The “Ample Alternative Channels” Flaw in First Amendment Doctrine
The “ Ample Alternative Channels” Flaw in First Amendment Doctrine
0 Thi s Article is brought to you for free and open access by the Washington and Lee Law Review at Washington & Lee University School of Law Scholarly Commons. It has been accepted for inclusion in Washington and Lee Law Review by an authorized editor of Washington & Lee University School of Law Scholarly Commons. For more information , please contact , USA
1 Enrique Armijo, Th e “Ample Alternative Channels” Flaw in First Amendment Doctrine , 73 Wash. & Lee
2 Elon University School of Law
The “Ample Alternative Channels”
Flaw in First Amendment Doctrine
In reviewing a content-neutral regulation affecting speech,
courts ask if the regulation leaves open “ample alternative channels
of communication” for the restricted speaker’s expression.
Substitutability is the underlying rationale. If the message could
have been expressed in some other legal way, the ample alternative
channels requirement is met. The court then deems the restriction’s
harm to the speaker’s expressive right as de minimis and upholds
the law. For decades, courts and free speech scholars have assumed
the validity of this principle. It has set First Amendment
jurisprudence on the wrong course.
Permitting a speech restriction because the speaker could have
communicated the same message another way distorts the First
Amendment. Ample alternative channels analysis instructs courts
to engage in counterfactual, post-hoc reasoning as to the expressive
choices the speaker could have made, but didn’t—i.e., to substitute
the court’s own value judgments for those of the speaker’s. The
modern communications world expands the doctrine’s pernicious
effects, since speech-facilitating technologies can always
theoretically grant an alternative means of expression to any
infringed speaker. And the origin of the doctrine, from Justice
Harlan’s concurrence in United States v. O’Brien, shows that
Associate Dean for Academic Affairs and Associate Professor, Elon
University School of Law and Affiliated Fellow, Yale Law School Information
Society Project. Thanks to colleagues at the Yale Information Society Project’s
Freedom of Expression Scholars’ 2015 Conference and the Loyola Chicago 2015
Constitutional Law Colloquium. Individual thanks go to Jack Balkin, Joseph
Blocher, David Han, Chad Golder, Randy Kozel, Kerry Monroe, Alex Tsesis, Mark
Tushnet, and Tinsley Yarbrough, to the librarians at Princeton University’s Mudd
Rare Manuscript Library for providing access to the John Marshall Harlan
Papers Collection, and to Spenser Tatum, Elizabeth Long, and Britney Boles for
ample alternative channels analysis was in its incipiency a
misguided afterthought—born, as historical Supreme Court case
files never examined before this Article show, as literally a margin
note to an unpublished draft.
In the place of ample alternative channels analysis, courts
should ask whether a speaker’s chosen mode is incompatible with
the government’s interest in the restriction in question. An
incompatibility rule would be more consistent with the Roberts
Court’s turn toward reviewing content-neutral speech restrictions rigorously, as evidenced in 2014’s McCullen v. Coakley. Table of Contents
I. Introduction ...................................................................1659
You are a Hare Krishna. As part of your spiritual obligation,
you must engage in Sankirtan, which requires, in addition to
public singing and chanting of mantras, the hand-to-hand
dissemination of religious literature and face-to-face solicitation of
donations for the church.1 These interactions between devotees
and potential recruits in public spaces are critical to the growth of
your chosen faith.2
Your community’s largest public gathering of the year, and
thus your best opportunity to proselytize to thousands of potential
new recruits to Krishna Consciousness, is the state fair.3 But the
fair’s vendor solicitation rules require you, or any other person or
organization seeking to distribute materials to fairgoers, to do so
only from behind a booth that is assigned to a fixed location within
the fairgrounds chosen by the fair’s organizers.
Note what the fair’s requirements do not bar you from doing.
They do not ban you from the fairgrounds altogether, and it
permits you to interact with fellow fairgoers who approach your
1. See E. Burke Rochford Jr., Recruitment Strategies, Ideology, and
Organization in the Hare Krishna Movement, 29 SOCIAL PROBS. 399, 401 (1982)
(noting that encounters in public places are an important way of recruiting new
members to the growth of the International Society for Krishna Consciousness);
see also ISKCON of Potomac, Inc. v. Kennedy, 61 F.3d 949, 952–53 (D.C. Cir.
1995) (describing a “Krishnafest” held by the International Society for Krishna
Consciousness of the Potomac).
2. See Rochford, supra note 1, at 401–02; (“Table 1 shows [that] 42 percent
[of ISKCON devotees] made contact in public places.”); see also E. Burke Rochford,
Jr., A Study of Recruitment and Transformation Processes in the Hare Krishna
Movement 19–20 (Apr. 1, 1982) (unpublished Ph.D. dissertation, University of
California, Los Angeles) (observing that, unlike other new religious movements,
in which preexisting “social network ties have played a prominent role in the[ir]
expansion . . . persons recruited into Krishna Consciousness most often made
their initial contact with the movement in public place encounters with Krishna
devotees.”) (on file with the author).
3. These facts are taken from Heffron v. Int’l Soc. for Krishna
Consciousness, 452 U.S. 640, 655 (1981).
assigned booth. You challenge the fair’s requirements on First
Amendment grounds. Should you prevail?
Now consider a second hypothetical. You own a piece of real
estate that you wish to sell.4 To attract interest in the property
from prospective buyers, you want to post a “For Sale” sign in the
home’s front lawn. However, your town bars owners from placing
sales-related signage on the lawns of their homes. Once again, note
what the restriction does not bar you doing: you may list your
property for sale with agents; you may place informational flyers
in a dedicated box on the lawn that passersby can help themselves
to (so long as the box is labeled “Free Information,” and not “For
Sale”); you may place your sign in the home’s windows (though
again, not the yard); craigslist, realtor.com, and the rest of the
Internet are available to you; and you retain the ability to show the
home to prospective buyers at times of mutual convenience.
You challenge the township’s signage ban on First
Amendment grounds. Should the aforementioned facts be
relevant? In other words, should the ban survive your challenge
because you are able communicate that the home is on the market
by other means despite the township’s restriction—even though
you, the restricted speaker, favor using a “For Sale” sign over any
of those means?
Alternatively, assume instead that the sign you wish to place
in your lawn expresses your opposition to the Iraq War.5 This time,
your township bars all yard signs except for “For Sale” signs.6 Is
the town’s abridgement of your speech cured by your ability to
express your distaste for the War through a range of other
constitutionally protected manners of expression, from picketing
and handing out antiwar flyers on your front lawn to placing
bumper stickers all over your car (parked right outside your
house)—expression that the township’s signage ban does not
implicate in any way?
These three hypotheticals, all based on actual cases, involve
different kinds of speech—religious; commercial; political. The
4. These facts are taken from Linmark Assocs., Inc. v. Willingboro, 431 U.S.
85, 98 (1977).
5. These facts are taken from City of Ladue v. Gilleo, 512 U.S. 43, 58–59
6. Id. at 45.
government interests underlying the regulations abridging that
speech vary widely as well—patron enjoyment of the state fair;
preventing white flight; aesthetic choices regarding visual clutter.7
But in all three cases, in deciding whether the restriction in
question abridged the speaker’s First Amendment right, the U.S.
Supreme Court considered the relative effectiveness of alternative
modes of expression that the speaker did not use—and, for all we
or the Court know, that the speaker may in fact have chosen not to
Current First Amendment doctrine finds that if ample
alternative channels of expression exist for a speaker to express
her views, then a content-neutral regulation foreclosing the
speaker’s chosen channel of expression will survive review.8 This
Article’s fundamental premise is that such a finding is at odds with
the First Amendment itself.
Giving the availability of alternative communication channels
dispositive significance in speech cases undermines speakers’
communicative choices with respect to their speech’s audience,
effectiveness, and reach—choices that both self-autonomy and
marketplace theory teach deserve constitutional respect.9 The
doctrine calls on judges to substitute their First Amendment
values for the restricted speaker’s; to engage in counterfactual,
post-hoc reasoning as to the choices the speaker could have made,
but did not; and to focus on such hypotheticals at the expense of
the relevant inquiry: whether the state has adequately justified its
7. See Heffron, 452 U.S. at 656–57 (justifying the state fair’s rule in the
interest of maintaining orderly movement of individuals at the fair); Linmark
Assocs., Inc., 431 U.S. at 94 (“[T]he vital goal this ordinance serves [is] namely,
promoting stable, racially integrated housing.”); Gilleo, 512 U.S. at 49 (“Ladue
relies squarely on that [esthetic values] content-neutral justification for its
8. See Hill v. Colorado, 530 U.S. 703, 726 (2000) (“[W]hen a content-neutral
regulation does not entirely foreclose any means of communication, it may satisfy
the tailoring requirement even though it is not the least restrictive or least
intrusive means of serving the statutory goal.”).
9. See id. at 751 (Scalia, J., dissenting) (stating that the “right to be let
alone” contemplated by Justice Brandeis in Olmstead v. United States is the right
of a speaker in a public forum to be free from governmental intrusion).
interference with speech.10 This approach leads to speech-averse
results in a range of cases.11 It is also a constitutional anomaly.
In no other area of constitutional law do courts excuse
government interferences with protected rights on the grounds
that despite the interference at issue, the affected party could have
exercised that same right as effectively in a different way. For
example, when a college student is denied admission to their public
university of choice on account of their race, no court asks whether
the student could have been admitted to another comparable
school (let alone whether the student did in fact apply and was
admitted to such a school), and if so, whether that fact minimizes
the harm caused by violating the student’s right to equal
protection.12 A city could not successfully defend a ban on firing
ranges against a Second Amendment challenge on the ground that
ranges are available in a jurisdiction nearby.13 And after last
term’s Obergefell v. Hodges,14 which held that the right to
samesex marriage is fundamental, 15 no state could save its ban on such
marriages by arguing that it permits same-sex civil unions, which
10. See Geoffrey R. Stone, Content-Neutral Restrictions, 54 U. CHI. L. REV.
46, 50 (1987) (“The Court does not seriously inquire into the substantiality of the
governmental interest, and it does not seriously examine the alternative means
by which the government could achieve its objectives. As a result, when the Court
applies this standard, it invariably upholds the challenged restriction.”).
11. Infra Part II.B.1.
12. These facts may at most raise a standing question, but even that issue is
far from clear. See, e.g., Fisher v. Univ. of Tex., 758 F.3d 633, 662 (5th Cir. 2014),
aff’d, 136 S. Ct. 2198, 2207 (2016) (arguing that nonadmitted student lacked
standing to challenge constitutionality of public university admissions decision
because the student had been admitted and subsequently graduated from another
Moreover, the Court has expressly rejected the argument that a state can
cure a race-based denial of a student’s university admission by providing a
substitute as an alternative. See Missouri ex. rel. Gaines v. Canada, 305 U.S. 337,
351–52 (1938) (finding that attendance at a university of any adjacent state for
law school is not a valid alternative for attendance at in-state institution that will
not accept black students).
13. See Ezell v. City of Chicago, 651 F.3d 684, 697, 709–10 (7th Cir. 2011)
(finding that Chicago had not established a strong enough public-interest
rationale for its ban on firing ranges and thus this plan likely violated the
plaintiffs’ Second Amendment rights).
15. See id. at 2607 (“The Court, in this decision, holds same-sex couples may
exercise the fundamental right to marry in all States.”).
provide all of the benefits of marital status under that state’s
substantive law, and thus the ban does not offend due process.16
The closest analogy is to free exercise claims.
PostEmployment Division of Oregon v. Smith,17 courts apply much less
searching review to laws of general applicability that incidentally
burden religious exercise.18 However, even post-Smith, if a
generally applicable law burdens a specific religious practice, a
court does not—indeed, could not—find as a basis for supporting
the law that the burdened party can exercise their religion just as
avidly despite the burden, due to a different but legally available
means for the same expression. In Smith itself, the majority did
not conclude—and in fact expressly declined to conclude—that
Native Americans could engage in their religious rituals without
peyote.19. But this is what happens in every First Amendment case
16. This was true pre-Obergefell as well. See, e.g., Kerrigan v. Comm’r of Pub.
Health, 957 A.2d 407, 414–17 (Conn. 2008) (rejecting the argument that the
state’s refusal to marry same-sex couples did not violate due process because civil
unions in the state entitled “gay persons . . . to all of the rights that married
18. See id. at 892 (“[T]he Court holds that where the law is a generally
applicable criminal prohibition, our usual free exercise jurisprudence does not
even apply. . . .).
Repeatedly and in many different contexts, we have warned that
courts must not presume to determine the place of a particular belief
in a religion or the plausibility of a religious claim . . . . It is not within
the judicial ken to question the centrality of particular beliefs or
practices to a faith . . . .
(quoting Hernandez v. Comm’r, 490 U.S. 680, 699 (1989))).
The Court’s undue burden analysis in its abortion jurisprudence might be
analogous. A law closing every abortion clinic in a given area might not violate a
woman’s right (from that area) to terminate a pre-viability pregnancy if the
woman can obtain an abortion in some other less convenient, though still
available, locality. See, e.g., Whole Woman's Health v. Hellerstedt, 136 S. Ct.
2292, 2313 (2015) (“We recognize that increased driving distances do not always
constitute an ‘undue burden.’”). But the undue burden test is about determining
whether inconvenience has crossed into impermissible interference with the right
the test protects. See Planned Parenthood of Southeastern Pennsylvania v. Casey,
505 U.S. 833, 920 (1992) (“A state-imposed burden on the exercise of a
constitutional right is measured both by its effects and by its character: A burden
may be ‘undue’ either because the burden is too severe or because it lacks a
legitimate, rational justification.”). In those First Amendment cases where a
choice as to mode, time, or place for speech is itself expressive, the alternative
saving the regulation’s constitutionality (as discussed infra notes 26, 39–41) is
involving a content-neutral law; such restrictions are upheld so
long as the reviewing court deems that the speaker could have
expressed the speech in question through a different mode.20 And
most importantly, this is so irrespective of the speaker’s
assessment of the equality of the substitute.21 This is the
equivalent of arguing that a state’s blanket ban on peyote does not
offend the free exercise rights of Native Americans because they
remain free to use sweat lodges.22
Unlike other anomalies in constitutional law, this one has
garnered near-total acceptance. In a rare display of unanimity in
an area where first principles have been contested for decades,23
courts and scholars have long found that it is not a constitutionally
significant intrusion upon free speech to limit a speaker’s preferred
mode of expression, so long as the intrusion leaves open other
available means by which the speaker may communicate.24 This is
usually a materially different expressive act from the one the speaker chose. See
City of Ladue v. Gilleo, 512 U.S. 43, 55–56 (1994) (stating that the alternatives
proposed to respondent for displaying opposition to war efforts “carr[y] a message
quite distinct” from displaying a sign from one’s residence).
See, e.g., Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989)
[E]ven in a public forum the government may impose reasonable
restrictions on the time, place, or manner of protected speech, provided
the restrictions “are justified without reference to the content of the
regulated speech, that they are narrowly tailored to serve a significant
governmental interest, and that they leave open ample alternative
channels for communication of the information.”
(quoting Clark v. Community for Creative Non-Violence, 468 U.S. 288, 293
21. See id. at 789 (finding alternative avenues of expression, but not
considering respondent’s input on the stated alternatives).
22. Post-Smith, the Religious Freedom Restoration Act sought to protect
religious exercise from “burdens” imposed by “laws [that are] neutral toward
religion,” which it deemed as offensive to religious practice “as laws intended to
interfere with religious exercise.” 42 U.S.C. § 2000bb(a)(2) (2012). The U.S.
Congress therefore expressly rejected any distinction between laws that could be
characterized, to use Speech Clause terminology, as “religious exercise-based”
and those that are “religious exercise-neutral.” Id.
23. See infra Part III.A (referring to the ongoing debate between
marketplace and self-autonomy theorists as to which theory best supports the
24. See C. Edwin Baker, Unreasoned Reasonableness: Mandatory Parade
Permits and Time, Place, and Manner Restrictions, 78 NW. U. L. REV. 937, 937
(1984) [hereinafter Baker, Unreasoned Reasonableness] (arguing that subjecting
time, place, and manner restrictions interfering with expressive conduct to the
equivalent of a mere reasonableness standard is “possibly the most universally
so as to self-autonomy, marketplace, and instrumentalist
On the self-autonomy front, Daniel Farber and John Nowak
argue that, “[a]lthough some people may be unable to express
themselves in the exact physical manner, location, or time they
find most satisfying, this inconvenience hardly seems a radical
intrusion into individual autonomy.”25 Similarly, Eugene Volokh
claims that “a typical law aimed at noncommunicative effects is
unlikely to excessively inhibit the communication of some
viewpoint of fact, because many different media would remain
available to the speakers.”26 As to marketplace theory, Geoffrey
Stone has argued that the content of the message is not blocked
from the speech market so long as that content can reach the
market via some other legal channel accessible to the speaker and
his audience.27 The restrained expression can thus still contribute
to the search for truth, so the harm the restriction causes, both to
the speaker and to listeners participating in the broader speech
market, is minimal.28 For the instrumentalists, Judge Richard
Posner compares restrictions on a speaker’s preferred mode of
communication to a “tax of variable severity on ideas and opinions
[that make] it more costly for the speaker to reach his audience,”
analogous to a “tax on newsprint or on broadcast air time, or for
that matter, an increase in second-class postal rates”29—a burden
paling in comparison to the “much heavier tax” of a content-based
restriction on a particular category of ideas.30
However, there are several unexamined problems with the
ample alternative channels approach—problems to which the
claims set out above do not respond. First, one can safely assume
that by dint of the chosen channel of communication alone, the
speaker herself has not found alternative channels of
communication to be analogous. It seems much more than a mere
“inconvenience,” to use Farber and Nowak’s term,31 to tell a
speaker she can be punished for using the mode of expression that
she believed to be most effective, but could have avoided
punishment if she had chosen a way to communicate the same
message that she likely viewed as less effective. Accordingly, the
notion that the availability of substitutes for expressing a given
idea minimizes the constitutional harm to a barred speaker’s
freedom of choice offends the self-autonomy-related justifications
for the First Amendment.32 It also seems wrong to find that the
marketplace of ideas is not harmed when speakers are barred by
generally applicable restrictions on the ground that speech could,
in theory, reach the market in some other way.33 The deprivation
of the market has already occurred, and from both the speaker’s
and his intended audience’s perspective, the alternative means by
which the speech could have reached the market are not true
contemporaneous substitutes. And Judge Posner’s analogy to taxes
on modes of speech delivery conflates the concepts of restriction
and proscription: a “tax on newsprint,” even a significant one, is
not the same as making newsprint illegal (as opposed to costly) and
forcing newspaper printers to become broadcasters, corner
speakers, or bloggers.34
30. See Posner, Free Speech in an Economic Perspective, supra note 29, at 16
(“A prohibition on all public expression of an idea is a much heavier tax.”).
31. Farber & Nowak, supra note 25, at 1237.
32. See Hurley v. Irish-Am. Gay, 515 U.S. 557, 573 (1995) (“[T]his general
rule, that the speaker has the right to tailor the speech, applies not only to
expressions of value, opinion, or endorsement, but equally to statements of fact
the speaker would rather avoid . . . .”).
33. See Stone, supra note 10, at 65 (stating that not all alternative means of
expression are perfect and that, in some circumstances, certain means of
expression may have specific advantages).
34. Cf. NLRB v. Fruit & Vegetable Packers & Warehousemen, 377 U.S. 58,
76–80 (1964) (Black, J., concurring) (justifying a restriction on expression on the
Speaking of bloggers: with the rise of speech-facilitating
technologies, there is now no limit to the damage that alternative
channels analysis can do to free speech.35 Courts have already
begun to find that thanks to the availability of such technologies,
an alternative means of expression is always available to any user
whose speech has been infringed.36 And in a world where the
existence of YouTube, Blogger, Twitter, or Facebook means that an
alternative channel to the one chosen is always available, even the
broadest content-neutral restrictions on speech will increasingly
survive judicial review.
To adduce precisely why ample alternative channels analysis
has won such a hold on First Amendment doctrine, it makes sense
to consider its origin. And if those foundations are shaky, we might
be more inclined to examine whether ample alternatives analysis
serves a proper purpose in balancing the liberty of the speaker
against the government’s interests in generally applicable
regulations that infringe on speech.
Ample alternative channels analysis arose more than fifty
years ago, in Justice John Marshall Harlan’s concurrence in
United States v. O’Brien.37 In that seminal case, Harlan stated the
First Amendment was not offended by O’Brien’s prosecution for
burning his draft card because O’Brien could have communicated
his antigovernment, antiwar message by other, legal means.38 The
ground that “other methods of communication are left open” and thus the
restriction is “on a par with holding that governmental suppression of a
newspaper in a city would not violate the First Amendment because there
continue to be radio and television stations”).
35. See Interstate Outdoor Adver., L.P. v. Zoning Bd. of Mt. Laurel, 706 F.3d
527, 535 (3d Cir. 2013) (“Potential alternative channels of communication [for
prohibited billboards] include on-premises signs, internet advertising, direct
mail, radio, newspapers, television, sign advertising, and public transportation
advertising.”); see also Gun Owners’ Action League v. Swift, 284 F.3d 198, 212
(1st Cir. 2002) (“The restriction challenged here ‘allows for reasonable alternative
channels of communication.’ Whatever messages the appellants seek to express
by shooting at human images on targets, those messages may be spread via
writing, the Internet, word of mouth, or other communication technologies.”).
36. See Bl(a)ck Tea Soc’y v. City of Bos., 378 F.3d 8, 14 (1st Cir. 2004) (finding
that the ability to communicate protest messages through mass media qualified
as a “viable alternative means . . . to enable protesters to communicate their
messages to the delegates”).
37. 391 U.S. 367, 388–89 (
) (Harlan, J., concurring).
38. See id. at 389 (“O’Brien manifestly could have conveyed his message in
many ways other than by burning his draft card.”).
development of Justice Harlan’s concurrence, however, as
evidenced through multiple drafts of that concurrence and the
majority opinion in the case—none of which have ever been
examined by scholars until this Article—shows that ample
alternative channels analysis was in its incipiency a misguided
afterthought.39 By dint of its being incorporated into the test for
content-neutral restrictions and applied in subsequent cases,
however, the concept now carries dispositive force in First
Amendment doctrine.40 It is time to save the doctrine, and the
interests it is intended to protect, from the tyranny of the
With the help of Justice Harlan’s case files in O’Brien, Part II
of this Article discusses the birth of the ample alternative channels
analysis and its role in the content-discrimination doctrine’s
intermediate scrutiny standard of review.41 Part III critiques
ample alternative channels analysis because in its blanket
treatment of alternatives as dispositive to the First Amendment
issue, it fails to differentiate those cases in which the speaker’s
chosen mode of expression is as worthy of protection as the
message expressed thereby.42 In terms of practical consequences,
this Part also shows that ample alternative channels analysis often
yields inconsistent and speech-averse results in a range of
contexts—a problem that will only be compounded by the current
emergence of technology-facilitated speech.43 In closing, Part III
argues that the Supreme Court’s longstanding distinction between
content-based and content-neutral restrictions cannot support the
use of ample alternative channels analysis in the latter type of case
but not the former.44
Turning to recommendations, Part IV proposes that in lieu of
ample alternative channels analysis, courts reviewing
contentneutral regulations should ask whether permitting the speaker’s
use of her chosen channel of communication is incompatible with
the government’s interest in adopting the regulation in question.45
Incompatibility is a more rigorous standard than the
reasonableness approach that has come to govern content
neutrality analysis.46 It will protect the speaker’s chosen mode of
expression in most cases, except those in which permitting the
speaker’s chosen mode of expression would frustrate a compelling
governmental interest in near-totality.47 In other words, the focus
is not where the current doctrine places it—on whether the
speaker could have met her communicative goals by expressing the
content of her message in another legal way. Rather, the inquiry is
whether the government could not have achieved its legislative
goals if the speaker had been able to express her message in the
desired manner. An incompatibility standard would also be able to
differentiate between conduct that merely facilitates speech and
conduct that is itself communicative or otherwise essential to the
speaker’s expressive act, which is a distinction that modern First
Amendment doctrine has merged right out of the law. Finally,
abandoning ample alternative channels analysis in favor of
incompatibility would be consistent with the Roberts Court’s turn
toward a pro-speaker view of the First Amendment—a view that
rigorously reviews even content-neutral restrictions on speech, as
evidenced in last Term’s McCullen v. Coakley.48
44. Infra Part III.
45. Infra Part IV.
46. See Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989) (articulating
the reasonableness standard that currently governs restrictions “on the time,
place, or many of protected speech”).
47. See Rosenbaum v. City of S.F., 484 F.3d 1142, 1167 (9th Cir. 2007)
(stating an “incompatibility” standard that must be proved by the State to deny
appellants free speech in the public forum).
48. See 134 S. Ct. 2518, 2541 (2014) (“[The Commonwealth has pursued
those interests by the extreme step of closing a substantial portion of a traditional
public forum to all speakers . . . . The Commonwealth may not do that consistent
with the First Amendment.”).
II. The Birth of Ample Alternative Channels Analysis: Justice
Harlan’s O’Brien Opinion
To better understand the role alternative channels analysis is
intended to play in Free Speech doctrine, we should start with the
analysis’s source. When looking for origin stories in constitutional
law, one can easily find statements in cases that were
afterthoughts at the time, but later serve as the foundations upon
which subsequent courts build legal doctrine.49 But excavating
those statements can uncover mismatches between the context and
principles animating the statements at the time of their making on
the one hand, and their application to modern controversies on the
other—mismatches that current law, building upon itself through
the common law process of decision, can be blind to.50
For present purposes, the statement being excavated first
appeared in a tent-pole First Amendment case: David Paul
O’Brien’s conviction under the Selective Service Act (SSA or Act)
for burning his draft card on the steps of the South Boston
Courthouse in protest of the Vietnam War.51 Alternative channels
analysis was born not in the O’Brien majority opinion, however,
but rather in Justice Harlan’s concurrence.52 And an earlier
version of that latter opinion had much bigger game in its sights:
what Harlan viewed as the dangerous logical fallacy of the Court’s
distinguishing between speech and content in First Amendment
A. Briefing, Argument, and the First Draft Majority Opinion
After being convicted in Massachusetts federal court for
violating the “no willful destruction” section of the SSA by burning
his draft card, O’Brien appealed to the U.S. Court of Appeals for
the First Circuit.54 That court held that O’Brien’s conviction
violated the First Amendment.55 In the Supreme Court, the United
States challenged the First Circuit’s reversal of O’Brien’s
On the First Amendment question, the United States’ merits
brief argued that “the decisive consideration” in determining
whether conduct like O’Brien’s was protected speech was whether
it fit within those “limited class of activities” that are
“[1.] inextricably tied to oral expression or [2.] where no reasonably
effective alternative means of communication [are] available.”57 As
to the first question, the United States claimed that burning a
draft card was not protected symbolic speech because it was not “a
natural extension of [a] verbalization,” not integral to a
concomitant “oral expression[’s] meaning,” and not “the manifest
equivalent of, or traditionally recognized substitute for, a verbal
statement.”58 As to the second alternative showing under its
proposed test, the Government argued that “other effective means
for expressing” the ideas communicated by the conduct that
O’Brien engaged in “plainly exist,” and that those means, unlike
United States v. O’Brien at 2 (Harlan, J., concurring), in John Marshall Harlan
Papers, Box 311, Seeley G. Mudd Manuscript Library, Princeton University
[hereinafter Harlan Papers, Box 311] (“This double-barreled approach seems to
me hopelessly to confuse two separate definitional problems presented by the
language of the First Amendment.”) (on file with the Washington and Lee Law
54. O’Brien v. United States, 376 F.2d 538, 539–40 (1st Cir. 1967), vacated
391 U.S. 367, 369 (
55. See O’Brien, 391 U.S. at 370–71 (“On appeal, the Court of Appeals for the
First Circuit held the 1965 Amendment unconstitutional as a law abridging
freedom of speech”).
56. See id. at 372 (“The government petitioned for certiorari in No. 232,
arguing that the Court of Appeals erred in holding the statute
unconstitutional . . . .”); see also Brief for the United States at 7–8, United States
v. O’Brien, 391 U.S. 367 (
) [hereinafter O’Brien United States Brief] (setting
forth the United States’ arguments for challenging reversal).
57. O’Brien United States Brief, supra note 56, at 8.
58. Id. at 15.
draft-card burning, “do not interfere in any significant way with
the orderly functioning of government.”59 O’Brien and like-minded
protestors had such “other effective means” to convey their views,
ranging from “the use of mass communication media” to “the public
meeting hall” to “peaceable demonstration” to “the distribution of
literature.”60 In response, O’Brien argued that the Court’s cases
established “a constitutional right to deliver one’s speech at the
place where, the time when, and the manner in which the speaker
deems it to be most effective.”61 That right, claimed O’Brien,
“include[s] the right to make the most dramatic and compelling
speech possible,” subject to narrow limitations.62
After oral argument, at which the Solicitor General reiterated
the Government’s argument that O’Brien was free “at all times to
express dissent by speech from the courthouse steps or on the
street corners, by letters to the editor, by pamphlet, by radio and
television,”63 the Government’s position prevailed. Chief Justice
Warren’s first draft opinion for the majority circulated on April 12,
1968.64 Consistent with only the first part of how the Government
had litigated the symbolic speech question, however, Warren’s
opinion turned exclusively on the speech-conduct distinction.65
Warren found that burning draft cards, even in protest of military
action, was conduct and not speech—and was thus not protected
by the First Amendment.66
Warren’s first analytical step was to affirm that the Act did
not facially infringe upon speech.67 As a general matter, there was
nothing “necessarily expressive,” Warren noted, about the conduct
that the Act prevented.68 Next, and with respect to O’Brien’s
asapplied claim alleging that his conduct was expressive and
therefore protected, Warren’s opinion held firmly that even though
the Court had “no reason to doubt” that O’Brien “intended by
burning his certificate to express his disagreement with the war
and the draft,” O’Brien’s “conduct of burning his certificate was not
speech within the meaning of the First Amendment.”69
With the Government’s position as a starting point for his
analysis, Warren rejected O’Brien’s position that the First
Amendment protected symbolic speech, driving a broad wedge
between words and conduct—even conduct intended to be
expressive—for First Amendment purposes.70 He did so first by
placing conduct as far subordinate to verbal utterances in the free
The view of the First Amendment advanced by O’Brien is
premised upon a definition of ‘speech’ that bears no resemblance
to the meaning and usage of that word in our society. Under this
view, any act done by an individual would be speech if the
individual intended by the act to express any idea, and at some
time made known his intent . . . . The multitude of decided cases
corroborate what is in any event apparent on the face of the
[Speech and Press] clauses—that their core meaning and
concern is with verbal expression, the spoken and written
utterance of words . . . . From its adoption through the present
time, the traditional, normal, and by far most important way
that people in our society have expressed their ideas is by using
So the “spoken and written utterance” was at the core of the
Speech Clause’s concerns, and to stray from that core was to depart
67. See id. at 8 (“[The 1965 Amendment] prohibits the knowing destruction
of certificates issued by the Selective Service System, and there is nothing
necessarily expressive about such conduct.”).
68. Id. at 8.
69. Id. at 16 (emphasis added).
70. See id. at 17 (stating that the cases decided on this confirm that the
clauses of the First Amendment concern “verbal expression, the spoken and
written utterance of words”).
71. Id. at 17–18.
not only from First Amendment text, history, and tradition, but
also from common conceptions regarding the sharing of ideas that
have been widely held from the Founding to the present day.
Warren admitted, however, that under the Court’s previous
cases, one narrow category of conduct did merit First Amendment
protection, noting that “the Amendment would have a rather
narrow compass if it embraced only the initial utterance of words.
The conception of freedom of communication embodied in the
Amendment by definition draws within its ambit behavior engaged
as a means of communicating, that is of disseminating or
transmitting, uttered words.”72 According to Warren, examples of
such verbal expression-disseminating conduct that could fall
within the Speech Clause’s protection included assemblies to
discuss matters of public interest,73 a speaker’s use of sound
amplification devices,74 union soliciting without a permit,75 and
“distributing . . . printed material.”76 Only conduct that was “a
natural extension of a verbalization” was protected.77
Speechfacilitating conduct, in other words, was protectable, but conduct
intended to communicate nonverbally, or what would later come to
be known as symbolic speech, did not implicate the First
Amendment at all.
Based on his distinction between conduct that was
unprotected even if intended to be expressive and conduct that was
“a means for the dissemination of verbal expression” and thus
protected,78 Warren concluded that “burning a document”79 fell
into the former category.80 Such an act was “wholly unrelated to
the employment of language, and consequently, its protection is of
no moment to the core concern of the First Amendment.”81
“Preventing people from burning things,” Warren concluded, “in no
73. Id. (citing Hague v. Comm. for Indus. Org., 307 U.S. 496 (1939)).
74. Id. (citing Saia v. New York, 334 U.S. 558 (1948)).
75. Id. (citing Staub v. City of Baxley, 355 U.S. 313 (1958)).
76. Id. at 19 (citing Ex parte Jackson, 96 U.S. 727 (1877); Schneider v. State,
308 U.S. 147 (1939); Martin v. City of Struthers, 319 U.S. 141 (1943)).
77. Id. at 20.
81. Id. at 20–21.
way impinges on their freedom to communicate ideas through
language.”82 Accordingly, punishing O’Brien for an act that was
“far-removed from what we mean by the word speech in our
society” did not implicate the First Amendment.83
Concluding, Warren summarized his proposed “commonsense”
holding for the Court, finding that
[a]n act unrelated to the employment of language is not speech
within the First Amendment if as a matter of fact the act has
an immediate harmful impact completely apart from any
impact arising by virtue of the claimed communication itself.
And if the Government . . . has attached legal consequences to
that noncommunicative impact, those consequences may be
enforced against the person who committed the act.84
B. Justice Harlan’s Response: Ample Alternative Channels Is Born
Upon receipt of Chief Justice Warren’s first draft for the
majority, Justice Harlan set out over the latter half of April 1968
to draft a concurrence.85 Harlan’s lengthy draft would express deep
concern with the Warren draft’s rejection of the idea that the First
Amendment could protect symbolic conduct. The first draft of
Harlan’s concurrence, circulated on May 1, 1968, stated the Justice
was “in full accord with the reversal” of the First Circuit’s holding
in O’Brien’s favor, yet Harlan was “unable to subscribe to the
process of reasoning by which the Court concludes that O’Brien’s
conviction for draft card burning . . . did not violate his right to free
speech as assured by the First Amendment.”86 That reasoning,
Harlan continued, employed “restrictions on the reach of the First
means of communication, however, the court found that ample
alternative channels existed because the petitioners could hold
signs, sing, call out to individuals entering the facility (including
through the use of amplification equipment), and engage in silent
and vocal prayer.302 Courts reach the same result in bubble zone
cases. Those cases find ample alternative means for the speaker
such as “peacefully hand[ing] leaflets” to the clinic-entering
individual “as they pass by” without entering the bubble zone
around that person.303 In sum, reviewing courts have consistently
found that ample alternative means exist for a speaker who is
barred from communicating face-to-face with abortion clinic
The long procedural history of the 2013–14 Term’s McCullen
v. Coakley305 demonstrates this principle. In McCullen, the Court
found that a Massachusetts statute establishing a thirty-five-foot
buffer zone around the entrances and driveways of reproductive
health service facilities was content neutral, but not narrowly
tailored and thus failed intermediate scrutiny.306 At every point
prior to McCullen’s reaching the Court, however, the lower courts
held that ample alternative channels existed for individuals
302. Id. at 629; see also United States v. Weslin, 156 F.3d 292, 298 (2d Cir.
1998) (per curiam) (finding the federal Freedom of Access to Clinic Entrances Act
constitutional because protestors are still “at liberty to hold signs, pass out
handbills, speak conversationally, and so forth”); American Life League, Inc. v.
Reno, 47 F.3d 642, 652 (4th Cir. 1995) (1995) (finding the Freedom of Access to
Clinic Entrances Act leaves open ample alternative means for communication
because the Act allows individuals to express antiabortion messages in a
“nonviolent, non-obstructive manner” via “voice, signs, handbills, symbolic gestures,
and other expressive means”).
303. See Madison Vigil for Life, Inc. v. City of Madison, Wis., 1 F. Supp. 3d
892, 895 (W.D. Wis. 2014) (analyzing Hill v. Colorado, 500 U.S. 703 (2000)); see
also McTernan v. City of York, 564 F.3d 636, 657 (3d Cir. 2009) (analyzing
Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753 (1994))).
304. See McTernan, 564 F.3d at 657 (citing Schenck v. Pro-Choice Network of
W. N.Y., 519 U.S. 357, 376 (1997); Madsen v. Women’s Health Cntr., Inc., 512
U.S. 753, 770 (1994) (holding that a thirty-six-foot buffer zone around the clinic
entrances burdens no more speech than necessary to accomplish the
governmental interest at stake); McGuire v. Reilly, 260 F.3d 36, 49 (1st Cir. 2001)
(upholding a six-foot floating buffer zone).
306. See id. at 2523 (explaining that even though the Act is content neutral,
it is not narrowly tailored because it burdens more speech than necessary to
further the legitimate interests of the government).
seeking to engage in face-to-face counseling of women entering
those facilities.307 The Court itself purported not to reach the ample
alternatives issue.308 But its tailoring analysis focused closely on
the petitioners’ preferred mode of communication, how the
Massachusetts statute severely hampered that mode, and how
alternatives to that mode proffered by the state in support of its
restriction failed to cure the alleged First Amendment violation:
[At each of the] Planned Parenthood clinics where petitioners
attempt to counsel patients, the zones carve out a significant
portion of the adjacent public sidewalks, pushing petitioners
well back from the clinics’ entrances and driveways. The zones
thereby compromise petitioners’ ability to initiate the close,
personal conversations that they view as essential to “sidewalk
These cases raise important and sensitive issues regarding the
conflict between the exercise of two fundamental constitutional
rights.310 But ample alternative channels analysis is no way to
decide them. The first question the pre-McCullen cases raise is
whether we want First Amendment rights to be defined via the
application of a judicial yardstick. If a one-hundred-foot buffer
zone is deemed to have not foreclosed alternative means of
307. See McCullen v. Coakley, 844 F. Supp. 2d 206, 217, 224 (D. Mass. 2012),
aff’d, 708 F.3d 1 (1st Cir. 2013) (analyzing the statute as applied to clinics in
Boston, Worcester, and Springfield, and finding ample alternatives existed in all
three cases); McCullen v. Coakley, 573 F. Supp. 2d 382, 413 (D. Mass. 2008), aff’d,
571 F.3d 167 (1st Cir. 2009) (finding the statute left ample alternative means in
response to a facial challenge because most expressive activity can be seen and
heard by people entering and exiting the buffer zone); McCullen v. Coakley, 571
F.3d 167, 180 (1st Cir. 2009) (finding protestors could “speak, gesticulate, wear
screen-printed T-shirts, display signs, use loudspeakers, and engage in the whole
gamut of lawful expressive activities,” and thus ample alternatives to the buffer
zone were available).
308. See McCullen v. Coakley, 134 S. Ct. 2518, 2540 n.9 (2014) (“Because we
find that the Act is not narrowly tailored, we need not consider whether the Act
leaves open ample alternative channels of communication.”)
309. Id. at 2535; see also id. at 2536 (noting that because of the difficulty
petitioners encounter identifying patients before the patients enter the buffer
zone, petitioners “often cannot approach them in time to place literature near
their hands—the most effective means of getting the patients to accept it”).
310. See, e.g., Leading Case: McCullen v. Coakley, 128 HARV. L. REV. 221, 228–
30 (2014) (arguing that the Court should analyze abortion protest restriction
cases by balancing protestors’ right to free speech against women’s constitutional
right to seek abortions at the place of protest).
communication for the barred protestor, for example, it is certain
that the fifty-foot zone in the next case will survive.311 Moreover,
the Court’s disposition in McCullen recognizes that a speaker’s
preferred mode of expression deserves meaningful First
Amendment protections.312 While purporting not to reach ample
alternative channels, the Court flatly rejected the lower courts’
consistent findings that substitutable alternatives existed for these
speakers despite the restrictions in question:
Respondents also emphasize that the Act does not prevent
petitioners from engaging in various forms of “protest”—such as
chanting slogans and displaying signs—outside the buffer
zones. That misses the point. Petitioners are not protestors.
They seek not merely to express their opposition to abortion, but
to inform women of various alternatives and to provide help in
pursuing them. Petitioners believe that they can accomplish
this objective only through personal, caring, consensual
conversations. And for good reason: It is easier to ignore a
strained voice or a waving hand than an outstretched arm.313
A results-oriented view of the Court’s First Amendment
jurisprudence might well conclude that McCullen is an abortion
speech case, that the Court treats abortion speech differently, and
that McCullen’s reach will thus be limited to those kinds of
cases.314 As we saw above, most First Amendment scholars view
the secondary effects doctrine as limited to its subject matter. But
another reading of the case reveals skepticism of, if not hostility to,
deciding First Amendment cases on ample alternative channels
arguments so readily, particularly when such alternatives are, by
the restricted speaker’s lights and to use the Chief Justice’s word,
irrelevant to the speaker’s communicative intent.315
311. See McGuire v. Reilly, 260 F.3d 36, 39 (1st Cir. 2001) (applying Hill to
find that if the one-hundred-foot buffer and eight-foot bubble zone in Hill left open
ample alternative means, then an eighteen-foot buffer and six-foot bubble zone
indisputably did as well).
312. See id. at 43 (discussing judicial review standards of First Amendment
313. McCullen, 134 S. Ct. at 2536.
314. See id. at 2541 (Scalia, J., concurring) (“There is an entirely separate,
abridged edition of the First Amendment applicable to speech against abortion.”).
315. See, e.g., id. at 2524 (“The Commonwealth has not shown that it seriously
undertook to address these various problems with less intrusive tools readily
available to it.”).
By favoring the mode of communication that the speakers
themselves “view[ed] as essential” in its narrow tailoring analysis,
and in giving that choice of mode significant weight in assessing
whether the restriction at issue burdened the speakers’ expressive
rights, the Court in McCullen took a first step toward diminishing
the power of ample alternative channels analysis. 316 It should take
the next step in that direction and abolish the inquiry from its
review of content-neutral laws.
C. Conflicts to Come: Online Speech
With the emergence of technology-enabled communication, we
can expect the harms associated with alternative channels
analysis to increase. As new methods of communication continue
to develop, alternatives to the speaker’s choice of communicative
mode will continue multiplying. Their existence tips First
Amendment analysis in content-neutral cases in the government’s
There is little doubt that online connectivity has expanded the
communicative opportunities and audience reach of nearly every
speaker.318 For present purposes, the substitute avenues for
speakers that ample alternative channels analysis renders
dispositive have increased exponentially.319 Following the court
decisions discussed in Part III.B.1 supra that have found
opportunities to communicate via television and radio were ample
alternatives to face-to-face protest activity, governments are
already relying on this argument in defending regulations against
First Amendment challenges.320 In last Term’s Reed v. Town of
316. See id. at 2535 (noting that because of the buffer zone, one speaker
claimed she had to “rais[e] her voice at patients from outside the zone—a mode of
communication sharply at odds with the compassionate message she wishes to
317. See, e.g., Bl(a)ck Tea Soc’y v. City of Boston, 378 F.3d 8, 14 (1st Cir. 2004)
(finding that the district court did not err in its First Amendment analysis
because means of communication existed through the internet and television).
318. See id. (indicating that the speaker possessed an alternative for
communication through the internet).
319. See id. (“At a high-profile event . . . messages have a propensity to reach
delegates through television, radio, the press, the internet, and other outlets.”).
320. See id. (“[W]e think that the appellant’s argument greatly
Gilbert,321 discussed in more detail infra, the town of Gilbert
adopted a complicated and comprehensive regime regulating the
size and duration of signs that could be displayed within the city
limits.322 In defending its ordinance under intermediate scrutiny,
the Town argued that the church whose speech was abridged could
have, and indeed did, express its views in an unfettered fashion in
a range of other ways not implicated by the ordinance’s
restrictions, arguing that the petitioners “have utilized a whole
host of avenues to spread the word about their mission and
location, including the internet, print advertising, personal
solicitations, pamphlets, telephone calls, and emails.”323 The Ninth
Circuit accepted this reasoning before the case reached the
The problems raised by this convergence of doctrine and
technological change are obvious and sobering. In 2016, it is
difficult to conceive of a case where a content-neutral restriction
on speech could not be defended on the ground the restricted
speaker could “spread the word” through “the internet . . . and
emails.”325 Digital connectivity has removed spatial and temporal
limitations on communication, but in a First Amendment world
where the availability of alternative channels is dispositive, it has
also expanded the government’s ability to restrict speech through
the use of content-neutral restrictions, because that connectivity
provides every speaker with alternative modes of communication
to those the speaker chose.
The advent of online speech spaces has compounded the
alternative ways by which practically any speech could be shared
and accessed, and has correspondingly compounded the problem
underestimates the nature of modern communication.”).
321. 135 S. Ct. 2218 (2015) (overruling Reed v. Town of Gilbert, 707 F.3d 1057
(9th Cir. 2013)).
322. See id. at 2224 (“The Sign Code identifies various categories of signs
based on the type of information they convey then subjects each category to
323. Brief for Respondents at 52–53, Reed v. Town of Gilbert, 135 S. Ct. 2218
(2014) (No. 13-502) [hereinafter Brief for Respondents].
324. See Reed v. Town of Gilbert, 707 F.3d 1057, 1075–76 (9th Cir. 2013),
overruled by Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015) (noting that the
ordinance “does not regulate any of the many other ways in which” the church
could engage in its intended speech).
325. Brief for Respondents, supra note 323, at 52.
that ample alternative channels analysis presents. It is now
literally impossible for a ban on one venue for speech to operate as
a complete ban on the speech in which that venue was to be
expressed. Accordingly, in the context of content-neutral
restrictions, the emergence of the Internet, ironically enough, will
result in the systematic underproduction of speech.
D. The Content Based vs. Content-Neutral Fallacy
Let’s back up for a moment. In its First Amendment cases, the
Supreme Court has long adhered to a distinction between
“restrictions that turn on the content of expression,” which “are
subjected to a strict form of judicial review,” and restrictions that
are “concerned with matters other than content,” which “receive
more limited examination.”326 As this Article details, this “more
limited examination” considers the availability of alternative
channels of communication for the abridged speech in question.327
This Section considers why First Amendment doctrine has tied
these two inquiries so closely together.
Ample alternative channels analysis and the justifications for
a lesser standard of review for content-neutral restrictions work
hand-in-hand. The very existence of alternative channels of
communication supports the conclusion that the content
restricting law in question is content-neutral.328 As noted above,
because a content-neutral regulation’s effects are deemed by a
reviewing court to infringe upon, at most, a means or locus of
expression and not the expression’s content, the speaker restricted
by such a regulation is free to shift to another means or locus to
express the same message.329 Accordingly, the alternatives
demonstrate that the law is not aimed at the content of the
speaker’s message and that content remains able to reach its
326. Redish, The Content Distinction, supra note 157, at 113.
327. Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 293 (1984).
328. See id. at 144 n.186 (“[T]he more alternative avenues of expression that
are available, the less justification the state must provide for restricting
329. See id. at 116 (“The reason that [content-neutral restrictions] are
generally unobjectionable is that they presume the existence of alternative
avenues of expression, alternatives that are by definition unavailable in the case
of absolute regulation.”).
intended listeners and contribute to public debate.330 Ample
alternative means analysis can thus be justified as an additional
way to smoke out whether the restriction in question is truly
content-neutral—if other ways to communicate the same message
are available, then the reviewing court can comfortably conclude
that the facially neutral restriction is not intended to restrict the
message, but rather its mode.331 Or so the theory goes, anyway.
But even if it makes sense, why is this inquiry irrelevant to
the analysis of content-based restrictions? Most content-based
restrictions (as opposed to apocryphal ones spoken of only in law
school classes and law review articles, such as “speech about
politics is banned”332) do not bar alternative channels of
communication either. As then-Professor Elena Kagan noted, even
a content-based restriction might make “the danger of distortion
insignificant” if it affects a “small quantity of speech” and leaves
“alternative means to communicate the ‘handicapped’ idea” readily
available to speakers.333 Yet under current law, the availability of
those alternative channels is irrelevant to a reviewing court once
a particular regulation is deemed content-based.334 Scrutiny here
is strict.335 Accordingly, the government’s burden is to justify the
330. See Quadres, Content-Neutral Public Forum Regulations, supra note 40,
at 480 (“[I]f one could argue that, despite the questioned regulation, a speaker
still has numerous alternative means by which to disseminate his message, the
degree of first amendment injury may seem insubstantial . . . . [because] [t]he
speaker can always make use of his alternative access.”).
331. This interpretation may support the conclusion that ample alternative
channels is simply a gloss on narrow tailoring analysis. But the two concepts are
distinct: tailoring analysis involves the government’s alternatives, while
alternative channels analysis focuses on the speaker’s alternatives. See Williams,
supra note 149, at 642 (stating that the requirement of alternatives “exists
because the Court believes that if adequate alternative channels of
communication remain, then a regulation restricting a particular alternative will
have no more than a minimal effect on speech.”).
332. See, e.g., Rubenfeld, supra note 133, at 793 (describing the hypothetical
arrest of an author for criticizing the President pursuant to a statute that
criminalizes such critiques).
333. Elena Kagan, Private Speech, Public Purpose: The Role of Governmental
Motive in First Amendment Doctrine, 63 U. CHI. L. REV. 413, 446 (1996).
334. See id. at 446–47 (commenting that despite the fact that some
contentbased restrictions have very minimal effect on skewing public discourse, First
Amendment doctrine does not distinguish “between content based laws of this
kind and [those] that wholly excise ideas from public discourse”).
335. See id. at 443–44 (“Content-based restrictions on speech-restrictions that
by their terms limit expression on the basis of what is said usually are subject to
level to which its interest in the restriction is compelling.336
Government arguments that a speaker’s message is not limited to
the mode of expression that the regulation bars, and that the
regulation’s harm to speech is thus minimal, are not a part of the
decision-making calculus for content-based laws.337
According to the relevant scholarship, the reason this is so is
that a law’s content-based nature is sufficient indicia standing
alone of improper governmental motivation.338 In other words,
content-based restrictions are suspect enough on their face that
there is no need for any further smoking out of government
intent.339 But even in the content-based context, the availability of
different modes to express the same content remains relevant,
particularly when a content-based law is aimed at a particular
mode of expression.340 A law that says “no draft card burning in
protest of the Vietnam War” says nothing about burning President
Johnson in effigy for the same reason. Can such a law safely be
deemed as aimed at protesting the Vietnam War and not at draft
card burning, merely by dint of its reference to content? The law’s
singling out of a particular mode of expression might lead one to
conclude that the law is not aimed at, or primarily concerned with,
restricting public debate, and thus a lesser standard of review
might be appropriate.341 But once the law makes reference to
content, the level of scrutiny is decided.342
far more rigorous scrutiny.”).
336. See id. at 444 (“Formulations of the standard used to review
contentbased action vary, but the Court most often requires the government to show a
compelling interest that could not be attained through less restrictive means.”).
337. Id. at 445.
338. See id. at 414 (“First Amendment law, as developed by the Supreme
Court over the past several decades, has as its primary, though unstated, object
the discovery of improper governmental motives. The doctrine comprises a series
of tools to flush out illicit motives and to invalidate actions infected with them.”).
339. See id. at 452 n.104 (“[T]he face of a law indicates more reliably than the
effects of the law what purposes underlie it.”).
340. See id. at 418 (concluding that a content-based ordinance, “while not
restricting a great deal of speech, thus restricted speech in a way that skewed
public debate on an issue by limiting the expressive opportunities of one side
341. See id. at 455 n.110 (stating that, “as the effect of a law on expressive
opportunities increases, so too should the government's burden of justification”
and thus, the standard of review used).
342. See id. at 499 n.237 (“To merit a stricter standard of review, a law would
If there is a place for alternative means analysis in First
Amendment doctrine at all, it should, as a matter of logic, apply to
both content-based and content-neutral restrictions. Nothing in
the distinction between content-based and content-neutral
restrictions calls for application of alternative channels analysis in
the latter set of cases but not the former.343 Martin Redish offers
one path, which he proposes should be applied in any case
involving regulations that adversely affect expression.344 In such a
case, a reviewing court should “balance the compellingness of the
state interest served by the law against the availability of
alternative means of expression to the speaker.”345 The less likely
it is “that the speaker will be able to find acceptable alternative
methods of expression,” Redish continues, “the more compelling
must be the government’s asserted justification.”346
As argued above, a primary flaw of ample alternative channels
analysis is that it favors judge-made alternative modes of
communication to those the speaker herself chose.347 One could
critique Redish’s proposal for significantly expanding the role
these alternatives would play by extending them to content-based
cases.348 But ample alternative channels analysis’s selective
application should trigger a critical reexamination of the doctrine,
the work it is intended to perform, and its effects. Redish’s
proposed test at least acknowledges the logical fact that
need to have a justification relating not to the restriction of speech generally
(which all content-neutral laws have), but to the restriction of speech of a certain
343. See Redish, The Content Distinction, supra note 157, at 129
(“[R]egulations that limit expression on content-neutral grounds should logically
be as suspect as content-based regulations, since they may also undermine this
344. See id. at 143 (suggesting that, like content-based inquiries by the court,
content-neutral inquiries also consider whether the government interests served
by the restriction are “compelling” enough to “justify significant invasions of free
345. Id. at 143; cf. Volokh, supra note 26, at 1307 (rejecting the application of
ample alternative means to content-based restrictions because of inconsistencies
in its application in content-neutral cases).
346. Redish, The Content Distinction, supra note 157, at 143.
347. Supra Part III.A.
348. See Volokh, supra note 26, at 1309–10 (criticizing the dangers of applying
the “vague ample alternative channels” analysis to content-based speech
alternative channels are relevant to either both kinds of cases or
The latter course is the better one. The availability of
alternative channels of communication may indeed be as relevant
to content-based restrictions as to content-neutral ones.350 But the
better conclusion is that considering such channels in the context
of reviewing any restriction on speech, especially facially neutral
ones, undermines longstanding rationales for the First
Amendment.351 Courts should thus no longer consider them. But
what should courts consider in their place?
IV. The Solution: Incompatibility
The foregoing has argued that First Amendment doctrine
should focus solely on the speaker’s preferred mode of speech and
the government restriction’s abridgement upon it, to the exclusion
of other hypothetical speech modes that the speaker has not used.
One way to achieve this goal is to apply an incompatibility test:
when a speaker’s expression is infringed by a law or regulation, a
reviewing court should ask whether the infringed speech act—in
the form the speaker intended to express it—is incompatible with
the law and its purpose. The law will survive as applied to the
speaker only if the speaker’s mode is incompatible with the
governmental interests asserted in the law’s support.
349. See Redish, The Content Distinction, supra note 157, at 83 (“Whatever
rationale one adopts for the constitutional protection of speech, the goals behind
that rationale are undermined by any limitation on expression, content-based or
350. See id. (noting that the proposed framework for evaluating the
constitutionality of “impaired speech” should include “whether ‘feasible’ less
restrictive alternatives are inadequate to accomplish that end; and whether the
speaker will have available adequate means to express the same views to roughly
the same audience.”).
351. But see id. (“Since the Court uses [this test] in reviewing content-based
regulations, it should have no greater difficulty in applying them to all
regulations of expression.”).
A. The Test
The concept of incompatibility is no stranger to First
Amendment doctrine. The Court’s initial foray into reviewing time,
place, and manner restrictions, 1972’s Grayned v. City of
Rockford,352 proposed an incompatibility test as the standard to be
applied to content-neutral laws that adversely affected speech.353
In Grayned, the speaker was convicted for protesting outside of a
high school regarding the school’s differential treatment of Black
students and prospective teachers.354 The school claimed that the
speaker and his fellow protestors’ activities disrupted classes and
caused other students to be tardy.355 The anti-noise ordinance that
the speaker was arrested for violating made it unlawful to “make
any noise or diversion” adjacent to a school in session “which
disturbs or tends to disturb the peace or good order” of that
In reviewing the ordinance, the Court held that “the nature of
a place, the pattern of its normal activities, dictate the kinds of
regulations of time, place, and manner that are reasonable.”357
“The crucial question,” the Court concluded, “is whether the
manner of expression is basically incompatible with the normal
activity of a particular place at a particular time.”358 Further, even
if the restriction in question is content-neutral on its face, a
reviewing court should “weigh heavily the fact that communication
is involved,” and the speech-suppressing regulation “must be
narrowly tailored to further the state’s legitimate interest.”359
352. 408 U.S. 104 (1972).
353. See id. at 120 (determining that the plaintiff’s noisy demonstrations were
disruptive and “incompatible with normal school activities” and thus, “may be
prohibited next to a school when classes are in session.”).
354. Id. at 105.
355. See id. (reporting that “the demonstrators repeatedly cheered, chanted,
baited policemen, and made other noise that was audible in the school; that
hundreds of students were distracted from their school activities”).
356. Id. at 107–08.
357. Id. at 116 (internal quotation marks omitted).
359. Id. at 116–17. The Court in Grayned held that the ordinance survived
incompatibility analysis because it “punishe[d] only conduct which disrupts or is
about to disrupt normal school activities,” as determined based on “an individual
basis, given the particular fact situation.” Id. at 119.
First Amendment law is also familiar with incompatibility
because of the concept’s use in forum doctrine. Forum doctrine is
often interpreted to allow governments to bar speech-related uses
of public property that are incompatible with the property’s
intended use.360 Incompatibility analysis in forum doctrine cases
focuses on whether there is physical incompatibility between the
intended mode of expression and the government’s intended use of
the property. For instance, in International Society for Krishna
Consciousness, Inc. v. Lee,361 the Supreme Court framed the issue
as whether in-person solicitation of travelers was incompatible
with the purpose of an airport terminal, which is to ensure those
travelers can timely reach their flight gates.362
What is called for here is more searching. Prima facie
incompatibility of the type discussed in public forum cases—what,
in application in Lee, amounted to mere inconsistency, which
sounds in rationality review363—should be insufficient for the
speech-restricting regulation to survive. Rather, the burden should
be on the government to show that its interests cannot be met if
the expression infringed by the restriction were permitted.364
Inconvenience to the government or its purposes in passing a law
should never be sufficient in such a case. And a prior use of the
same government property or other resource for expressive
purposes that occurred without incident, as was the case in
360. See Cornelius v. NAACP Legal Def. & Educ. Fund, Inc., 473 U.S. 788,
800 (1985) (“Because a principal purpose of traditional public fora is the free
exchange of ideas, speakers can be excluded from a public forum only when the
exclusion is necessary to serve a compelling state interest and the exclusion is
narrowly drawn to achieve that interest.”); see also Int’l Soc’y for Krishna
Consciousness, Inc. v. Lee, 505 U.S. 672, 698 (1992) (Kennedy, J., concurring) (“If
the objective, physical characteristics of the property at issue and the actual
public access and uses which have been permitted by the government indicate
that expressive activity would be appropriate and compatible with those uses, the
property is a public forum.”).
361. 505 U.S. 672, 698 (1992).
362. See id. at 682–83 (describing the traditional use and expectations of
363. See id. at 682 (stating that the appellants’ expressions in the airport were
inconsistent with the forum’s purpose because “terminals have never been
dedicated to expression in the form sought to be exercised here: i.e., the
solicitation of contributions and the distribution of literature”).
364. See id. at 679 (stating that “the government has a high burden in
justifying speech restrictions relating to traditional public fora”).
Clark,365 would cut the incompatibility analysis in the speaker’s
favor. In sum, a default rule of the type incompatibility analysis
would impose would return a speaker-favoring presumption to the
Court’s First Amendment cases—a presumption to which the
Court’s historical First Amendment cases express sympathy for.366
There is no denying that content-neutrality doctrine serves an
important overarching purpose: to ensure that there is no
freestanding speech-related defense to laws of general
applicability. But incompatibility analysis would not stand in the
way of most generally applicable laws aimed at criminal conduct.
It would not, for example, undermine the longstanding rule that
the First Amendment does not protect speech facilitating illegal
activity.367 A piece of false advertising is incompatible with the
government’s compelling interest in protecting consumers.368
Similarly, a Craigslist ad that solicits prostitution or offers to sell
illegal drugs is incompatible with the interest in criminalizing the
conduct that the barred speech proposes. It would be incompatible
with the government’s purpose in criminalizing homicide, property
damage, or the like to permit those crimes on the claim the conduct
underlying the violation was expressive. Where a law is clearly
aimed at an important governmental interest unrelated to
expression and any claimed harm to the speaker punished under
the law’s expressive interest is truly de minimis, an
incompatibility test would not stand in the state’s way every time
a defendant proffers a First Amendment defense.
365. See supra notes 158–161 and accompanying text (noting that the
National Mall had been previously used for “expressive sleeping” purposes prior
to the Petitioner’s request).
366. See, e.g., Thomas v. Collins, 323 U.S. 516, 529–30 (1945) (suspending
usual presumption of constitutionality and placing burden of persuasion on
government where economic legislation adversely affected the “freedoms secured
by the First Amendment”).
367. See Giboney v. Empire Storage & Ice Co., 336 U.S. 490, 502 (1949) (“[I]t
has never been deemed an abridgment of freedom of speech or press to make a
course of conduct illegal merely because the conduct was in part initiated,
evidenced, or carried out by means of language, either spoken, written, or
368. See, e.g., Va. State Bd. of Pharmacy v. Virginia Citizens Consumer
Council, 425 U.S. 748, 781 (1976) (“[T]he elimination of false and deceptive claims
serves to promote the one facet of commercial price and product advertising that
warrants First Amendment protection its contribution to the flow of accurate and
reliable information relevant to public and private decision-making.”).
An incompatibility analysis of content-neutral restrictions
would also prove flexible enough to accommodate the parade of
horribles that many claim would follow if those restrictions were
to receive greater scrutiny because of their effects on speech.369
Facially neutral government considerations such as traffic flow
and safety could justify denials of parade permits, for example, so
long as the chosen mode of expression and its chosen time and
place were truly incompatible with the government’s interest in
denying such requests for uses of public space.370 As Justice
Marshall wrote when expounding on the incompatibility principle
[T]wo parades cannot march on the same street simultaneously,
and government may allow only one. A demonstration or parade
on a large street during rush hour might put an intolerable
burden on the essential flow of traffic, and for that reason could
be prohibited. If overamplified loudspeakers assault the
citizenry, government may turn them down. . . . Although a
silent vigil may not unduly interfere with a public library,
making a speech in the reading room almost certainly would.371
Under the current “substantial government interest” prong of
the content-neutral intermediate scrutiny standard, courts barely
scrutinize the substantiality of the government’s asserted interest
at all.372 Justice Marshall’s Grayned opinion also suggests another
potential feature of an incompatibility test: the validity of a
particular law under the test will place sharper focus upon the
government interest asserted in defense of that law.373 Where a
369. See Redish, The Content Distinction, supra note 157, at 142 (“Proponents
of the content distinction [between content-based and content-neural restrictions]
may be concerned that increasing the level of judicial scrutiny for content-neutral
restrictions may result in a generally reduced skepticism for all content-based
Supra note 353 and accompanying text.
Grayned v. City of Rockford, 408 U.S. 104, 117 (1972) (internal citations
372. See William E. Lee, Lonely Pamphleteers, Little People, and the Supreme
Court: The Doctrine of Time, Place, and Manner Regulations of Expression, 54
GEO. WASH. L. REV. 757, 782–84 (1986) [hereinafter Lee, Lonely Pamphleteers]
(“[M]easuring the substantiality of the government’s interest is not a critical part
of the Court’s time, place and manner methodology. The Court rarely tells
legislatures or Congress that their concerns are insubstantial; therefore the
balance usually will be struck in favor of governmental interests.”).
Grayned, 408 U.S. at 117.
speaker’s First Amendment rights are implicated, courts should
not assume government interests are substantial simply because
the laws purportedly supporting those interests were products of
the legislative process. Speech interests cannot be put to majority
An incompatibility test can also solve one of the First
Amendment’s most bedeviling doctrinal problems. Since the
mid1980s, the Supreme Court has collapsed the review of time, place,
and manner regulations, and of regulations infringing symbolic
speech into a single strand of intermediate scrutiny that includes
ample alternatives analysis.375 Much harm to speech has
followed.376 Incompatibility analysis might untie these two lines of
Specifically, when characterizing the barred speaker’s chosen
communicative mode for incompatibility analysis purposes, the
court might first categorize the mode according to a distinction
that Susan Williams has drawn as “communicative” versus
“facilitative.”377 In a symbolic speech case such as United States v.
374. See Lee, Lonely Pamphleteers, supra note 372, at 784 (illustrating that
scholars have criticized “the minimal scrutiny the Court applies to the
substantiality of the government's interest,” stating courts have given substantial
weight to government interests merely because they were “not imaginary” and,
thus, have failed to honestly “weigh the interest against the impact on freedom of
375. R.A.V. v. City of St. Paul, 505 U.S. 377, 386 (1992) (“[T]he O’Brien test
differs little from the standard applied to time, place, or manner restrictions”); see
also Williams, supra note 149, at 619–20 (“The Court, arguing that the two
standards were always functionally identical, has melded them into one test.”
(citing Ward v. Rock Against Racism, 491 U.S. 781, 798 (1989))); Clark v. Cmty.
for Creative Non-Violence, 468 U.S. 288, 298 (1984) (noting that “the four-factor
standard of United States v. O'Brien, 391 U.S. 367 (
), for validating a
regulation of expressive conduct” is “little, if any, different from the standard
applied to time, place, or manner restrictions”).
376. See Williams, supra note 149, at 620 (“Although the two separate lines
of doctrine were far from robust even when they were independent, the new
unified doctrine provides even less protection for speech.”).
377. See id. at 660–61 (defining the “communicative” mode as conveying a
message “through symbols that represent [the] message” and the “facilitative”
mode as “part of the speech activity which the speaker uses to aid in the
transmission . . . of the message, but which does not itself play a role in the
O’Brien or Clark v. CCNV,378 the mode of expression—burning a
draft card, sleeping in Lafayette Park, or burning an American
flag379—the conduct at issue is communicative, as it plays “a role
in the representation of the message.”380 In many cases, the
speaker’s choice of mode is itself expressive; upholding the
restriction in such a case on ample alternatives grounds forces the
speaker to change her message to something different
altogether.381 It is in these symbolic speech cases where ample
alternative channels analysis does the most harm to speech.382
Alternatively, in a case traditionally characterized as
involving the time, place, or manner of expression, the choice of
mode is facilitative of the underlying expression.383 The mode aids
in the message’s transmission and is intended to expand the
potential audience for the speech—the use of loudspeakers on a
sound truck384 or of one’s own PA system rather than the one a host
provides385—but it is not itself communicative. Here, finding the
representation of the message”).
468 U.S. 288 (1984)
Texas v. Johnson, 491 U.S. 397, 406 (1989).
Williams, supra note 149, at 661–63.
381. See id. at 644 (“In a true symbolic speech case . . . [where] the
communication takes place through symbolic action––the regulation would have
to be aimed at . . . non-speech activities, rather than . . . content categories of
speech, and the government’s purpose would have to be to prevent some
noncommunicative harm caused by such activities.”).
382. See id. at 654 (“[T]he range of doctrinal tools available to deal with
complex first amendment problems has been reduced, and real first amendment
protections have been lost.”).
383. See id. at 706 n.330 (“If the physical activity about which the government
is concerned is expressive, we have a symbolic speech case; if it is facilitative, we
have a [time, place, or manner] case.”).
384. See, e.g., Kovacs v. Cooper, 336 U.S. 77, 78 (1949) (examining the
constitutionality of an ordinance that made it unlawful to use a sound truck for
“advertising purposes, or for any other purpose whatsoever, on or upon the public
streets, alleys or thoroughfares”).
See, e.g., Ward v. Rock Against Racism, 491 U.S. 781, 784 (1989)
This case arises from the city’s attempt to regulate the volume of
amplified music at [an amphitheater] so the performances are
satisfactory to the audience without intruding upon those who [reside]
in its vicinity. The city’s regulation requires [amphitheater] performers
to use sound-amplification equipment and a sound technician provided
by the city.
speaker could have used an alternative mode may do less harm to
the speaker’s expressive right.
And a third kind of case, not delineated by Williams but highly
relevant, involves government infringements of the speaker’s
choices of how, when, and where to speak that do not constitute
symbolic speech, but nevertheless go to the core of the expressive
act to a much greater degree than in the run-of-the-mill time,
place, or manner restriction on facilitative conduct. Here, recall the
kinds of cases discussed in Parts III.B.1. and B.4. supra: public
protests at specific events intended for specific audiences, of the
type discussed in the Menotti386 and Bl(a)ck Tea Society387 cases, or
on-site abortion counseling of the type discussed in McCullen.388
We might call these cases “associative conduct” cases, because
though they do not involve symbolic speech, the relevant speech’s
intended message and effects, along with its particular audience,
are inextricably associated with the message’s mode, time, and
place—so much so that communicative content can be ascribed to
the speech-accompanying conduct.
Incompatibility analysis could take into account these three
categories by granting greater or less deference to the speaker’s
choice of expressive mode depending on where along the
communicative-associative-facilitative continuum the mode falls.
In other words, a presumption in favor of the speaker could be
applied where the conduct in question is communicative or
associative prior to determining whether permitting the mode
would be incompatible with the government interest at issue. The
more communicative a speaker’s choice of mode, the more likely
the content-neutral restriction that has infringed upon that mode
will be found to have violated the speaker’s First Amendment
Adopting incompatibility analysis can encourage courts to
closely analyze the role that speech-accompanying conduct plays
in a speaker’s expressive act. Under current doctrine, it is a court’s
characterization of a particular law as content-based or
content386. Menotti v. City of Seattle, 409 F.3d 1113 (9th Cir. 2005).
387. Bl(a)ck Tea Soc’y v. City of Boston, 378 F.3d 8 (1st Cir. 2004).
388. McCullen v. Coakley, 844 F. Supp. 2d 206 (D. Mass. 2012).
neutral, not the speaker’s intent, which decides the answer to that
C. Incompatibility and Underinclusivity in the Review of Content
Another important doctrinal advance that incompatibility
analysis would accomplish is to place underinclusivity in the
foreground when analyzing content-neutral speech restrictions.
Despite the fact that scrutiny of such restrictions is nominally
rigorous, courts have failed to take underinclusivity seriously in
analyzing content-neutral laws.390 First Amendment review of sign
regulations offer a good example.
Municipalities justify signage restrictions based on
government interests in traffic safety and aesthetics.391 The local
ordinance at issue in the abovementioned Reed facially
distinguished between signs based on the message that those signs
conveyed and was thus, as the Supreme Court eventually found,
clearly content-based; in effect, the size of a sign or the length of
time a sign could be shown depended on what that sign said.392 As
noted above, the Town of Gilbert had argued to the Supreme Court,
consistent with the Ninth Circuit’s holding in its favor, that its
ordinance was content-neutral; in offering the government
interests supporting the ordinance, the Town claimed that it
389. See Redish, The Content Distinction, supra note 157, at 121–27
(discussing the development of the content-based and content-neutral distinction
within the Supreme Court).
390. See Harnish v. Manatee County, 783 F.2d 1535, 1539 (11th Cir. 1986)
(finding that bans on portable signs that were justified for aesthetic reasons were
not fatally underinclusive, even though such signs “represent[ed] a small fraction
of the total number of sign advertisements” in those cities); Mark Cordes, Sign
Regulation After Ladue: Examining the Evolving Limits of First Amendment
Protection, 74 NEB. L. REV. 36, 67 (1995) (“Potential problems arise, however, with
regard to content-neutral restrictions which prohibit or more severely restrict
particular types of signs within the same area, thus posing underinclusiveness
391. See Cordes, supra note 390, at 1 (recognizing one primary concern of
municipal efforts to regulate signs and billboards to be “supporting regulation,
most notably traffic safety and aesthetics”).
392. See Reed v. Town of Gilbert, 135 S. Ct. 2218, 2227 (2015) (finding the
town ordinance “content based on its face”).
“serve[d] to minimize visual clutter and confusion for people
traveling to an event that has already concluded.”393
The Town’s arguments as to the relationship between these
two asserted state interests and the means taken to further them,
while claiming to apply intermediate scrutiny to its own ordinance,
sound more in rationality review. As discussed supra,
contentneutral intermediate scrutiny is, in effect, rational basis review,
and those arguments thus would likely have been sufficient to
uphold the Town’s restrictions if the Ninth Circuit had deemed the
ordinance content-neutral.394 However, the interests in preserving
such “visual clutter” or in protecting “confuse[d] travelers”395 would
not survive incompatibility analysis. As to the “visual
clutter”related interest, the permissibility of a range of other signs—
political signs (the display of which had to be allowed under state
law), or signs that the Town called “Ideological Signs,” whose use
was much less restricted under its ordinance—shows that the
ordinance’s limits on a particular subset of signs is drastically
underinclusive.396 And as to the “confused traveler”-related
interest,397 travelers can be confused for a range of reasons, the
overwhelming majority of which have nothing to do with a local
church sign promoting an event in that community that has
already passed. In both cases, the underinclusivity of the
ordinance demonstrates that the restricted speech at issue is not
393. Brief for Respondents, supra note 323, at 4. The Court’s cases have
expressed sympathy for such arguments. See, e.g., Lehman v. City of Shaker
Heights, 418 U.S. 298, 304 (1974) (plurality opinion) (explaining that “[w]here a
city consciously has limited access to its transit system advertising space in order
to minimize chances of abuse, the appearance of favoritism, and the risk of
imposing upon a captive audience, these are reasonable” restrictions and there is
no constitutional violation); id. at 307 (Douglas, J., concurring) (“In my view the
right of the commuters to be free from forced intrusions on their privacy precludes
the city from transforming its vehicles of public transportation into forums for the
dissemination of ideas upon this captive audience.”).
394. See supra Part III.A (discussing self-autonomy theory and the
395. Brief for Respondents, supra note 323, at 48.
396. See Reed v. Town of Gilbert, 707 F.3d 057, 1061 (9th Cir. 2013) (“Section
4.402(D) lists nineteen different types of signs that are allowed without a permit.
Three of the types of exempted signs are of particular relevance: ‘Temporary
Directional Signs Relating to Qualifying Event,’ ‘Political Signs,’ and ‘Ideological
397. Brief for Respondents, supra note 323, at 48.
incompatible with the relevant state interests, and the ordinance
By focusing on the potential underinclusivity of government
action suppressing speech, incompatibility analysis will force the
government to act much more narrowly when burdening speech
through content-neutral restrictions. For instance, while a
generalized interest in aesthetics and visual clutter will often be
insufficient to demonstrate sufficient incompatibility for a law to
survive First Amendment scrutiny, preservation of a particular
area’s historic or aesthetic character might be.398 While this kind of
incompatibility-based tailoring may or may not be as demanding as
the least restrictive means requirement that is applied to
contentbased restrictions—a test whose applicability to the content-neutral
context the Court has rejected399—it will hold the government to its
obligation to limit as little speech as possible when acting.
Incompatibility will ensure that the burden of persuasion remains
on the state to justify even an incidental restriction.400
In 1939’s Schneider v. State of New Jersey,401 the Supreme
Court was faced with four challenges to municipal ordinances
passed by cities in California, Massachusetts, New Jersey, and
Wisconsin.402 These laws prohibited or restricted distributing
398. See, e.g., Messer v. City of Douglasville, 975 F.2d 1505, 1511 (11th Cir.
1992) (finding a restriction on signs within 300 feet of historic site to be
399. See Volokh, supra note 26, at 64 (“While this principle [that the speaker
chooses his means of communication] generally applies when the law restricts the
content of speech . . . and thus interferes with the speaker's choice of content, it
generally doesn't apply when a content-neutral law restricts the manner of
speech, [interfering] with the speaker's choice of manner.”).
400. Cf. Schad v. Borough of Mount Ephraim, 452 U.S. 61, 79–80 (1981)
(Stevens, J., concurring in judgment) (“[I]f one starts . . . from the premise that
appellant’s claims are rooted in the First Amendment, it would seem reasonable
for the Borough to overcome a presumption of invalidity.”).
308 U.S. 147 (1939).
402. See id. at 153–54 (1939) (“Four cases are here, each of which presents the
question whether regulations embodied in a municipal ordinance abridge the
freedom of speech and of the press secured against state invasion by
the Fourteenth Amendment of the Constitution.”).
handbills or other preprinted material, regardless of the
distributor’s identity or the handbill’s content.403 In all four cases,
the convictions for violating the ordinances were affirmed by the
courts below, which found that the laws struck the proper balance
between “the right of free expression” and the cities’ power to pass
“reasonable” regulations supported by the governmental interests
in preventing “littering of the streets” and “protecting the
occupants” of homes “from disturbance and annoyance.”404 Two of
those courts, foreshadowing the dispositive role that alternative
channels analysis would come to play in First Amendment doctrine,
upheld the ordinances in question on the ground they excluded only
“the public streets” from handbill distribution, and “leave open for
such distribution all other places in the city, public and private.”405
The Supreme Court invalidated all four ordinances.406 After
noting that all of the ordinances were content-neutral but
nevertheless burdened free expression,407 the Court took particular
exception to the claim that the ordinances permitted handbillers to
distribute their messages in other ways, and that this fact saved the
laws’ constitutionality.408 The Court proclaimed that “one is not to
have the exercise of his liberty of expression in appropriate places
abridged on the plea that it may be exercised in some other place.”409
In the fifty years since United States v. O’Brien, the Court has
abandoned this pronouncement. In its place, the Court has
403. See id. at 156 (“An ordinance of the City of Worcester, Massachusetts,
provides: ‘No person shall distribute in, or place upon any street or way, any
placard, handbill, flyer, poster, advertisement or paper of any description.’”).
404. Id. at 165.
405. See id. at 157 (quoting the Massachusetts court’s decision upholding the
Worcester ordinance); see also id. at 163 (referring to the Los Angeles ordinance).
406. See id. at 160 (“Although a municipality may enact regulations in the
interest of the public safety, health, welfare or convenience, these may not abridge
the individual liberties secured by the Constitution to those who wish to speak,
write, print or circulate information or opinion.”).
407. See id. at 163 (noting that one of the ordinances “bans unlicensed
communication of any views or the advocacy of any cause”).
408. See id. (stating that the streets are an appropriate place to distribute
printed manner to the public and just because one could theoretically distribute
those materials elsewhere does not mean that the ordinance is constitutionally
409. Id.; see also Southeastern Promotions Ltd. v. Conrad, 420 U.S. 546, 556
(1975) (citing Schneider for the very same proposition); Spence v. Washington,
418 U.S. 405, 411 n.4 (1974) (per curiam) (same).
entrenched an approach that permits an entire category of burdens
on speech-related conduct on the ground that those burdens are
instrumental rather than substantive. And, in support of this
approach, it has ratified an inquiry that permits judicially created
speech norms to override speaker communicative choice—the very
kind of governmental imposition on expression that the First
Amendment was designed to protect speakers from.
A return to Justice Roberts’ Schneider First Amendment
baseline will likely result in more protestors sleeping in public parks
to raise awareness about homelessness; more Hare Krishnas
attempting to hand us leaflets as we stroll along the fairgrounds or
rush to our airport gates; and more political signs in our public
rightof-ways. These are minor prices, and they are well worth paying for
a society that is committed to free expression. It also likely means
that women seeking to exercise their right to choose will be
confronted by anti-abortion activists who believe that those women
may be about to make a tremendous mistake,410 or that a candidate
or her supporter seeking to persuade a voter need not stand back
one-hundred feet from the entrance of the voter’s polling place.411
Perhaps some of us might be more equivocal about those prices. But
again, they are worth paying in a society that is committed not only
to the individual’s liberty to decide what to say, but also of how one
may say it.
It is indeed so that the First Amendment feeds “[h]umanity’s
innate desire for truth.”412 But the Speech Clause also leaves to each
of us to choose how to fulfill that desire, and to find that truth. In
analyzing content-based restrictions on speech, the Supreme Court
insists that “the First Amendment mandates that we presume that
speakers, not the government, know best both what they want to
say and how to say it.”413 By abandoning ample alternative channels
analysis, the Court can respect that same principle in its review of
II. The Birth of Ample Alternative Channels Analysis: Justice Harlan's O'Brien Opinion..................1670 A. Briefing , Argument, and the First Draft Majority Opinion ..................................................... 1671 B. Justice Harlan's Response: Ample Alternative Channels is Born..................................1675 C. Chief Justice Warren's Revised Majority Opinion and Justice Harlan's Withdrawal of His Draft Concurrence ...................................................1685
III. The Conflict....................................................................1689 A. Conflict with Theory................................................ 1689 1 . Self-Autonomy Theory ....................................... 1689 2 . Marketplace Theory........................................... 1695 B. Conflict Within Case Law ....................................... 1705 1. The “Free Speech Zone” Cases........................... 1706 2. The Adult Entertainment Cases ....................... 1711 3 . The “Prohibited Means” Cases .......................... 1714 4 . The Abortion Clinic Protest Cases ....................1717 C . Conflicts to Come: Online Speech ........................... 1722 D. The Content Based vs . Content-Neutral Fallacy....1724
IV. The Solution: Incompatibility........................................1728 A. The Test ................................................................... 1729 B. Untying Doctrinal Knots .........................................1733 C. Incompatibility and Underinclusivity in the
accepted tenet of first amendment doctrine”) . 25. Daniel A. Farber & John E. Nowak , The Misleading Nature of Public
Forum Analysis: Content and Context in First Amendment Adjudication , 70 VA.
L. REV . 1219 , 1237 ( 1984 ). 26 . Eugene Volokh , Speech as Conduct: Generally Applicable Laws , Illegal
Courses of Conduct, “ Situation-Altering Utterances ,” and the Uncharted Zones, 90
CORNELL L. REV . 1277 , 1305 ( 2005 ). 27 . See Stone, supra note 10 , at 67 ( “In some cases, these 'time, place , and
impact on free expression.”) . 28 . See id. at 68 ( “In light of the availability of alternative means of
the total quantity of public debate . ”) . 29. Richard A. Posner, Free Speech in an Economic Perspective , 20 SUFFOLK
U.L. REV . 1 , 16 - 17 ( 1986 ) [hereinafter Posner, Free Speech in an Economic
Perspective]; see also RICHARD A . POSNER, FRONTIERS OF LEGAL THEORY 76 ( 2001 )
(explaining the impact of modern thinking on free speech). 39. See id. (concurring with majority's determination of when a government
alternative ways for O'Brien to communicate his message ). 40 . See Ashutosh Bhagwat , The Test that Ate Everything: Intermediate
Scrutiny in First Amendment Jurisprudence , 2007 U. ILL. L. REV . 783 , 790 ( 2007 )
Interest , the Fall of Judicial Scrutiny , 37 HASTINGS L.J. 439 , 490 ( 1986 )
process” in assessing content-neutral regulations ). 41. Infra Part II. 42. Infra Part III. 43. Infra Part III. 49. See , e.g., Geoffrey P. Miller , The True Story of Carolene Products , 198
SUP. CT. REV . 397 , 398 - 99 ( 1987 ) (discussing a footnote about strict scrutiny of
Act”); David A. Strauss , Is Carolene Products Obsolete?, 2010 U. ILL. L. REV . 1251 ,
1263- 65 ( 2010 ) (discussing the doctrinal implications of Justice Stone's footnote
in Carolene Products) . 50 . See Miller , supra note 49 , at 398 ( listing numerous legal doctrinal
not the stuff of great decisions”) . 51 . See United States v. O'Brien , 391 U.S. 367 , 369 ( 1968 ) (“David Paul
for the District of Massachusetts.”) . 52 . See id. at 389 (Harlan, J., concurring) ( “O'Brien manifestly could have
conveyed his message in many ways other than by burning his draft card . ”) . 53 . See Justice Harlan, Concurring Draft Opinion Circulated Apr . 1968 : 59 . Id . at 19 . 60. Id. at 19- 20 . 61 . Brief for O 'Brien at 11, United States v. O'Brien , 391 U.S. 367 ( 1968 ),
(Nos. 232, 233) [hereinafter O'Brien Brief]. 62. Id. at 40 . 63. Transcript of Oral Argument at 10, United States v. O'Brien , 391 U.S.
367 ( 1968 ) (Nos . 232 , 233); see also id. at 11 (stating the SSA is “not fairly to be
(emphasis added)) . 64 . Chief Justice Warren, Majority Draft Opinion Circulated Apr . 12 , 1968 :
United States v. O'Brien at 1 in Harlan Papers, Box 311 [hereinafter Majority
First Draft ] (on file with the Washington and Lee Law Review) . 65 . See id. at 18 (“[T] he core concern of the [First] Amendment is . . . verbal
expression . . . .”). 66 . See id. at 7-8 (“The 1965 Amendment [subjecting to criminal liability
no connection with speech . ”). 82. Id. at 21. 83. Id. at 22. 84. Id. at 23 . 85. See Justice John M. Harlan , First Draft Concurring Opinion Dated May
1968: United States v. O'Brien and Justice John M. Harlan , First Draft
Concurring Opinion Circulated May 1 , 1968 : United States v. O'Brien
[hereinafter First Draft Concurrence] in Harlan Papers, Box 311 ( “I find myself
O'Brien 's conviction for draft card burning , pursuant to 50 U.S.C. App. § 462
(b)(3), did not violate his right of free speech assured by the First Amendment .”).
The quotations cited here are from the draft opinion dated “ May 1968 ,” but the
two drafts cited in this section are materially similar . 86. Id .