.2d 666, 674 (11th Cir. 1984) (“[I]n the

unique context of first amendment challenges upon the facial validity of licensing

statutes, it is the very existence of official discretion that gives rise to a threat of

injury sufficient to warrant an injunction.”). In the absence of objective criteria

establishing when individuals attending large public gatherings may be searched,

the ad hoc imposition of such a burden on free expression and assembly in this

case must be invalidated.

B.

Even putting aside the question of the Chief’s unlimited discretion in

deciding whether to require mass searches at the SAW protest, the searches are

still unconstitutional because they constitute a prior restraint on free speech and

assembly. “A prior restraint of expression exists when the government can deny

access to a forum before the expression occurs.” United States v. Frandsen, 212

F.3d 1231, 1236-37 (11th Cir. 2000). To enter SAW’s protest site, individuals

16
For the reasons discussed in the following Subparts, the searches in this case would
still have been unconstitutional, even if the Chief’s discretion in implementing the policy had
been appropriately cabined.

33
must be waived through a checkpoint by a police officer. Officers are authorized

to prevent people from assembling at the protest site and engaging in expression if

they refuse to submit to a search or possess certain otherwise legal items that have

been prohibited from the protest area. Because individuals must, in effect, receive

the permission of police officers to enter the area of the protest and to exercise

their rights to freedom of speech and assembly, the search policy establishes a

prior restraint on protected expression. Cf. id. at 1237 (“Because the

superintendent can deny the use of the park for expression by denying a permit,

[the law at issue] is a prior restraint on expression.”).

Because the searches constitute prior restraints, “there is a strong

presumption against their constitutionality.” Id. To be valid, a prior restraint

must, at the very least, provide constitutionally adequate procedural safeguards for

potential speakers. The policy here fails to provide such safeguards.

In Frandsen, we held, perhaps ambiguously, that under Supreme Court

precedent, prior restraints must satisfy “at least some” of these requirements:

(1) the burden of going to court to suppress the speech,
and the burden of proof once in court, must rest with the
government; (2) any restraint prior to a judicial
determination may only be for a specified brief time
period in order to preserve the status quo; and (3) an
avenue for prompt judicial review of the censor’s
decision must be available.

34
Id. Meaningful judicial review is the touchstone of the test. “[P]rompt judicial

review must be available to correct erroneous denials” of access to expression.

Café Erotica of Fla., Inc. v. St. John’s County, 360 F.3d 1274, 1283 (11th Cir.

2004). Here, however, there was no available route through which protestors

excluded at the checkpoints from the protest site could seek meaningful judicial

relief prior to the end of the protest. Furthermore, the City never availed itself of

the courts in seeking initially to suppress the speech, nor did it limit its speech

restraint for a “brief time period” pending judicial review. As a result, none of the

Frandsen factors are satisfied, placing this prior restraint plainly outside of what is

permissible under the First Amendment.

C.

Even assuming the searches did not constitute a “prior restraint,” they were

still invalid. The Supreme Court has recognized two categories of restrictions on

expression—content-based and content-neutral. The district court found that the

decision to institute searches was content-neutral; that is, it was motivated by a

legitimate governmental concern other than disagreement with the message

conveyed. In support of this conclusion, the court offered only the following

observation—“[T]he fact that everyone will be searched in the same manner

35
indicates that the restriction is content neutral.” This is a non sequitur; the fact

that all the protestors were searched does not suggest that the decision to search

them was content-neutral; it suggests only that the City treated each SOA protestor

equally vis-a-vis the other SOA protestors. The decision to search all the SOA

protestors but not other persons or groups is entirely consistent with the notion

that the City targeted them precisely because of the message they were sending.

Under the Supreme Court’s holding in Forsyth County v. Nationalist

Movement, the decision to institute the mass search policy was content-based. In

Nationalist Movement, Forsyth County enacted a policy requiring individuals

seeking permits for public gatherings to pay a fee to defray the cost of law

enforcement and the county administrator’s time in processing the permit. The

county administrator was permitted to determine the amount of the fee, based on

the amount of law enforcement assistance he anticipated would be necessary to

maintain order. The Court observed:

In order to assess accurately the cost of security for
parade participants, the administrator must necessarily
examine the content of the message that is conveyed,
estimate the response of others to that content, and judge
the number of police necessary to meet that response.
The fee assessed will depend on the administrator’s
measure of the amount of hostility likely to be created by
the speech based on its content. Those wishing to
express views unpopular with bottle throwers, for

36
 example, may have to pay more for their permit.
. . . . [I]t cannot be said that the fee’s justification
has nothing to do with content.

505 U.S. at 134, 112 S. Ct. at 2403 (quotation marks and citations omitted).

In this case, the Chief conducted a somewhat similar analysis in deciding to

require magnetometer searches. Apparently anticipating how other organizations

such as the Anarchists would react, the Chief infringed upon SAW’s right to

protest by requiring magnetometer searches. Under Nationalist Movement, when

a government official decides that certain expressive activity will lead others to

break the law, he is making a content-based distinction. Cf. R.A.V. v. City of St.

Paul, 505 U.S. 377, 391-92, 112 S. Ct. 2538, 2547-48, 120 L. Ed. 2d 305 (1992)

(striking down as content-based a local ordinance that banned only fighting words

that targeted specific minority groups, even though such fighting words might well

have produced the type of violence the city sought to prevent). Having concluded

that the magnetometer searches are content-based, we must next determine

whether the City’s policy is a permissible content-based regulation of speech.

Content-based restrictions on speech in a public forum are subject to strict

scrutiny, which means that we must “ascertain whether [the policy] employs the

least restrictive means to meet a compelling government interest.” Burk, 365 F.3d

at 1255 (citing United States v. Playboy Entm’t Group, Inc., 529 U.S. 803, 813,

37
120 S. Ct. 1878, 1886, 146 L. Ed. 2d 865 (2000)).17 Neither side disputes that the

protest takes place in a public forum, though none of the briefs indicate with any

specificity the situs of the demonstrations. Bourgeois indicates only that the

“large, non-violent demonstration and memorial procession [take place] outside

the front gate of Fort Benning.” Appellants’ Brief at 2. It appears from

uncontradicted testimony on the record, however, that the demonstrations take

place on some combination of public roads and parks, the use of which SAW

secured by permit. V2-R15-54. We assume it is this area to which Bourgeois

refers in his brief. See Appellants’ Brief at 23-24 (“[S]treets are natural and

proper places for the dissemination of information [. . . .]” (quoting Schneider v.

State, 308 U.S. 147, 163, 60 S. Ct. 146, 151, 84 L. Ed. 155 (1939) (quotation

marks omitted))).18 Thus, we proceed on the assumption that strict scrutiny

17
The City asserts that it need only establish a “significant” government interest to which
its policy is narrowly tailored. Appellees’ Brief at 38-39. This argument assumes, however, that
the ordinance is a content-neutral regulation, which it is not. Although the City’s brief describes
as merely “significant” the government’s purported interest in “maintain[ing] public safety,
security, and order, which includes physical protection of demonstration participants, spectators,
passers-by, and law enforcement,” Appellees’ Brief at 41 (quoting the district court), we take the
City to be arguing that this interest is not just “significant,” but “compelling.” See Appellees’
Brief at 38-39 (“[T]he actual test is whether Columbus had a ‘significant’ government interest.
However, the evidence of record supports not only a ‘significant’ interest, but also supports a
heightened finding of a ‘compelling’ interest.”).
18
We recognize that we determined in another case involving protests of the School of
Americas that Fort Benning itself is a nonpublic forum. United States v. Corrigan, 144 F.3d 763,
767 (11th Cir. 1998). That case did not purport to rule on areas surrounding the Fort and thus
does not bind us here. The City does not claim that its purpose in administering magnetometer
searches is to burden speech only in the Fort itself, nor could it, for the policy is manifestly not

38
applies, and we begin by identifying the interest in the policy.

The asserted government interest here is in “maintain[ing] public safety,

security, and order” for the protection of participants, law enforcement, and

innocent bystanders. Appellees’ Brief at 41. Here again the City asserts that the

post-September 11 environment further substantiates the government’s interest.

Appellees’ Brief at 41 (“First, there is a new and real threat of terrorist attack

[that] did not exist previously. This threat is so authentic that the President . . .

created the Department of Homeland Security. This Department has requested

[that] local governments implement precautionary measures to assist in the

prevention and detection of terrorist attack.”). Although this argument was never

before the district court, the City asserts that “this fact is one of which the Court

can take judicial notice.” Appellees’ Brief at 41 n.15 (citing Fed. R. Evid. 201).

The City then somewhat tautologically posits that magnetometer searches are

narrowly tailored because they are the “least intrusive manner of removing the

threat of knives, bombs, guns, and incendiary devices.” Appellees’ Brief at 43.

In fact, magnetometer searches do not seem narrowly tailored at all to the

City’s professed interest in maintaining safety. As discussed in Part II, there are

other ways of ensuring public safety at this event, and the availability of

tailored to such a purpose.

39
alternatives casts serious doubt on any narrow tailoring analysis. See Carey v.

Brown, 447 U.S. 455, 465, 100 S. Ct. 2286, 2292, 65 L. Ed. 2d 263 (1980)

(“[C]ertain state interests may be so compelling that where no adequate

alternatives exist a content-based distinction—if narrowly drawn—would be a

permissible way of furthering those objectives . . . .” (citing Schenck v. United

States, 249 U.S. 47, 39 S. Ct. 247, 63 L. Ed. 470 (1919))). Furthermore, the list

the City provides of problems historically posed by the SAW protests includes a

number of incidents against which magnetometer searches would provide little

utility, suggesting underinclusivity. See Appellees’ Brief at 42-43 (noting that

affinity groups will attend the protest, but failing to indicate any proclivities of

such groups to bring metallic objects to protests; that each year violence and

illegal activity has occurred, but failing to indicate that even a single instance of

such activity involved the use of a metallic object; that a large crowd will be in

attendance, but failing to show any meaningful relationship between the size of a

crowd and the likelihood that anyone would carry metallic objects to such a

crowd; that the 2001 protest included an effort to “gather[] items and [an]

attempt[] to ignite a large fire,” but failing to explain how a metal detector could

prevent flammable (and therefore presumably nonmetallic) items from being

brought into the protest; and that attendees once brought feces and blood to the

40
protest, but failing to show that a magnetometer could detect these “weapons”). In

addition, it does not strain belief to imagine that a magnetometer search could be

triggered by a number of objects that have no combative or militaristic utility. In

any event, the magnetometer searches are, at a minimum, substantially

underinclusive to meet the government’s purported interest in public safety.

The City’s policy is underinclusive along another dimension as well.

Bourgeois points out that the magnetometer search policy has not been

implemented for any other large gatherings the City has faced. In particular,

sporting events—at which large crowds gather; where authorities have no way of

knowing who is coming; at which “affinity groups” of various sorts are regularly

in attendance; where “violent acts had in fact occurred,” Appellants’ Brief at 33;

and which confine large numbers of people into a compact area (presumably even

more so than a protest in a public park)—would seem to implicate the

government’s purported interest even more gravely than the SAW protests. Thus,

the City’s apparently arbitrary application of its magnetometer search policy from

one type of event to the next underscores both our “unbridled discretion” concerns

outlined above and our doubts here about whether the policy is narrowly tailored

to any kind of government interest, whether compelling or even simply

“significant.”

41
 Because we hold that the policy here is not narrowly tailored to serve the

government’s interest, we need not address the question whether the interest is

compelling. It bears mentioning here, however, that the City’s attempt to employ

terrorism to bolster its alleged interest in the search policy, without having

developed a record at the district court level indicating that concerns about

terrorism motivated the adoption of that policy, is unavailing. The City asks us to

take “judicial notice” that the Department of Homeland Security has asked local

governments to “implement precautionary measures” to sniff out terrorist threats.

Appellees’ Brief at 41 & n.15. The question, however, is not whether we may or

may not take notice; it is whether the City actually took this directive into

consideration when drafting its policy. Otherwise, the City merely invites us to

engage in post hoc rationalizations of its policy, which is precisely one of the

dangers that attaches to the sort of uncabined, impulsive policymaking practice at

issue in this case. Cf. Plain Dealer, 486 U.S. at 758, 108 S. Ct. at 2144 (“Without .

. . guideposts, post hoc rationalizations by the licensing official and the use of

shifting or illegitimate criteria are far too easy, making it difficult for courts to

determine in any particular case whether the licensor is permitting favorable, and

suppressing unfavorable, expression.”). The fact that the City cannot even decide

from one stage of litigation to the next what precisely was the basis for its decision

42
to search protestors precludes a determination that the policy was in any way

“narrowly tailored.” Thus, the policy here is an impermissible content-based

regulation of speech.

D.

Even if we were to accept that the searches were entirely motivated by

content-neutral concerns, they still violate the First Amendment. Content-neutral

restrictions are permissible so long as they amount to “reasonable time, place, and

manner restrictions” on speech. Clark v. Community for Creative Non-Violence,

468 U.S. 288, 293, 104 S. Ct. 3065, 3069, 82 L. Ed. 2d 221 (1984) (“Expression,

whether oral or written or symbolized by conduct, is subject to reasonable time,

place, or manner restrictions. . . . [R]estrictions of this kind are valid provided that

they are justified without reference to the content of the regulated speech.”).

Because these searches were not a form of “time, place, or manner restriction,”

they were an impermissible burden on the protestors’ First Amendment rights.

Here, the City contends that the search was a permissible restriction on the

manner of the protest. The City might be correct to suppose that a mere ban on the

use of weapons or incendiary devices at the protest would withstand a “reasonable

manner restriction” analysis. Examples of “manner” restrictions that have been

upheld by the Supreme Court include bans on certain types of amplification

43
equipment or requirements that speakers do not exceed a particular volume. Ward

v. Rock Against Racism, 491 U.S. 781, 109 S. Ct. 2746, 105 L. Ed. 2d 661 (1989)

(upholding restrictions on the volume of a concert in a public bandshell), Kovacs

v. Cooper, 336 U.S. 77, 69 S. Ct. 448, 453, 93 L. Ed. 513 (1949) (“We think it is a

permissible exercise of legislative discretion to bar sound trucks with broadcasts

of public interest, amplified to a loud and raucous volume, from the public ways of

municipalities.”).

Simply identifying a manner of speech that legitimately may be restricted is

not the end of the matter, however. We have said that any restriction must be

“reasonable,” meaning “not the anemic simulacrum of a constraint on

governmental power found in the Due Process Clause’s ‘rational basis’ test, but

rather a more robust notion of ‘reasonableness’ such as that applied in the Fourth

Amendment context.” Holloman ex rel Holloman v. Harland, 370 F.3d 1252,

1271 (11th Cir. 2004) (citations omitted). Specifically, a reasonable restriction on

a manner of speech “must leave open alternative channels of communication and .

. . not restrict substantially more speech than necessary to further a legitimate

government interest.” Burk, 365 F.3d at 1251.

An antecedent mass search for weapons or incendiary devices goes far

beyond merely regulating the manner in which the protest is conducted. Such a

44
search is, instead, a prophylactic measure designed to implement and enforce an

otherwise permissible “manner” restriction and is conceptually distinct from the

underlying restriction itself. Under the City’s policy, every person who wishes to

speak by taking part in the SAW protest is burdened by a magnetometer search.

The scope of this far-reaching policy alone would pose difficult First Amendment

problems without the additional fact that, as noted above, the magnetometer

searches will fail to catch a number of additional dangers that implicate the City’s

alleged interest. Because the search policy thus “restrict[s] substantially more

speech than necessary to further a legitimate government interest,” id., it is

impermissible.

E.

The City may contend that the searches are permissible because they are

entirely voluntary. No protestors are compelled to submit to searches; they must

do so only if they choose to participate in the protest against the SOA. This is a

classic “unconstitutional condition,” in which the government conditions receipt

of a benefit or privilege on the relinquishment of a constitutional right.

See Adams v. James, 784 F.2d 1077, 1080 (11th Cir. 1986) (“The doctrine of

unconstitutional conditions prohibits terminating benefits, though not classified as

entitlements, if the termination is based on motivations that other constitutional

45
provisions proscribe.”). This case presents an especially malignant

unconstitutional condition because citizens are being required to surrender a

constitutional right—freedom from unreasonable searches and seizures—not

merely to receive a discretionary benefit but to exercise two other fundamental

rights—freedom of speech and assembly.

Our circuit has roundly condemned the use of unconstitutional conditions.

See, e.g., Bertrand v. United States, 467 F.2d 901, 902 (5th Cir. 1972) (ordering

resentencing due to the trial judge’s questioning of the defendant because “[t]he

effect of the trial judge’s questioning was to impose an unconstitutional condition

on the petitioner’s Fifth Amendment rights: he could go into the details of the

other offense . . . that might constitute a confession or he could exercise his right

to be silent and receive a long sentence”);19 Boykins v. Fairfield, 370 F.2d 847,

851 (5th Cir. 1967) (“Even if the school board were under no obligation to provide

public education to children of military personnel on the air base, it could not

provide that education subject to an unconstitutional condition.”). We are

especially careful to ensure that the government does not use unconstitutional

conditions to chill speech. See, e.g., Leary v. United States, 431 F.2d 85, 89 (5th

19
In Bonner v. Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981) (en banc), we adopted as
binding precedent all decisions of the former Fifth Circuit rendered prior to October 1, 1981.

46
Cir. 1970) (“[I]f the appellant’s eligibility to be enlarged on bail under the Eighth

Amendment may be lost because he exercises his first amendment right to freedom

of speech and to freedom of the press, then [the law at issue] imposes an

unconstitutional condition.”).

The ability of protestors to avoid the searches by declining to participate in

the protest does not alleviate the constitutional infirmity of the City’s search

policy; indeed, the very purpose of the unconstitutional conditions doctrine is to

prevent the Government from subtly pressuring citizens, whether purposely or

inadvertently, into surrendering their rights. Similarly, the existence of other

vehicles through which protestors could voice their disagreement with the SOA

(e.g., letters to Congress) does not in any way alleviate the unconstitutional

conditions problem. “The applicability of the unconstitutional conditions doctrine

does not turn on whether conferral of the discretionary benefit is conditioned upon

completely foregoing the right to engage in expression or instead upon foregoing

the right to engage in that expression in certain places or manners or at certain

times.” Sammy’s, Ltd. v. City of Mobile, 140 F.3d 993, 1000 n.2 (11th Cir. 1998).

The one situation in which the doctrine applies with slightly reduced force, public

employment, is inapplicable here. See Pickering v. Bd. of Educ., 391 U.S. 563,

568, 88 S. Ct. 1731, 1734, 20 L. Ed. 2d 811 (1968) (“[T]he State has interests as

47
an employer in regulating the speech of its employees that differ significantly from

those it possesses in connection with regulation of the speech of the citizenry in

general.”).

As discussed in Part II, the searches violated the Fourth Amendment. The

speech and assembly in which the protestors sought to engage were

unquestionably protected by the First Amendment. Consequently, the City’s

decision to require protestors to waive their Fourth Amendment rights and submit

to searches in order to exercise their First Amendment rights was an

unconstitutional condition that violated the First Amendment.

IV.

The City's search policy violated both the First and Fourth Amendments to

the United States Constitution. The plaintiffs are entitled to a permanent

injunction against its implementation. The judgment of the district court is

VACATED and the case is REMANDED for the entry of appropriate injunctive

relief.

SO ORDERED.

48